« Back to Results
AEA Poster Session
Friday, Jan. 7, 2022
7:00 AM - 6:00 PM (EST)
Saturday, Jan. 8, 2022
7:00 AM - 6:00 PM (EST)
Sunday, Jan. 9, 2022
7:00 AM - 3:00 PM (EST)
American Economic Association
"The Public Banking Act " Should Allow for Individual Federal Reserve Bank Accounts To Transform Monetary Policy (E5, E4)
This research, which is based on the author’s forthcoming book Distorted Money Flow, examines the diversion of the flow of money in recent decades from Main Street to Wall Street that has resulted in a buildup of both private debt and public debt without which the people on Main Street would be unable to buy back the value of the goods and services they are producing. and examines a way of achieving much tighter and immediate control of overall consumer demand. This is achieved by modifying The Public Banking Act as recently proposed in Congress to re-create the postal banking system that existed from 1910 until 1966 and creating a digital cybersecurity block-chained postal Federal Reserve smartphone bank account for every American.
To focus these accounts on consumers with a high marginal propensity to consume, these accounts could be restricted to one account per Social Security number and interest could be earned only on some base amount (e.g. $10,000 or less). Initially, $1,000 would be put into each account, which could not be withdrawn except by heirs after death. However, account holders could add additional money up to the base amount, which could be withdrawn at any time along with interest earned. Money above the base amount would earn no interest.
Whenever excessive inflation threatened, the Federal Reserve could set high interest rates for these accounts to encourage savings and discourage consumer demand, while keeping interest rates somewhat lower in the financial markets to encourage an increase in the supply of goods and services to tightly control inflation.
In the face of an economic downturn, interest rates could be lowered and the Federal Reserve could inject money into these individual Federal Reserve bank accounts to stimulate consumer demand as necessary to restore full employment.
'An Agent-based Stress Testing Framework of Central Clearing (G4, G2)
We put forward a partial equilibrium model of a central counterparty (CCP) and clearing members (CM). Agents are profit optimizing and are exposed to market risk. We simulate the balance sheets and the leverage of the CCP and the CMs as a result of the market price movements. We also model the margin shortfalls and the entire default waterfall in case of CM defaults. Moreover, we analyze the stability of the network of counterparties in a system with and without a CCP. We compare the stability of the networks with and without a CCP by analyzing i) endogenous feedback effects between CM defaults and market liquidity, ii) the contagion to non-defaulting CM as well as iii) the contagion from service functions that CMs provide to the CCPs, such as liquidity provision or collateral and investment services.
We find that in a system with a single CCP overall less defaults occur than in a system without a CCP. In addition, a CCP mitigates contagion measured by lower amount of 2nd-round defaults. We also find that mark-to-market price feedbacks have considerable adverse impact on the stability of the system, while that the mutualization of losses has a strong stabilizing effect.
A Distributional Analysis of Gender Gaps in Wages and Numeracy Skills (J1, I2)
This paper investigates gender-specific patterns in the distribution of numeracy skills among
adults, measured by standardized tests in the international PIAAC survey. We report three
patterns: first, we document that both wages and returns to skills are higher for individuals
with higher numeracy skills. Second, we show that women with the highest numeracy
levels experience lower returns to skills compared to men with the same numeracy levels.
Lastly, we find that women in general have lower numeracy skills than men, especially at
the top of the numeracy distribution. In order to shed light on potential determinants,
we perform a decomposition of gender gaps across the numeracy distribution and find that
especially children and fields of study play an important role. These findings contribute to
the discussion about glass ceilings for women in the labor market and emphasize the role of
potential labor-market discrimination of women.
A Free and Fair Economy: A Game of Justice and Inclusion (C7, D7)
Frequent violations of fair principles in real-life settings raise the fundamental question of whether such principles can guarantee the existence of a self-enforcing equilibrium in a free economy. We show that distributive justice basic principles ensure that a pure-strategy equilibrium exists in an economy where agents freely (and non-cooperatively) choose their inputs and derive utility from their pay. These principles hold that: 1) your pay should not depend on your name, and 2) a more productive agent should not earn less. When these principles are violated, an equilibrium may not exist. Moreover, we uncover an intuitive condition---technological monotonicity---that guarantees that the equilibrium is unique and Pareto-efficient. Significantly, our findings generalize to economies with social justice, implemented in the form of progressive taxation and redistribution, and guaranteeing a basic income to unproductive agents. Our analysis uncovers a new class of strategic form games by incorporating normative principles into non-cooperative game theory. Our results rely on no particular assumptions, and our setup is entirely non-parametric. Illustrations of the theory include applications to contagion and self-enforcing lockdown in a fair networked economy, exchange economies, and bias in the academic peer-review system.
A Joint Top Income and Wealth Distribution (C5, D3)
Top distributions of income and wealth are still incompletely measured in many national statistics, particularly when using survey data. This paper develops the technique of incorporating the joint distributional relationship to enhance the estimation of these two top distributions by using the best data available for Germany. We leverage the bivariate copula to extrapolate both income and wealth distributions from German PHF (Panel on Household Finance) data under the incidental truncation model. The copula modelling grants the separability in choosing the estimation domain as well as the parametric specification between the marginal distribution and dependence structure. One distinct feature of our paper is to complement the model fit with external validation. The copula estimate can help us to perform out-of-sample prediction on the very top of the tail distribution from one margin conditional on the characteristics of the other. The validation exercises show that our copula-based approach can approximate much closer to the top tax data and wealth “rich list” than those unconditional marginal extrapolations. The data and effectiveness of our copula-based approach also verify our presumption of incidental truncation and differential detectability in the top lists.
A Multi-Dimensional and Dynamic Screen Tool of Financial Toxicity Among Cancer Patients (I1, I3)
Increasing cancer care costs have led to a tremendous economic burden on society and severe financial hardship for many cancer patients and their caregivers. For any evidence-based assistance to ease the burden, a systematic screening to proactively identify those patients at risk for financial hardship is essential. However, systematic financial toxicity measures are limited for routine screening due to data availability in actual clinical practice and constrained clinical time, which is often not feasible to monitor the financial hardship dynamically across the cancer survivorship continuum.
To bridge the gap between literature and practice, we develop a composite financial toxicity measure that is practical to collect by the current health system. Using electronic health records, clinic screen and cancer registry data collected from the University of Virginia Health System, we consider all three conceptual domains of financial toxicity, including material conditions (e.g., income, insurance, and payment histories), coping behaviors (i.e., treatment adherence), and psychological distress.
Based on multidimensional poverty index literature, we use a fuzzy set approach to convert outcomes with different scales into a uniform scale between 0 and 1, where 0 signals the absence of deprivation and 1 signals absolute deprivation. Next, a composite index will be constructed via stochastic multicriteria acceptability analysis that accounts for the individual heterogeneity in the sets of weights assigned to different financial hardship domains. Using the longitudinal nature of the data, we will also apply this composite index to assess the same patient’s financial toxicity level multiple times beyond the initial cancer diagnosis and identify risk factors using supervised machine learning algorithms.
Identifying financial toxicity based on existing medical records will greatly improve screening scalability and sustainability. The insight into dynamic financial toxicity patterns is essential for identifying both the magnitude and the scope of this evolving healthcare financing problem for cancer care.
A Neural Network Ensemble Approach for GDP Forecasting (E3, C5)
We propose an ensemble learning methodology to forecast the future US GDP growth release. Our approach combines a Recurrent Neural Network (RNN) with a Dynamic Factor model accounting for time-variation in mean with a Generalized Autoregressive Score (DFM-GAS). The analysis is based on a set of predictors encompassing a wide range of variables measured at different frequencies. The forecasting exercise is aimed at evaluating the predictive ability of each model's component of the ensemble by considering variations in mean, potentially caused by recessions affecting the economy. Thus, we show how the combination of RNN and DFM-GAS improves forecasts of the US GDP growth rate in the aftermath of the 2008-09 global financial crisis. We find that a neural network ensemble markedly reduces the root mean squared error for the short-term forecast horizon.
A New Theoretical Foundation for the Shape of Utility Functions over Wealth (D9, D6)
This paper proposes a new theoretical foundation for the shape of utility functions that holds regardless of contextual details. In particular, when decisionmakers face an underlying distribution of consumption values for which they allocate their wealth, their utility over that wealth is governed by that distribution. When the distribution has a Pareto tail, the utility function exhibits a constant elasticity. As its exponent approaches one following Zipf's Law, the utility function becomes approximately logarithmic. Based on Zipf's Law, a new utility function that nests canonical utility functions is proposed. As distributions induced by plausible stochastic processes replicate labor supply elasticities that are in accord with real-world data, this finding has practical benefits for economists.
A Pandemic Trap: Juggling Stringency between Unemployment and COVID-19 Cases (H1, I1)
Using simultaneous equations models with panel data of 50 U.S. states, we examine the bi-directional and dynamic relationships between COVID-19 cases and unemployment in response to stringency policy. These relationships are critical to controlling COVID-19 spread while preserving the economy. When COVID-19 cases increase, we find that stringency policy becomes more restrictive and is effective in lowering cases. However, a high unemployment rate leads the government to reduce stringency, since implementing it increases the unemployment rate. We also find that the governor’s approval ratings positively affect state-level stringency policy; unemployment is affected negatively by retail sales and positively by labor participation, and temperature negatively affects the infection cases.
A Quantitative Model of Corporate Reputation Building in Debt Markets and Firm Dynamics (G3, G2)
I estimate a heterogeneous firm model on dynamic adverse selection with screening and signaling in public debt markets for privately informed persistent productivity, and quantify the effect of asymmetric information on misallocation of capital and financial liabilities. The firm's manager chooses between public debt (e.g., corporate bonds), priced on the accumulation of borrower's history, and monitored private debt (e.g., bank loans), allowing costly monitoring to be an imperfect substitute for reputation building. Cross-subsidization caused by asymmetric information in public debt markets leads to overinvestment of low productivity firms and capital misallocation. A good reputation lowers borrowing costs from public lenders. The firm's manager uses leverage and equity to send a good signal. However, neither reputation building nor monitored private debt substantially remove capital misallocation. I find precise information associated with productivity increases total factor productivity, decreases the variance of the marginal product of capital, and increases consumption by 1.4%. Less demand for monitoring lowers aggregate bank debt ratio by 6 percentage points. In the counterfactual policy analysis, the taxation of debt forgiveness under Chapter 11 reorganization generates a rise in consumer welfare by reducing low productivity firms' incentive to overborrow.
A Structural Analysis of Simple Contracts (C0, L0)
This paper provides an econometric framework for analyzing simple contracts where an
agent chooses between a fixed-price option and a cost-reimbursement option provided by a
principal. First, we propose a consistent procedure for testing the null hypothesis of a corresponding cost function being linear, which is widely assumed for tractability in the literature. Motivated by the rejection of such a null based on our empirical data, next we establish nonparametric identification, without restricting the cost function to be linear, for all model primitives conditioned on the agent exerting nonzero effort. These primitives include agent's cost and disutility functions, distribution of agent efficiency type, and parameters that characterize agent's bargaining power and intertemporal preference. Moreover, we propose a consistent procedure to implement the identification results for estimation. In our empirical study, we find strong evidence against linearity of the cost function. The importance of this empirical finding is further evidenced by a welfare analysis, which shows the welfare assessment to be sensitive to the specification of cost function.
A Structural Model of Subjective Well-Being (I3, J2)
This study attempts to minor the gap of the literature of SWB by opening the black box of relative income and income aspiration in a uniformed framework that is empirically testable by data. Borrowing from the classic life-cycle labor supply models, we argue that aspirations can be decomposed into two components, income desire and consumption desire. Income desire is defined as the desired future income or work discounted to present, and consumption desire can be interpreted as desired future consumption discounted to present. Income desire depends on one's own current income, while consumption desire depends on the income of one's reference group. Moreover, neighborhood income affects one's reference group income. We justify that impacts of income on SWB have two channels of mechanisms: (1) increasing SWB directly through consumption and (2) decreasing SWB indirectly by boosting up one's income desire. The impacts of reference income on SWB transmit through two channels as well: (1) welfare loss due to status comparison and (2) positive effects due to revealing information on one's life-cycle consumption. The directions of the impacts of income and reference income on SWB remain to be an empirical issue.
Our model predicts the long-run impact of income can be isolated if we have information on income desire, reference income, and neighborhood average income. Using our unique survey data from more than 80% of all households in two villages in Shaanxi Province, China, we find the data fit our model as expected using life satisfaction as a measurement for SWB.Doubling one's reference income decreases life satisfaction by 0.1 - 0.6. We also find, despite the negative impacts of relative income and income desire, the long-run impact of absolute income on life satisfaction is positive and substantial. Each doubling of household income increase life satisfaction by 0.2-0.5 in the long run.
A Tale of Gold and Blood: The Unintended Consequences of Market Regulation on Local Violence (O1, Q3)
How can a small change in fiscal accountability boost violent disputes for valuable natural resources? In this paper, we investigate a regulatory change in Brazil that greatly reduced governmental monitoring capacity against gold laundering and we show how this affected violence in illegal gold-mining sites. Because the new regulation introduced in 2013 made it harder for authorities to find illegal gold transactions between miners and first-buyers, demand for it increased and boosted competition for illegal mining sites, leading to more violence. To verify this, we devise a theoretical model and test its implications using a unique database combining gold deposits, Indigenous Territories, Natural Conservation Areas, environmental crimes, deforestation, and homicide rates. With a difference-in-differences design, we find that municipalities more exposed to illegal mining experienced extra 8 homicides per 100,000 people - or an increase of roughly 20% - after the regulation was passed. Moreover, the level of illegal activity, as well as deforestation, increased in places exposed to illegal gold-mining after the regulatory change.
A Tale of Two Firms (E3, E5)
How aggressively can the Fed increase the credit supply during an economic crisis caused by the 2020 COVID-19 pandemic crisis without surging inflation? While the labor market has not been recovered yet, some part of the economy concerns about surging inflation. Should the Fed of the 2020s consider tapering expansionary policy already because of like-the-1970s-inflation concerns, while there are 2%p of the labor force without their paychecks due to a recession? If the aggregate price level has risen slower since the 1990s than it used to until the 1980s, as Sargent and Surico (2011) suggest, the Fed has more room to care about the labor market. According to Diez, Fan, Villegas-Sanchez (2019), there has been a polarization of markups among firms at least since 2000. Under the polarizing economic environment, the transmission mechanism of monetary policy to the aggregate price level might have been altered. This paper explores a dynamic general equilibrium model with two production sectors and pictures an alternative inflation scenario after expansionary monetary policy. The distinctive assumption is that there are heterogeneous firms: the high-end producers and the low-end producers. Consumers have different elasticities of substitution toward these two types of goods, and thus, the markups of the two types of goods diverge. In such an economic environment, an exogenous disutility shock to labor in the low-end sector, due to the fear of COVID-19, induces an expansionary response of the central bank by a Taylor rule. However, the change in price is weighed on the high-end sector which takes a smaller fraction of the aggregate consumption basket. In this way, the aggregate price level does not rise as much as a conventional dynamic general equilibrium model suggests. In a new polarizing economic environment, the Fed's current monetary policy seems well balanced between price stability and maximum employment.
A Theory of Debt Maturity and Innovation (G3)
The financing of innovative firms must balance two goals. On the one hand, since innovation is inherently risky, the entrepreneur must receive adequate protection after failure. Simultaneously, the firm must be liquidated when its assets can be redeployed more efficiently elsewhere. Meeting these two goals can be especially challenging when contracts are incomplete and shaped by renegotiation. I show how an appropriate choice of debt maturity, together with ex-post contract renegotiation, embeds a "put option" into the firm's capital structure. The put is exercised when liquidation is efficient, and it partially insures the entrepreneur against failure and thus motivates innovation. The theory has novel empirical implications for the financing patterns of innovative firms.
A Theory of Hedonic Experience and Time Allocation (D9, D0)
In standard consumer theory, preferences are regarded as primitive. People are assumed capable of ranking any two bundles, and rationality means choosing the most preferred among available alternatives. Failure to optimize is considered anomalous. But what if optimization is not usually the motivating force? What if, as suggested by some neuroscientists, preference itself does not lie at the heart of human cognition?
We present a theory of the human experience that is independent of preference, and can also be free of the optimizing impulse. Our primitive, akin to Jevons’s quantity of feeling, is a function of time, h_i(t), that captures an individual’s “instant hedonic experience” from engaging in activity i. When accumulated over time, the integral of this function captures “hedonic experience.”
An essential distinction is maintained between a subject, who owns h_i(t), and an observer who wishes to measure or estimate it. We call the observer’s measure of the primitive a “quantity of feeling” function, f_i(t). Hedonic experience is measured, from the observer’s perspective, by the integral of f_i(t) during an interval of time.
We prove the existence of a family of functions f_i(t) that characterize the subject’s hedonic experience. Our framework rests upon three assumptions: 1) a finite number of activities is under consideration; 2) a single activity is engaged at any time; and 3) the rate of change in f_i(t) is the difference between the proportional change in f_i(t) for activity i and the sum of proportional changes, f_j(t), in the quantity of feeling functions for the other activities, up to coefficient functions of proportionality. These functions, for the chosen activity and all non-chosen activities, act simultaneously on an individual at every moment in time.
We apply the theory to time-allocation decisions, showing how the framework can explain behavior either with or without imposing an optimizing impulse.
Access to Financial Navigation Services: Insights from a Discrete Choice Experiment (D9, I0)
High out-of-pocket spending on cancer treatments often causes financial problems for many families. Although several cancer centers offer financial navigation programs to help patients manage the financial impact, a large number of patients at risk fail to seek financial navigation services due to several barriers, such as being overwhelmed with cancer diagnoses and treatment decisions. Particularly, the low utilization rates are prevailing among racial minorities and people with low education levels. This study aims to address three research questions: 1) What limiting factors restrict cancer patients from seeking financial navigation? 2) What attributes and associated levels of financial navigation programs are appealing to patients and influence their participation decisions? 3) Whether the preferences vary across different groups of patients partitioned by selected characteristics.
From the University of Virginia Cancer Center, we survey patients on barriers of uptaking financial navigation programs and assess their stated preferences via a discrete choice experiment. Pairs of hypothetical programs are constructed using five attributes (i.e., type of financial navigation services, appointment scheduling methods, contact sources, meeting locations, and program starting time). In addition to collecting health and sociodemographic information, the survey also assesses patients’ financial status, psychological distress, health literacy and numeracy levels to test for systematic differences in preferences based on these characteristics.
Results from mixed-logit models show the trade-offs between program features and how choice probabilities vary with changes in attribute levels. The modeling process incorporates preference heterogeneity by interacting program attributes with selected characteristics and test whether there are significant differences between patient groups. Finally, the predicted choice probabilities will promote patient-centered care by identifying customized assistance strategies appealing to different groups of cancer patients, especially those disadvantaged populations. A better understanding of patients’ needs will help improve the designs of future financial assistance services.
Accounting for Transshipment: The Case of Kosovo (F1)
In November 2018, Kosovo imposed 100 percent tariff on goods imported from Serbia and Bosnia and Hercegovina. The tariff caused immediate and significant disruptions in regional trade flows. Imports from Serbia, the second largest trading partner of Kosovo after the EU, dropped to nearly zero by early 2019 and remained negligible thereafter. This paper empirically traces the transshipment from Serbia to Kosovo via neighboring countries. Using detailed monthly imports data for 2017-19 and difference‐in‐difference approach to identify products that were previously being exported from Serbia to Kosovo (treatment group), this paper provides evidence of Serbian exporters’ evasion of Kosovar tariff through transshipment via Macedonia and Albania.
Adaptive Importance Sampling for Large DSGE Models (E1, C1)
The paper introduces a new adaptive methodology for the estimation of dynamic stochastic general equilibrium (DSGE) models based on the Mixture of $t$ by Importance Sampling weighted Expectation-Maximization (MitISEM). The use of Importance Sampling and of an adaptive scheme based on Expectation Maximization allows estimation of large-scale models and highly complex forms. Simulation exercises show how our method achieves identification of posterior distributions in several cases, including standard New Keynesians and macroeconomic models. We apply the MitISEM to estimate a large-scale open economy model encompassing international trade between two countries, namely Canada and the US. For both countries, we consider a rich fiscal policy sector that includes two different types of public expenditure: productive and unproductive government spending. Our findings show that, in the presence of nominal rigidities, an increase in productive spending generates a crowding-in on domestic private consumption whereas unproductive spending induces a fall in domestic private consumption. Finally, we find that irrespective to the type of government expenditure, an increase in public spending for the domestic economy induces an exchange rate appreciation and an improvement in the trade balance.
Aggregate Fluctuations from Clustered Micro Shocks (E3, E2)
This paper investigates the hypothesis that idiosyncratic firm-level shocks account for macroeconomic fluctuations where they are correlated across firms within an industry (cluster). Most business cycle research has ignored the cross-firm correlation because a firm's cross-sectionally demeaned fluctuations are orthogonal to a cross-sectional mean and asymptotically uncorrelated to other firms. This argument breaks down if the variance-covariance of idiosyncratic shocks is heterogeneous across firms, as I find evidence on the heterogeneity from the US firm-level data. I show how the correlated idiosyncratic movements across firms within clusters lead to macroeconomic fluctuations. In the US, these clustered origins well explain the evolution of the aggregate GDP volatility. Also, their contributions to the GDP volatility start to rise from 5% to 20% in the recent two decades.
Aggregate Liquidity Allocation and Endogenous Uncertainty (E1, G0)
This paper provides a continuous time DSGE model to shed light on an endogenous mechanism of liquidity allocation between real economy and financial system. This theory is helpful for us to understand the weak investment and saving glut of non-financial corporates in main advanced economies during the current recovery from the Great Recession. Physical capitals have less liquidity than their corresponding equities. With financial frictions, investment depends not only on the equity price but also on the capital structure of corporates. When the net worth of entrepreneurs who are risk-averse are low, they will disinvest and hold more financial assets to hedge against risks. This leads to more funds flowing into financial system so that a financial boom with high system risks but a slow economic recovery.
Aggregate Uncertainty and the Micro-Dynamics of Firms (E3)
Using firm level micro-data, I find evidence that firms with lower growth prospects are more sensitive to aggregate shocks. I interpret these findings using a model of demand accumulation and endogenous entry and exit decisions, which I then estimate on French data. The resulting cyclical dynamics of firms provide an explanation for the observed counter-cyclical dispersion in firms' growth rates. They suggest that cyclical dispersion is the result of a pre-existing and persistent characteristic of the firm and caution against its use as a proxy for time-varying uncertainty. The estimated negative correlation between a firm's sensitivity to aggregate shocks and its expected future growth rate is shown to have important consequences for the cyclical characteristics of entering and exiting firms. The quantitative model suggests that this compositional effect is sizeable and equivalent to around 10.5% of the drop in aggregate employment between 2008 and 2009.
Ambiguity, Date Labels, and Food Waste (D9, D8)
The paper aims to understand how date labels (e.g., ‘best by’, and ‘use by’) would influence the consumer’s consumption decision. Existing studies have documented that date labels may cause consumers’ confusion about the edibility of food items and hence contribute to food waste. However, no study has investigated the underlying mechanism through which date labels may affect consumers’ food consumption decisions. This gap is important because legislators in the United States are contemplating regulating date labels to combat food waste (e.g., Food Date Labeling Act of 2019). The present study sheds light on the issue by theoretically modeling consumers’ decision under ambiguity and by using data collected from a series of experiments to test the hypotheses derived from the theoretical model.
We utilize an α-MaxMin model to depict a representative consumer’s decision making under ambiguity regarding whether or not to consume a food item with a certain date label. Testable hypotheses are derived from this theoretical. We conducted 16 series of laboratory experiments in two different cities in the United States (Auburn, AL and Ithaca, NY). The experiment includes an assessment of subjects’ ambiguity aversion parameters and solicitation of their food consumption decisions under various food item and date label scenarios. Results show that ‘use by’ date label will significantly increase waste of deli meat when compared with a ‘best by’ date label. A consumer who is more ambiguity averse would be more likely to discard deli meat labeled as ‘use by.’ These effects are not found for spaghetti sauce or eggs.
This study reveals the complexity of the interaction between date labels and consumers’ attitude toward ambiguity. It shows that simply changing food date labels might not be able to reduce food waste caused by consumers’ confusion on the information conveyed by these date labels.
An Explanation of Real U.S. Interest Rates with an Exchange Economy (E4, G1)
The paper offers an explanation of historical quarterly real 3-month US Treasury bond interest rates in the United States economy since 1975. This offers one solution to the puzzle of standard asset pricing models being unable to explain rates through fundamentals alone. The model adds in a premium that can be viewed as reflecting liquidity by extending a standard exchange economy through the optimal use of bank supplied exchange credit. The paper contributes the closest fit to date of real interest rates in an exchange economy without liquidity elements added into the utility function. It explains business cycle filtered real ex-post interest rates, as well as procyclic output and labor. It also explains closely the actual level of real interest rates by adding back in the HP filtered trend. Three alternative shock construction methods are used and compared within this monetary business cycle framework that has a sparse set of shocks including bank productivity for the exchange credit sector. Results include robustness and sensitivity analysis such as comparison to a money-only economy. The Bayesian and iterative convergence methods yield a high correlation of the model implied real interest rate with data, and a relative volatility near one.
Ancestral Connection and Alliances (Z1, G4)
This paper studies the cultural determinants of firm boundaries. We find that the historical ancestral composition of the area where firms locate plays an important role in their choices of partners when forming business alliances and of the location of the new venture. Partnering firms experience significantly better performance when the ancestral connection between their headquarters states is stronger. Using data on ancestral connections between partners’ key employees, we find that shared values and better communication between firms’ stakeholders (e.g., inventors), as opposed to connections between corporate leaders, likely underline the role of ancestral connection. Our results highlight the importance of ancestral connection in shaping firm boundaries, above and beyond geographic and economic determinants, thus the important role culture plays in inducing cooperation between firms when facing contractual incompleteness in the real business world.
Animal Spirits in Regulation: Evidence from Banking (G2, G4)
Are banking regulation and supervision consistent over time and space? We posit they are not, and that inconsistency in supervisory practice is influenced by variation in media sentiment toward banks. Further, we argue that this regulatory volatility is reflected in banks’ operational losses. We conjecture that this causal channel may lead to cyclicality in financial intermediation orthogonal to fundamentals.
To gauge variation in media sentiment toward banks, we perform sentiment analysis on a large textual corpus of US news publications; to measure regulatory response, we draw from proprietary Federal Reserve System data on confidential supervisory actions and bank operational losses, including legal damages, as well as court-docket information sourced from the Administrative Office of the United States Courts. Our empirical strategy relies on an IV approach to infer causality, using M&A activity among local media to instrument for potentially endogenous variation in media sentiment.
Are Historic Districts A Backdoor for Segregation? Yes and No (R2, J1)
Historic districts preserve the heritage of designated areas and attract tourism income. However, these districts also come with restrictions and increased housing prices that could cause segregation. We study how two historic district programs impact residential segregation in Denver. We find that homebuyers are more likely to be White within historic districts, but that the official historic designation has no effect on this probability. Similarly, we find that most transactions flow from White sellers to White buyers, regardless of official designation. Thus, while historic districts tend to be more segregated, official designation does not seem to amplify this existing problem.
Are Housing Rental Markets that Competitive? A Dynamic Monopoly Approach (D4, R0)
This paper develops a search model of the housing rental market with moving costs and imperfect information: the dynamic monopoly model. Landlords set the rent and, in equilibrium, there is an interval of rents charged for similar housing units, a consequence of landlords' market power. Within such interval, tenants turnover is higher the higher the rent charged, which results from a downward-sloping demand faced by individual landlords. The less elastic this demand is, the larger landlords' market power, and the larger the scope for public policy to regulate the market and improve efficiency. These implications are in stark contrast to those of the conventional perfect competition theory. Using data from the American Housing Survey, I create a data set of rental spells in New York City and estimate the average of that elasticity of demand, a measure of landlords' market power. To identify that estimate I use variation in rent that results from whether the housing units are rent controlled/stabilized, publicly owned or subsidized. The results suggest that the average NYC landlord has substantial market power. These findings are important for the public policy debate on housing inequalities, metropolitan housing crises and gentrification.
Are IMF Loans Diverted to Offshore Bank Accounts? (A Quarter Century of Evidence Suggests Not) (F3, H0)
This paper examines whether IMF lending is associated with increases in outflows to offshore
financial centers (OFCs) —known for bank secrecy and asset protection—relative to other
international destinations. Using quarterly data from the BIS on bilateral bank deposits, we are
unable to detect any positive and statistically significant effect of IMF loan disbursements on bank
deposits in OFCs. The result holds even after restricting the sample to the duration of the IMF
program, where disbursement quarters and non-disbursement quarters should be subject to similar
degrees of macroeconomic stress. It is also robust to using the scheduled tranching of disbursements
as an instrument for actual disbursements. While the effects vary by the type and conditionality of
the IMF program, as well as by the amount of lending, none of the effects are positive and
statistically significant. Using real-time data on international interbank flows from SWIFT, we also
examine whether the recent COVID-related surge in emergency lending is associated with an
increase in outflows to OFCs, but find no evidence to support this.
Are the Supporters of Socialism the Losers of Capitalism? Conformism in East Germany and Transition Success (P2, P3)
The empirical literature is inconclusive in whether democratization of a country goes together with a redistribution of economic resources (Acemoglu et al., 2015; Aidt et al., 2020). With newly available individual-level panel data on East Germany's socialist past, we analyze how former supporters and opponents of the system performed after reunification with the capitalist West. Opponents, who helped to overturn the communist system in the Peaceful Revolution, show increases in life satisfaction, income, and employment. Former members of the communist party and employees in the Stasi supervised area become less satisfied but have still higher incomes than the average East German population. The politically inactive silent majority lost strongly and show persistently lower levels of life satisfaction and labor market outcomes. Adverse effects are especially strong when individuals silently supported the communist system. Additional results show that conformism is also associated with political party preferences several decades after communism. The contribution to the literature on persistence of socialist elitism for economic outcomes are i) a novel dataset to analyze the results for the special case of East Germany, ii) the use of panel data to analyze differences in outcomes additionally to persistence effects, iii) the analysis of former dissidents of the system and their outcomes over a period of almost three decades. The paper relates to the economic literature on the long-lasting effects of socialism (Alesina and Fuchs-Schündeln, 2007; Becker, Mergele, and Woessmann, 2020) and the persistence of privileges of Communist elites (Geishecker and Haisken-DeNew, 2004; Vevcernik, 1995; Otrachshenko, Nikolova, and Popova, 2021).
Are You a Zombie? A Supervised Learning Method to Classify Unviable Firms and Identify the Determinants (G3, C5)
We examine the determinants of zombie companies using a comprehensive firm-level dataset of public corporations from Europe and the United States. We show that US zombie companies differ from their European peers on a number of firm-specific and industry-specific factors. Using decision trees, we document that income and leverage-related variables are among the main drivers classifying zombie companies in Europe, while dividends and stock-related variables are the most important in the US. We observe a frequent mislabeling of zombie firms into other unviable types of firms. To account for this, we also examine the determinants of distressed firms and compare them to the zombie. We find that zombie and distressed are not comparable types of companies, rather companies at a different stage of their financial unviability. We also document that zombification is especially a European phenomenon, while distressed-type of firms are mostly populating the US economy. We find no major differences in terms of zombie company-specific determinants between crisis and non-crisis periods.
Assessing the Impact of COVID-19 on Trade: A Machine Learning Counterfactual Analysis (F1, O1)
This paper aims to estimate the causal effect of the COVID-19 shock on a firm's probability of survival in the export markets and study the heterogeneity of this effect. The main hurdles for this evaluation task are related to the pervasiveness of the COVID-19 shock, which makes it impossible to find a control group, coupled with the complex interdependencies between firms and products belonging to different sectors and countries. This paper's first contribution is exploring and comparing the effectiveness of different Machine Learning (ML) techniques in predicting firms' trade status in two different scenarios, a COVID-19 and a non-COVID-19 setting. ML techniques have been successfully applied to predict firm performances and help companies (and public agencies) in their decision-making in complex environments. Up to what we know, this is the first time that ML techniques are used to predict firm-level international trade performance. This paper's second contribution is to use these predictions to estimate the causal effect of the COVID-19 shock at the individual firm level. We use the estimated ML model with the best performance in predicting the 2019 export status of firms exporting in 2018 to build a 2020 non-COVID counterfactual outcome for firms exporting in 2019. Then, we compare these counterfactual non-COVID firm-level export probabilities with the predicted probabilities obtained by using the best performing ML model using the characteristics of 2019 exporters to predict their export status in 2020. These estimated probabilities summarize the information on the observed COVID scenario and express it in a metric that is comparable with the estimated counterfactual non-COVID outcomes. Finally, we use recursive partitioning methods to identify subgroups with differential treatment effects. We find that, besides the temporal dimension, the main factors predicting treatment heterogeneity are interactions between firm size and industry.
Asset Growth and Bond Performance: The Collateral Channel (G1, G0)
Firms often use financial leverage to fund their asset growth, particularly under low interest rate environment. This study documents a pervasive inverse relationship between corporate growth in theirs assets and subsequent bond returns among non-investment and low-investment grade bonds. The spread of average annual return between high asset growth and low growth decile portfolios is as large as 5.03% for equal-weight portfolios. We interpret this finding as a dominance of the collateral effect that more tangible assets are accompanied with high asset growth effect to the leverage effect. While this finding supports the conventional risk story, we cannot not dismiss the possibility that investors overreact to corporate asset growth. We find that yield and default probability of high asset growth firms increase more than those of low asset growth firms. This effect is stronger in low-interest rate years.
Asset Pricing In a World of Imperfect Foresight (G1, G4)
We consider a canonical asset pricing model, where agents with quadratic preferences are allowed to re-trade a limited set of securities over multiple periods, after which these securities expire, and agents consume their liquidation values. A key assumption in this model is that agents have perfect foresight: for all future contingencies, they correctly foresee the corresponding equilibrium prices. We show that, under myopia, prices generically are as if agents had perfect foresight. Yet their choices are wrong, because of neglected re-trading opportunities. In an experiment, we find both prices and choices to be consistent with myopia.
Asset Pricing with Option-Implied Consumption Growth (G1, E2)
Prompted by the asset pricing puzzles, particularly the equity premium puzzle (Mehra and Prescott, 1985) and the risk-free rate puzzle (Weil, 1989), the consumption-based asset pricing literature has been in search of more appropriate measures or models of consumption growth such that the resulted stochastic discount factors can be consistent with the asset returns for reasonable preference parameters. This paper shows that once the Epstein and Zin (1989) recursive preference, which is commonly used in the asset pricing literature, is assumed in a representative agent model, the market index option prices largely reveal the market subjective beliefs about the distribution of consumption growth and, thus, impose strong restrictions on the specification of consumption growth risk. The asset pricing with this option-implied consumption growth explains the data well. The model gives the data of option prices and stock returns enough flexibility in picking either a combination of low consumption risk and high risk aversion or a combination of high consumption risk and low risk aversion. The data indeed picks the latter with a risk aversion of 4.72, solving the equity premium puzzle. Moreover, the option-implied equity premium in real time predicts the future market returns better than the historical mean forecast as well as the popular predictors. The equilibrium risk-free rate not only matches the sample moments but also is highly (97%) correlated with the data. The model also implies a market conditional volatility in real time which not only matches the sample volatility on average but also predicts the future volatility much better than the lagged realized volatility or the option-implied risk-neutral volatility. Finally, the model also captures the dynamics of dividend-price ratio ratio and necessarily explains any asset pricing puzzles in the options market, e.g., the implied volatility skew, given that the option prices are taken as inputs.
Bad News, Good News: Coverage and Response Asymmetries (C3, E3)
We construct two measures of media coverage of bad and good unemployment figures based on three major US newspapers. Using nonlinear SVAR techniques, we document four facts. (i) There is no significant negativity bias in media coverage of economic events. The asymmetric responsiveness of newspapers to positive and negative economic shifts is entirely explained by the higher persistence of bad shocks. (ii) Bad news are more informative than good news. (iii) Bad news increase agents’ agreement about economic outcomes and modify their expectations more than good news. (iv) Consumption reacts to bad news, but not to good news.
Bank Capital Requirements, Lending Supply and Economic Activity (E3, G0)
We evaluate the relation between bank capital, lending supply and economic activity using
Italian data over 1993-2015, a period which covers three key post-crisis regulatory and supervisory
measures (the Basel III reform; the 2011 EBA Stress test; the ECB’s Comprehensive Assessment
and launch of the Single Supervisory Mechanism – SSM). We quantify the effects of increased
bank capital requirements adopting a novel procedure which recovers the size of the policy actions
relying on scenario analysis and Bayesian VARs with a rich characterization of the banking sector.
We document that the EBA and SSM measures unpredictably raised Tier 1 ratio by about 2.5
percentage points, leading to an average reduction in credit to firms and households by 5 and 4%,
respectively, and to a decline in real GDP by over 2 and 4%. The Basel III bank capital increase is
instead correctly anticipated in out-of-sample forecasting. These findings are robust to time-varying
model parameters and consistent with narrative sign restrictions techniques.
Bank Risk and Bank Rents: The Franchise Value Hypothesis Reconsidered (G2, G2)
The predictive relationship between banks' Tobin Q and a theory-based measure of bank risk of insolvency is highly non-linear. Using large samples of publicly quoted banks in the US, Europe, and Asia during 1985-2017, we find that higher values of Q predict lower bank risk of insolvency up to values of Q close to 1, but higher values of Q predict higher bank risk of insolvency when Q exceeds 1 and franchise value is priced. The franchise value hypothesis (FVH) postulating a negative relationship between bank rents and risk is thus rejected in our samples. We then construct proxy measures of bank efficiency rents, loan and deposit pricing power rents, and rents due to government guarantees as the potential sources of franchise value, and show that an increase of any if these rents associated with higher franchise values, hence they predict higher bank risk of insolvency. We provide an explanation of the rejection of the FVH using a calibration of two standard financial models of the banking firm, and illustrate a simple model of an "exciting life hypothesis" that can rationalize the positive correlation between pricing power rents and a measure of cost inefficiency found in the data.
Bank Specialization and Zombie Lending: Evidence from Belgium (G2, G3)
Zombie lending has been shown to have a detrimental effect on economic growth and business dynamism. Using a unique dataset drawn from the Corporate Credit Register of the National Bank of Belgium, we analyze a novel determinant of zombie lending and show that higher bank specialization leads to relatively lower lending to zombie borrowers. We also demonstrate that banks act to prevent the negative zombie spillover in sectors, where they are more specialized, by curbing lending to zombie borrowers relatively more when sectors are more populated by zombie companies and in more rapidly growing sectors. Additionally, we document that more specialized banks are less willing to reduce lending to zombie companies in sectors with higher asset specificity. Overall, our results suggest that higher bank specialization may benefit financial stability as it reduces a bank's risky exposures.
Banks, Shadow Banks, and Business Cycles (E3, G2)
Credit spreads on household and business loans move in lockstep and spike in every recession. We propose a theory as to why banks tighten their lending standards following a drop in market sentiment. The key feature is a procyclical shadow banking sector that shifts risk from traditional banks to investors through securitization. We fit the model to euro area data and find that market sentiment shocks are the main driver of business and financial cycles over the past two decades.
Better Safe than Sorry: The Impact of Green Card Delays on the Propensity of Foreign STEM Doctorates to Work at Startups (J6)
In October 2005, doctorates from China and India began to face multiple-year delays in the processing of their EB-2 permanent residency visa applications due to newly-binding country-specific quotas on employment-based green cards. This created an incentive for Chinese and Indian doctorates to seek employment at established firms over startup firms as the former are generally less likely to shut down prior to the resolution of visa delays. Using a difference-in-differences identification strategy, I find that temporary resident STEM doctorates from quota-constrained countries were 7.2 percentage points (42%) less likely to work at a startup in their first decade of employment after the emergence of visa delays, conditional on being employed at any US for-profit firm within ten years of graduation. When allowing for the possibility that Chinese and Indian doctorates respond differently to visa delays, I find that the main result is driven by a 9.2 percentage point (48%) drop in the propensity of Chinese STEM PhDs to work in startups after visa delays were announced, whereas Indian PhDs do not appear to reduce their rate of working in startups in response to visa delays. Additionally, I find that Chinese STEM PhDs are 1.7 percentage points less likely to work at a startup within ten years post-PhD for each additional year of delay.
Biased News and Irrational Investors: Evidence from Biased Beliefs about Uncertainty and Information Acquisition (G1, G4)
Investors who use biased information from news media subsequently tend to make irrational decisions about acquiring firm-specific information compared to rational expectations. This model of information acquisition yields testable predictions that are verified by using a novel dataset of news stories. First, when sentiment in news articles, as a proxy for biased public information, is more optimistic, investors tend to acquire less earnings-relevant information before the earnings announcement and vice versa. Second, the return predictability from firm-specific news sentiment confirms that it contributes to variations in asset information risk due, in a biased belief equilibrium, to the proportion of informed investors deviating from rational expectations. Overall, these findings suggest that biased public information inherent in news sentiment serves to irrationalize investors’ acquisition of firm-specific information through a biased perception of uncertainties in the risky asset payoff.
Birds of a Feather Invest Together - The Effect of National Culture on Financial Decisions (Z1, G5)
The empirical finance literature has until recently ignored the relevance of cultural traits in explaining financial dimensions. However, it seems reasonable to assume that agents are influenced by their national culture while making investment decisions, allocating financial assets, and forming risk preferences. In fact, as highlighted by Zingales (2015) “homo economicus was embedded in a cultural context and this context affected people’s choices in a relevant way”.
This study seeks to understand the impact of national culture on asset allocation and investment decisions based on a cross-country analysis. We base our baseline analysis on OECD countries and rely on Hofstede’s (2001) six-dimension model as our primary source of national cultural traits. We determine the relationship between national cultural traits and financial decisions by conducting a cross-sectional analysis. To establish causality, we employ the empirical methodology recommended by Zingales (2015). Namely, we “isolate the cultural component of beliefs and preferences by instrumenting them with their cultural determinants”. We closely follow the implementation devised by Berger et al. (2020) and apply it to the “OECD Household’s Financial Assets and Liabilities database”. The database is available from 1995 to 2019 and provides internationally comparable asset allocation observations of household financial assets broken down into six main categories.
We find that, in more individualistic societies, households hold less currency and deposits and invest more in equity and mutual funds than in more collectivist societies. In nations characterized by a greater uncertainty avoidance index, households hold more currency and deposits, invest more in money market funds and bonds but less in equity and mutual funds. In countries in which the power is distributed unequally, households hold more currency and deposits and, unexpectedly, invest less in pension funds and life insurances.
Bond Financing Channel of Monetary Policy: Evidence from Bank Lending (E5)
Using the loan level data from a global leading policy bank, we propose the bond financing channel of monetary policy transmission. This paper studies the lending behavior of the policy bank in response to funding costs fluctuation caused by monetary policy shock. The main findings show that tighter monetary policy would increase loan rate and loan spread. In addition, it would also reduce loan volume and increase loan quality. We also explore the heterogeneous effect of monetary policy. The pass-through of monetary policy is stronger in marketization reform period and tight monetary policy period, as well as in areas with lower economic development, weaker fiscal capacity and smaller bank penetration.
Bond Pricing and Business Cycles with Central Bank Asset Purchases (E3, E5)
This paper studies the term premium in a general equilibrium model with a financial constraint and central bank asset purchases. Structural estimates of the term premium match past empirical measures. Term premium dynamics are policy dependent, with Federal Reserve quantitative easing programs reducing the term premium by 1.25% at peak and over 0.9% on average from 2009 to 2019. The model introduces household debt with a loan-in-advance constraint. Other models introduce household debt with relatively impatient agents. The loan-in-advance specification nests the relative impatience setup. Endogenous debt rollover under the loan-in-advance specification dampens the estimated effect of asset purchases on GDP.
Relative to past papers, I use a simple New Keynesian model to show how the term premium acts as an endogenous cost channel in the Phillips curve. Expected changes to the term premium act as a credit wedge in the IS curve. Current inflation is a function of the forward-looking paths of the output gap and term premium.
Structural estimates of the term premium are consistent with measures from Kim and Wright (2005) or Adrian, Crump, and Moench (2013). The loan-in-advance setup from this paper generalizes the "Four-equation New Keynesian Model" from Sims, Wu, and Zhang (2020), allowing the term premium to act as an endogenous source of propagation for non-financial shocks. Alternatively, under the four-equation setup, the term premium is only the origination point by which financial shocks propagate to the economy.
Finally, the loan-in-advance specification alters traditional forward guidance concerns in the New Keynesian literature. The presence of a financial wedge dampens the effect of forward guidance in the short- to medium-run, but does not alleviate long-run forward guidance perplexities common to the New Keynesian literature.
Born in the COVID-19 Pandemic: The Effects of Elective Medical Procedure Delay on Infant and Maternal Health (I1, J1)
This paper studies the impacts of the elective medical procedure delay order induced by the COVID-19 pandemic on infant and maternal health outcomes. 32 states in the United States issued executive orders to delay elective or non-urgent procedures in order to conserve personal protective equipment and free up staff and facilities for COVID-19 patients. By utilizing the Regression Discontinuity design on a nationwide claims dataset, this paper finds that infants born after the elective medical procedure delay order are less likely to do general examination, get vaccinated, and visit the ER within 90 days from birth. Meanwhile, infants born after the order are more likely to be diagnosed with conditions originated in the perinatal period. As infants grow up to 180 days, symptoms related to development, such as osteochondrosis and mental disorder, are more likely to be detected among those born in the post elective-procedure-delay period. Additional analysis suggests that pregnant women and new moms are more likely to develop anemias, nutritional diseases, and mental issues after elective procedures are temporarily banned.
Bounding Omitted Variable Bias Using Auxiliary Data (C1, J0)
This paper provides an estimator to bound omitted variable bias using auxiliary data including proxies for the omitted variable. The R package (’bndovb’) implementing the estimator is provided. This estimator can complement popular methods, such as Altonji et al. (2005) and Oster (2019), which bound the omitted variable bias using the identifying assumption that the degree of selection in treatment due to an unobservable is smaller than selection in treatment due to observables. Compared to Altonji et al. (2005) and Oster (2019), this method does not require making such an assumption and therefore may provide a robust bound when a researcher does not have a good prior about the degree of omitted variable bias. In addition, this method does not suffer from a multiple solution issue. An empirical example of estimating the Mincerian wage regression using auxiliary data including proxies for cognitive
and noncognitive ability is provided.
Brahmin Left versus Merchant Right: Changing Political Cleavages in 21 Western Democracies, 1948-2020 (P1, P5)
This paper provides new evidence on the long-run evolution of political cleavages in 21 Western democracies by exploiting a new database on the vote by socioeconomic characteristic covering over 300 elections held between 1948 and 2020. In the 1950s-1960s, the vote for democratic, labor, social democratic, socialist, and affiliated parties was associated with lower-educated and low-income voters. It has gradually become associated with higher-educated voters, giving rise to “multi-elite party systems” in the 2000s-2010s: high-education elites now vote for the “left”, while high-income elites continue to vote for the “right”. This transition has been accelerated by the rise of green and anti-immigration movements, whose key distinctive feature is to concentrate the votes of the higher-educated and lower-educated electorate, respectively. Combining our database with historical data on political parties’ programs, we provide evidence that the reversal of the educational cleavage is strongly linked to the emergence of a new “libertarian-authoritarian” axis of political conflict. We also discuss the evolution of other political cleavages related to age, geography, religion, gender, and the integration of new ethnoreligious minorities.
Breaking the Sovereign-Bank Nexus (G2, E4)
This paper develops a quantitative dynamic general equilibrium model that features endogenous bank failure and sovereign default risk. It studies the feedback loop between sovereign and banking crises, and evaluates the effectiveness of bank capital regulation in addressing it. In the model, bank failure contributes to an increase of sovereign default risk through the government bailout of bank creditors. Meanwhile, holding high-yield risky sovereign bonds may be attractive to banks protected by limited liability. By increasing banks' failure risk and their funding costs, sovereign exposures hurt bank lending and contribute to further contractions in aggregate economic activity. Capital requirements shape banks' incentives to invest in sovereign debt. More stringent capital regulation makes banks safer, weakening the sovereign-bank nexus. This comes at the cost of constraining the overall supply of credit.
Can a ‘Pet’ Take a Bite Out of the Savings Shortfall? (D9, G4)
This paper explores the potential for gamification to develop and nurture a saving habit among Americans. Students are introduced to ‘pet’ savings accounts that require continual care. The owners of these ‘pets’ are reminded periodically to feed them at the touch of a button, with positive reinforcement built in. The feeding of the digital savings pets is linked to an actual bank account, automatically transferring funds from a checking account to a dedicated savings account. The amount of money needed to feed the pet is small — several dollars a day — to help reframe the problem and reduce loss aversion. Converting the process of saving for retirement into one of caring for a pet today changes the reference point for individuals. Savers check in on their pets by getting indicators of their portfolio performance, translated into easily understood communication. The savings ‘pets’ will be launched among students at a university, to provide a proof of concept. Results will analyze the success of this approach in increasing saving accumulations among students, and evaluate its promise for tackling the savings shortfall among Americans.
Can Air Pollution Save Lives? The Impacts of Air Quality on Risky Behavior (I1, Q5)
In this paper, we find a rare “benefit” of air pollution: reducing the number of road accidents. Although studies have documented that elevated air pollution increases the number of traffic accidents and argued that impaired cognition is the main channel of such impact. Still, medical studies have found evidence showing that exposure to air pollution can increase the levels of stress hormone cortisol and make individuals more risk averse. We therefore hypothesize that air pollution can affect road safety through both cognitive impairment and higher level of risk aversion, where the former and latter would increase and decrease the number of accidents, respectively. Using the administrative individual-level traffic accident data from Taiwan between 2009 and 2015, this paper investigates the effect of air pollution on the number of traffic accidents with casualties. To treat the endogeneity between air quality and the number of traffic accidents, we use wind direction as an instrument variable to introduce exogenous variation in the level of air pollution. In addition, we apply a land-use regression model accounting for population, elevation, and emission data to interpolate 3km*3km level air quality data. Our results show that a 1 ug/m3 increase in PM2.5 concentration leads to a 0.3 – 0.5% reduction in the number of accidents associated with rule violations and a 0.3 – 0.6% reduction in speeding violations. However, the impacts of elevated air pollution on the number of accidents associated with carelessness and errors as well as violations that are “less risky,” such as illegal parking, are imprecisely estimated. These findings suggest that air pollution can save lives through less risky behavior. Moreover, those negative effects on violations are stronger among non-enclosed vehicle drivers, such as motorcyclists, so we argue that direct exposure is the main mechanism explaining how air pollution affect road use behavior.
Can Mentoring Alleviate Family Disadvantage in Adolescence? A Field Experiment to Improve Labor-Market Prospects (I2, J2)
We study a mentoring program that aims to improve the labor-market prospects of schoolattending adolescents from disadvantaged families by offering them a university-student mentor. Our RCT investigates program effectiveness on three outcome dimensions that are highly predictive of adolescents’ later labor-market success: math grades, patience/social skills, and labor-market orientation. For low-SES adolescents, the one-to-one mentoring increases a combined index of the outcomes by half a standard deviation after one year, with significant increases in each dimension. Part of the treatment effect is mediated by establishing mentors as attachment figures who provide guidance for the future. The mentoring is not effective for higherSES adolescents. The results show that substituting lacking family support by other adults can help disadvantaged children at adolescent age.
Can Whistleblowers Root Out Public Expenditure Fraud? Evidence from Medicare (I1, H4)
This paper analyzes private anti-fraud enforcement under the False Claims Act, which compensates whistleblowers for litigating against firms that overbill the government. I analyze several case studies of large, successfully prosecuted whistleblower lawsuits from the Medicare program, pairing new data on whistleblower lawsuits with large samples of Medicare Fee-for-Service claims from 1999 – 2016. I estimate that deterrence from $1.9 billion in whistleblower settlements generated future cost savings of nearly $19 billion, while imposing relatively few costs on the federal government. In a case study of fraudulent spine surgery, whistleblower-induced changes to care modestly improved patient health. These results suggest private enforcement by whistleblowers is a cost-effective way to combat public expenditure fraud.
Capital Allocation, the Leverage Ratio Requirement and Banks' Risk-Taking (G2, G0)
This paper examines how the level at which regulatory constraints are applied affects banks' asset risk. We develop a theoretical model and calibrate it to UK banks. Our main finding is that the impact differs depending on which regulatory constraints are binding at the group consolidated level. As long as at the consolidated level, only the leverage ratio constraint is binding, the allocation of constraints down does not imply any negative impact on banks' resilience. However, it could bring about an increase in banks' asset risk in the case where only the risk-weighted constraint binds at the group level. We also find that the impact on banks' asset risk of the application level differ across banks' business models.
Causal Impacts of Teaching Modality on U.S. COVID-19 Spread in Fall 2020 Semester (I1, I2)
In response to the COVID-19 pandemic and its containment measures U.S has implemented partial or full business closures to mitigate the spread. Many U.S. colleges temporarily closed or switched to online in spring 2020, and over six out of ten colleges reopened in fall 2020 with an in-person or a combination of in-person and online teaching plans. College students mainly fall in the age cohort of 18 to 29 years, which has a lower death rate (0.4%) from COVID-19, but a greater chance of socialization than the other age-cohorts.
Given the substantial risk of spread from college campuses to the community, a policy question is whether colleges should hold in-person classes or switch to online or hybrid mode. We combine web-scraped college-level data, NCES, College Crisis Initiative, U.S. census, New York Times, and HealthData.gov to predict the chance of a college adopting one of the three modes of teaching: in-person, online, and hybrid; and then estimate the effects on county-level COVID-19 cases and deaths.
In a quasi-experimental approach, we use a logistic model and a gradient boosting algorithm to estimate the propensity scores for the three groups to adjust the pre-treatment imbalances of college and county-level covariates. We find greater enrollment, fewer republican supporters, and prevailing individual mask ordinance are common predictors of adopting online or hybrid teaching modalities over in-person. Treatment effects provide evidence that college re-openings increase daily new cases and deaths, especially counties where colleges reopened with an in-person mode experienced a statistically significant rise of 2.5 new cases and 0.049 new deaths compared to the pre-semester level. Daily new COVID-19 deaths increase about 0.04 per 100,000 population with campus reopening. Finally, a percent increase in the county population who stayed home decreases new COVID-19 cases and new deaths by about 0.7 and 0.035 per 100,000 population.
Central Bank Digital Currency and Balance Sheet Policy (E5, G2)
This paper studies a stylized economy in which the central bank can hold either treasuries or risky securities against central bank digital currency (CBDC) deposits. The key mechanism driving the results is the reduction in bank deposits that follows the introduction of a CBDC and its impact on the banking sector. With CBDC funds invested in treasuries, the central bank channels funds back to the banking sector via open market operations and the introduction of a CBDC is neutral, consistently with the equivalence theorem of Brunnermeier and Niepelt (2019). However, it is not neutral when accounting for liquidity requirements, quantitative easing or for CBDC deposits held against risky securities. We reach three main conclusions. First, current monetary policy regimes do matter for CBDC equilibrium effects. Second, a CBDC might render current quantitative easing programs a quasi-permanent policy, as commercial banks might use their excess reserves to allow depositors to switch from bank to CBDC deposits. Third, there is a trade-off between bank lending to the economy and taxes, as holding risky assets against CBDC deposits leads to lower expected taxes and lower bank lending.
Central Bank Digital Currency in Brazil (E5, G2)
Digital currencies---Central Bank backed digital payment instruments---pose several tradeoffs including the removal of anonymity that characterizes the use of cash. We study the implications of this trade-off in terms of credit, output, and welfare in the context of an emerging market, Brazil, where concerns are particularly pronounced. We consider a model of means of payment choices and households with different preferences over anonymous cash, non-anonymous deposits, and a digital currency overseen by a central authority that can choose the anonymity level. The financial sector is monopolistically competitive breaking the link between borrowing and lending rates. Calibrating the model to Brazil, we show that a sufficiently attractive digital currency reduces the holdings of both cash and bank deposits. Since the use of cash is costly, the use of the digital currency may increase welfare. However, if banks are liquidity constrained, the digital currency may result in lower credit and output leading to a reduction in welfare. We show that the digital currency interest remuneration can be set optimally to balance this trade-off dominating other policy choices.
Centralized Admission Systems and School Segregation: Evidence from a National Reform (I2)
This paper investigates whether the adoption of a centralized school admission system can alter within-school socio-economic diversity. We assess the importance of two factors: residential segregation and outside options. In theory, both have the potential to increase school segregation under centralized systems. We provide evidence confirming this premise. We take advantage of the largest school-admission reform implemented to date: Chile’s SAS, which in 2016 replaced the country’s decentralized system with a Deferred Acceptance algorithm. We exploit its sequential introduction across regions to quantify its heterogeneous impact on segregation. The empirical analysis is carried out using administrative data and a Difference-in-Difference strategy. SAS increased within-school segregation in areas with high levels of pre-existing residential segregation. School districts with the higher provision of private education experienced an uptick in school segregation as well. The migration of high-SES students to private schools emerges as a driver.
CEO International Background and Cross-Border M&As (G3, G4)
We investigate if having a CEO with an international background affects U.S. firms’ cross-border merger and acquisition (M&A) activities. By defining international background as having non-U.S. nationality, overseas education, foreign work experience, or combinations of these, we find that when a CEO has these characteristics, the firm is more likely to acquire international targets, and these deals are more value-enhancing. Moreover, our results indicate that having multiple international characteristics increases the likelihood and announcement returns of cross-border deals. The observed gains are related to the mitigating effect on due diligence, acquisition premium, and the financing of these deals mostly with equity. Additionally, CEOs with an international background realize most of the gains from conducting cross-border deals when the deals are announced within five years of their tenure as CEO of the firm. This suggests that U.S. firms are aware of the advantages that an international background provides. We address potential endogeneity concerns and our results are robust to alternate definitions of international background, a placebo test, and within-firm analysis.
Child Health and Parental Responses to an Unconditional Cash Transfer at Birth (I1, I3)
Baby Bonuses have been criticized for potentially harming unborn children because of strategic birth-shifting. Surprisingly little is known about the health consequences of unconditional cash transfers tied to the birth of a child. We estimate the causal impact of the unanticipated introduction of the Australian Baby Bonus (ABB) on child health from birth until age 5, using hospital and emergency room records linked to birth certificates and perinatal data from South Australia. We find that treated babies had fewer preventable, acute, and urgent hospital presentations, medical care that is accessible without copayments, in the first two years of life. In the second year of life, the ABB increased demand for elective care, which requires medical referrals and entails co-payments. Our findings suggest that the ABB improved child health through increased parental health investments, especially among disadvantaged families. Back-of-the-envelope calculations indicate that up to 34% of the ABB payouts were recouped within the first year of life of the beneficiary child due to improved health.
China's Currency Intervention on Economic Development (E5, O1)
During the past three decades, the spectacular economic growth pattern of China has been under increasing scrutiny in the literature. However, existing treatises have concentrated on either analyzing the causes/effects of the RMB exchange rate misalignment or theorizing the channels of investment/speculation resources. The contribution of the currency intervention regime on China's economic takeoff, in comparison, still remains one of the most contentious topics. To shed light on this issue, the present study develops a demand-driven dynamic stochastic model which particularly addresses two cores features of the Chinese-style currency intervention: fast commodity markets liberalization and draconian financial controls. Based on the model, I further investigate the implications of macroeconomic controls on economic short-run and long-run economic developments. The conditions of "economic catch-up" and "middle income trap" have also been explored in the paper.
China’s Rebalancing and Gender Inequality (J1, E2)
This paper examines gender inequality in the context of structural change and rebalancing
in China. In contrast with the predictions of the standard structural transformation literature,
which suggest that rising service sector share is associated with narrowing gender gaps, we
document that women’s relative wages and labor force participation in China have declined
significantly during the last two decades. We provide a set of explanations for this finding.
First, using household-level data, we show that women’s labor supply elasticity to spouse’s
wages has increased dramatically between 1995 and 2013. This evidence is consistent with
a U-shaped relationship between economic development and female labor force
participation found in the literature. Second, using a theoretical model of structural
transformation, we show that labor market wedges have increased over time, negatively
impacting both women’s relative hours and earnings, which could in part be due to
worsening employer discrimination and lack of affordable childcare options. Model
counterfactuals suggest that removing discriminatory practices against women in the labor
market and raising service sector productivity can boost both gender equality and economic
growth in China.
Class Rank and Sibling Spillover Effects (I2, D1)
Siblings are perhaps the most important childhood peers, yet we know little about sibling spillover effects on school achievement and their potential mechanisms. I estimate the effect of children’s rank in primary school on their younger sibling’s schooling outcomes using administrative records from the Netherlands. In this setting, variation in sibling rank is credibly exogenous and isolates sibling spillovers driven by behavioral and psychological mechanisms, as opposed to direct transmission of human capital. A 1SD increase in child rank in test scores increases their younger sibling’s test scores by 4.3 percent of a standard deviation, showing that behavioral mechanisms in sibling spillovers are empirically relevant. However, child rank also increases the chance that their sibling is recommended for the academic school track by 5 percent, even after accounting for younger sibling’s test scores. This recommendation is given exclusively by teachers, suggesting that teachers track children based on arguably meaningless information on their siblings. I argue that this is a form of teacher bias in expectation formation and show that it only occurs for non-migrant children. This points towards cultural proximity as an important factor in the formation of biased expectations, widening achievement gaps between migrant and non-migrant children. Overall, my findings show that school inputs can be important drivers of within-family human capital spillovers.
College Education and Income Contingent Loans in Equilibrium: Theory and Quantitative Evaluation (E6, I2)
We investigate the welfare implications of income-contingent loans (ICLs) used for financing college education in presence of the dropout risk that depends on unobservable effort. Using a simple model, we show that the laissez-faire enrollment is inefficiently low due to missing insurance against dropping out. However, providing this insurance generates a moral hazard cost of lowering effort. We show that ICLs can implement the second best allocation. Then, we construct a heterogeneous agent OLG life-cycle model, calibrate it to the US and show that ICLs significantly increase welfare and that their non-linear structure is essential to delivering high welfare gains.
Collusion and Land Market Auctions (H7, K4)
This paper investigates the presence of collusion in auctions. Using a unique institutional feature of China's land market, we compare two formats of auctions and find that the two-stage auction generates a much lower price than the English auction. We exploit an exogenous variation created by an anti-corruption campaign to show that "corrupted" officials chose the two-stage auction less often during the inspection. Thus corruption can explain the price difference. We further document the existence of another type of collusion by showing that the security deposit rate has a reverse-U shape impact on the transaction price. Bidder collusion can explain the low price in the lower end of the security deposit rate axis, and corruption can rationalize the low price on the other end. We also find evidence consistent with the presence of a bid rotation scheme by showing a positive correlation between the closeness of firms and the timing of when they won the land parcels.
Combining Financial Incentives with Nudges to Increase Preschool Parental Engagement (I2, D1)
Disadvantaged children arrive at kindergarten behind their more advantaged peers in indicators of school readiness. Previous research shows that family engagement is a crucial aspect to improving child outcomes. As such, Head Start and other publicly supported preschools are required to spend substantial funds promoting family engagement. In order to increase parental attendance at school-sponsored family-engagement events, we designed an intervention using a combination of financial incentives and tools from behavioral economics. Our intervention was a 17-week Randomized Control Trial (RCT) with 319 parents across 6 preschools in Chicago. The treatment group was given a $25 per event incentive to attend 8 events sponsored by their preschool, as well as weekly text message reminders of the events. The financial incentive was framed using loss aversion, where the parents were given $200 in a virtual account, and lost $25 for missing each event. The overall likelihood of attending an event was 12.5% in the control group and 17.5% in the treatment group, representing a 40% increase. There was little heterogeneity by event time and type. Attendance was not substantially higher for events later in the day, where work conflicts are less likely to arise. While this increase is substantial in relative terms, the fact that the absolute level of attendance was still less than 20% despite high incentives in the treatment group might imply that parents put little value on such school-sponsored events.
Commitment and Conflict in Multilateral Bargaining (C7, D7)
We extend the Baron and Ferejohn (1989) model of multilateral bargaining, allowing players to take an aggressive bargaining posture by attempting to commit to a bargaining position prior to negotiating. Any such attempted commitment fails with an exogeneously given probability. If successful, commitment binds a player to reject any proposal which allocates to her a share below a self-imposed threshold. We characterize the symmetric stationary subgame perfect equilibria. Under unanimity rule, there are potentially many equilibria which can be ordered from the least to most inefficient, according to how many commitment attempts must fail in order for an agreement to arise. The most inefficient equilibrium exists independently of the number of players. More efficient commitment profiles cannot be sustained in equilibrium if the number of players is sufficiently large. Expected delay increases (setwise) with the number of players. Under any (super)majority rule, the unique equilibrium is efficient. The results suggest that the unanimity rule is particularly damaging if the number of legislators is large and the time lags between consecutive sessions are long.
Common Knowledge and Collective Action on Directed Communication Networks: Models and Experimental Findings (C7, D8)
Social media platforms are critical tools in organizing collective action, such as the Occupy Wall Street protests and the Arab Spring, and more recent Black Lives Matter protests, and the 2021 Storming of the US Capitol. In collective action problems, individuals are willing to participate if there is a sufficient number of other participants, and this requires that individuals know about each other's willingness to participate (threshold) and that this information is common knowledge (CK). Social networks can help the spread of information and facilitate CK and coordination, making collective action possible.
In this paper, we propose game-theoretic models of collective action on Twitter and Facebook communication networks. We model collective action as a coordination game in which individuals post their thresholds. Our goal is to understand how actionable information spreads locally on social media platforms and to theoretically characterize the necessary and sufficient conditions for CK and the minimal substructures necessary for CK to occur. Previous models of CK and collective action (Chwe 1999, 2000) assume that the network structure is commonly known by everyone. We characterize the communication patterns that facilitate CK and coordination where the network structure is (i) known by everyone, and (ii) locally known. A model of collective action on Facebook (Korkmaz et al. 2014) assumes that communication is undirected. Our model relaxes the assumption of bidirectional communication and assumes that communication can occur in one direction. Prior works do not discuss CK on Twitter-type communication networks, and this work proposes game-theoretic models of CK on Twitter networks, focused on Twitter “retweets”.
To test the predictions of the models, we conduct human subject experiments to identify the effects of both network structure and communication on CK. We find that in networks that satisfy our theoretic predictions, a higher number of participants choose to participate.
Competition and Corruption: Highway Corruption in West Africa (K4, O1)
Petty corruption in the developing world impedes citizens from receiving public services
and operating their businesses. In this paper, we show the importance of market structure
in determining a corruption equilibrium. We do this in the context of highway merchandise
transportation in West Africa, where checkpoint ocials frequently stop truck drivers for petty
bribes. We exploit a road system with two alternative corridors to develop a model which
predicts that checkpoints in the two competing corridors follow a Bertrand game as they set
price equal to the marginal cost. Moreover, when costs to pass through one corridor increase
due to road construction, checkpoints in the other corridor raise prices and keep drivers waiting
for longer. We estimate a dierence-in-dierences model to conrm that road construction did
increase both bribes and enforced delays for stops in the unaected corridor. This work demon-
strates the importance of competition among corrupted ocials to facilitate public services for
drivers and suggests that the eectiveness of a local intervention can be oset by reallocating
customers towards ocials who are not aected by it.
Compression as an Alternative to Central Clearing (G2, G0)
In the wake of the 2008 financial crisis, regulators have turned to central clearing in an effort to reduce systemic risk. As part of the 2009 Pittsburgh Summit, the G20 leaders agreed that standardized over-the-counter (OTC) derivative contracts “should be traded on exchanges or electronic trading platforms, where appropriate, and cleared through central counterparties” (G20 Leaders 2009).
However, central clearing is not without its downsides. Having all trades routed through a single central counterparty (CCP) mechanically introduces a systemically important participant which itself may require more regulatory oversight. Furthermore, the failure of the CCP could have disastrous consequences for the entire market. The loss sharing mechanisms designed to reduce risk also introduce moral hazard problems for market participants. Margin requirements may be excessive, especially in the presence of multiple CCPs across markets (Duffie 2011).
In this paper, we first develop a tractable model for analyzing the risk structure of different bilateral networks with the same net exposures, and study the implications for systemic risk. We conduct counterfactual analyses based on the current market structure of foreign exchange swaps. We find that a slight modification of trade compression can mimic many of the desirable features of central clearing, and in some cases address the limitations of central clearing as well. Specifically, we show that including intermediary financial agents taking both long and short positions with different counterparties can result in many of the risk-sharing benefits of a centrally cleared market.
Our findings suggest trade compression is a viable alternative to CCPs for markets where central clearing is impossible, such as cross-border markets or markets where similar but not identical products are traded. Furthermore, as the post-compression market is still a bilateral market, and compression ultimately requires consent from all parties involved, compression is not as likely to introduce moral hazard problems.
Consensus Among Economists in 2020 (A1)
Based on survey research of American Economic Association (AEA) members, this paper updates previous work that has been conducted roughly every decade since the 1970's and that has been published in the American Economic Review, Papers and Proceedings, and the Journal of Economic Education. Economists are asked about their level of agreement with 46 economic propositions. Using a number of measures to classify the distribution of responses, the study is able to discern areas where economists show consensus, but also areas where economists' consensus has shifted over time. Previous research has shown that economists show the strongest consensus for propositions that deal with the global economy and free trade. Historically, the least consensus is found for macroeconomic propositions. The 2020 survey keeps almost half of the propositions identical to the initial survey conducted in the late 1970's, 1990, 2000 and 2011. Therefore, meaningful changes in the opinion of economists regarding international, macro-, micro-, and distributional propositions can be observed over time. The current survey includes new propositions that focus on issues of diversity, equity, and inclusion; climate change; and the pandemic. The current survey also reaches a larger sample of the AEA membership as this is the first time the survey is distributed online to members who have agreed to participate in survey research.
Consequences of Cambodian Refugees (O1, Z0)
Using the complete count 1998 Census microdata, this paper examines consequences of Cambodian refugees who emerged around the collapse of the Pol Pot regime (1975-1979). Our focus is the two major groups of returnees from the neighboring countries, Thailand and Vietnam, who had a distinct repatriation process: The former stayed in refugee camps with humanitarian assistance prior to repatriation and the latter did not. This distinct circumstances reflected the Cold War situations: During the 1980s, when it intensified, the two groups were under the control of different great powers, East vs. West. Our analysis reveals that the returnees from Thailand (Vietnam) experienced better (worse) educational outcomes through forced displacement, while the two
groups both generally experienced worse labor market outcomes. Such adverse displacement impacts are severer for later returnees. We provide suggestive evidence that congested labor markets due to limited access to land mainly drive the results.
Contagion, Migration and Misallocation in a Pandemic (I1, E6)
We present a multi-city migration model to study the endogenous choice of migration and to evaluate various policy alternatives during a pandemic. We combine a traditional epidemic dynamic model with residents' endogenous migration decisions between cities as well as constraint of hospital resources of each city, and provide analytical solutions under two situations: laissez-faire equilibrium and social optimal policy. By studying the relationship between migration rates of different types of residents and the abundance of hospital resources in different cities, we find that migration rates under laissez-faire equilibrium and under social optimal policies diverge, especially for infected people. We also provide cross-country empirical facts that countries with higher COVID-19 mortality disparity across sub-regions (calculated as the normalized standard deviation of mortality rates of a country's all provinces or states) tend to have lower nationwide mortality rates. This new fact is inconsistent with the prediction of the misallocation literature, nevertheless is aligned with our theoretical model, which can be seen as an extension of the traditional misallocation framework with cross-region contagion and externality. In addition, through numerical analyses we observe that total social welfare can be increased by around 25% when collective actions are taken in a two-month span. This research will shed light on policy design under pandemic threat and provide a new insight to enrich the misallocation literature.
Corporate Asset Pricing (G1, G3)
Koijen and Yogo (2019, 2020) show that unexplained demand movements are responsible for the majority of asset price changes. This paper helps explain this unexplained demand. Using corporate demand, I provide the first demand-based convenience yield explanation. I show that when managers are exposed to moral hazard, corporate demand will be determined by their idiosyncratic risk. I see in the cross-section that idiosyncratic volatility leads to higher financial asset holdings, higher savings, and has decreased corporate investment by 5% on average annually. In the time series corporate demand predicts 29% of the convenience yield variation. This effect dominates other investor classes such as financial intermediaries. I isolate my demand-based effect from confounders by using exogenous cross-sectional variation from corporate size and industry exposures.
Corporate Leverage: Business as Usual and Speculative Cycles (C5, G3)
This paper explores if the recent trends in the U.S. corporate debt should raise a concern about possible speculative borrowing cycle. The conceptual foundation is offered by the analysis of the Minskyan cycles and Kalecki’s principle of increasing risk. Empirically, the paper develops a model of corporate capital structure, based on Stein (2011, 2012) for four time periods: before, during, and after the 2008 global financial crisis, and before the Covid-19 crisis. The sample consists of 86 corporations from six industries: technology, financial, pharmaceutical, auto, airline and energy, for the time span from 2000 to 2018. Calculated for each firm, the aggregated model allows to infer a default risk on an industry basis by measuring overleveraging, defined as the difference between actual and optimal debt. Here, optimal debt is the debt capacity above which borrowing becomes unstable in the medium term. As a result, the excess debt represents the degree of overleveraging at a firm and industry basis. The results suggest that excess debt, rather than the mere holding of debt, was the reason behind the severe financial meltdown in 2007–2009 and the excess debt moves up in deep recessions. With broad variations in vulnerability by industry, and as the pandemic exacerbates aggregate demand trajectory, the results of this paper also conform with the general Minsky-Kaleckian analytical framework suggesting a possibility of rising speculative bubble in the economy over the medium term.
Corporate Social Responsibility and Financial Reporting Quality (G3, M0)
This paper studies the relationship between corporate social responsibility (CSR) and financial reporting quality. Our findings based on a data sample over the period from 1991 to 2018 confirm a positive relationship between CSR and reporting quality and demonstrate that firms with higher levels of corporate social responsibility investment are associated with higher accuracy of financial forecasts, fewer earnings surprises, and greater coverage by financial analysts. Empirical results hold after we account for potential endogeneity in this relationship. Additional analyses reveal that the positive relationship between CSR and financial reporting quality is stronger for firms that face low agency concerns, and firms that have a higher level of customer awareness and more long-term institutional ownership. We also find that when firms are financially unconstrained, the link between CSR investments financial reporting quality becomes stronger. Finally, we document that firms with higher CSR and better financial reporting are more likely to experience lower risk and higher levels of information disclosure. The evidence supports the stakeholder value maximization view of CSR and identifies the areas
of positive impact of CSR investment at a firm.
Correlation of Idiosyncratic Asset Return and Wage Risk of U.S. Households (E2, D3)
This paper documents the degree of idiosyncratic asset returns risk, serial correlation, and correlation with wage risk for US households. Novel panel-data measures for returns on household assets are proposed, which, notably, are the first to include net investment in the measures of capital gains. Sizeable transitory idiosyncratic return risk is documented for total household assets and for returns within each asset class. This idiosyncratic returns risk exists concurrently with permanent heterogeneity in household-specific returns and exhibits negative serial correlation. On average, households do not exhibit correlations between the idiosyncratic risk to wages and total asset returns. However, this masks the correlated wage and return risk to private business assets and secondary housing assets and capital gains to primary housing. Secondary housing assets are negatively correlated with permanent shocks to wages and hence play a key role in insuring against permanent wage risk. These estimates inform the structure of covariance risk in quantitative models that include uninsured idiosyncratic returns and wage risk for the examination of portfolio choice and inequality.
Cost of Research and Education Activities in Us Colleges - Complementarity, Scalability, and Heterogeneous Efficiency (I2)
Colleges and universities in the United States are remarkably diverse in their efficiency, both in terms of research output and educational achievement, but existing literature has so far failed to adequately describe this heterogeneity due to a lack of appropriate data and methodological limitations. In this paper, we address this problem by using new data and a recently proposed method to infer cost functions for universities.
We utilize three main data sources, two of which became only recently available. One is Microsoft Academic Graph, which contains detailed information on academic publications, including author and institutional affiliation. Another is the mobility report card collected by Chetty et al (2017), which reports income data for graduates and their parents by universities. Third, we use IPEDS data, which provides budgetary information. By combining these datasets, we can construct comprehensive measures of university outputs with research and education as outcomes.
Our model combines neural network techniques with recent advances in ``interpretable machine learning'' algorithms. Specifically, we first train a neural network that predicts the institutional cost from all the available variables including the outputs, then approximate the derivative of the predicted total cost along each dimension of outputs utilizing the recently developed SmoothGrad technique (Smilkov et al, 2017). This algorithm allows us to reliably estimate heterogeneous marginal costs with relatively short time series without imposing structural assumptions on the functional form. Furthermore, we apply SmoothGrad again to the estimated marginal costs, yielding the second derivative estimates that allow us to analyze the complementarity between research and education production. Our analyses reveal the trade-off between research and education efficiency, the economy of scale in research outputs, and efficiency differences across universities. We further adopt SHAP method to explore what explains the heterogeneity in cost efficiency (Lundberg and Lee, 2017).
Could Intra-Firm Informational Misalignment Explain Price-Setting Patterns? (E3, D8)
We propose a simple model for firm's pricing decision, based on the interplay between communication within firm and the provision of incentives. Such mechanism, endogenously, generates discrete prices and explain price stickiness even though there is no cost on adjusting prices or acquiring information. We embed this firm structure in a multi-sector general equilibrium model and derive a new Phillips curve where the misalignment of incentives and the number of divisions of a given firm drive the slope of the Phillips curve, exposing a new channel to monetary policy and illustrating its impact. Empirically, we take the model into a new retail daily database across countries and supermarkets to stress the relationship between within firm incentives and price-setting discussion, providing an empirical estimation of the theoretical mechanism. Our model matches the moments of microeconomic evidence on price-setting, as well as the macro-impact of real effect of monetary policy.
Covered Bonds and Bank Portfolio Rebalancing (G2, G3)
We use administrative and supervisory data at the bank-, loan- and firm-level to investigate the impact of covered bond issuances on bank lending and real economic outcomes. We show that the introduction of covered bonds leads to a rebalancing of bank portfolios from mortgages to corporate loans. We provide a theoretical framework for analyzing the impact of covered bonds on bank portfolio allocation, and highlight two opposing forces: On the one hand, covered bonds encourage banks to issue more mortgage loans due to lower funding costs. On the other hand, covered bonds enhance the liquidity of existing mortgages which allow banks to substitute mortgages with riskier corporate lending for higher yields. If initial bank liquidity is sufficiently low, the latter mechanism dominates. We provide empirical support for this by showing that the observed portfolio reallocation is driven by low-liquid banks. The increase in corporate credit leads to more favorable outcomes at the firm-level.
COVID-19 Disease Threat Perceptions and Migrants’ Willingness to Return to Work in India (J6, I1)
In this paper, we explore the link between the likelihood of re-migration to cities and the perceived threat of contracting COVID-19 among a unique sample of reverse migrant workers in India. We find that individuals who perceive a significant chance of contracting COVID-19 have a significantly lower stated likelihood to return to their urban work centres. We observe heterogeneity with respect to the duration of migration as long-term migrants perceive a lower disease threat than short-term migrants. We also find that COVID-19 threat perception increases with a rising number of unique sources of information and decreases with a greater recall of preventive measures. With respect to behavioural factors, we find that loss-averse individuals display lower disease threat perception, whereas more impatient individuals have higher COVID-19 risk perception. Thus, a key policy takeaway from our results is that along with standard economic incentives, behavioural factors and access to information regarding COVID-19 crucially determine migrants' potential return to urban workplaces.
COVID-19 Epidemic and Generational Welfare (E6)
We study the effects of COVID-19, and the ensuing lockdown and fiscal policies, on the welfare of different age-groups within a life-cycle macroeconomic scheme, adapted from Gertler (1999), where the pandemic is represented as a shock to the mortality rate. We obtain two main results. First, we can show that lockdown policies have a negative impact on the dynamics of economic welfare of younger agents relative to that of older agents, thus providing analytical support to the idea that the management of the COVID-19 pandemic through lockdown policies has hit mainly the young generations. Second, we show that expansionary fiscal policies aimed at supporting income after the lockdown affect the relative welfare index of age-groups mainly through the repayment scheme of the consequent public debt; the more the repayment scheme entails a postponement of the debt repayment, the more older agents are favored (in relative terms).
COVID-19, Housing Prices and Macroprudential Policies (E4, E5)
The COVID-19 crisis has been unprecedented in many angles. Unlike the 2008 global financial crisis, this crisis came from real factors caused by lockdown measures. Theoretically, in times of recession, when businesses close, unemployment rises and uncertainty prevails, house prices should be falling. However, unexpectedly, the lockdown effects and the easing of monetary policy have unevenly impacted consumers and sectors. This has provoked a redistribution towards housing, which has translated into an increase in house prices. In this context, the composition of policies to shape the recovery should take into account a coherent mix between monetary and macroprudential policies. This paper builds a multi-sector DSGE model, which features a housing market. Consumers are divided into savers and borrowers. Borrowers need collateral to obtain loans. Macroprudential policies are represented through changes in loan-to-value ratios. First, we use the model to assess how corona-shocks affect both the macroeconomy and housing markets. Then, we study the optimal monetary-macroprudential combination of policies to increase production without compromising financial stability.
COVID-19, Policy Interventions and Credit: The Brazilian Experience (G2, H7)
The COVID-19 pandemic caused a global health and economic crisis to which governments responded with massive policy interventions. We investigate whether and how the COVID-19 pandemic and ensuing policy interventions affect local credit. We use Brazil as a testing ground because, although the country has been strongly hit, there has been fundamental disagreement between the federal government vs. state/municipal governments on how to react to the pandemic. Policy intervention decisions in Brazil were ultimately taken at the municipality level, implemented in a staggered way, and varied in speed, duration and intensity. First, we perform panel data regression analysis at the bank-municipality level to examine how the local pandemic severity influences the monthly credit. We find that the pandemic has a significantly negative impact on local credit. Second, using a novel dataset on local policy interventions in 920 major metropolitan municipalities, we examine how different types of policy interventions influence local credit in a difference-in-differences analysis. We find heterogenous effects of interventions: positive effects of soft intervention (social distancing, mass gathering restrictions and closure of schools and universities) and late reopening, and negative effects of hard intervention (closure of public venues and/or non-essential services) and early reopening. Third, we address potential endogeneity problems in an instrumental variable analysis using pre-pandemic local political preference as instrument for municipal government policy interventions. We show the local political preference is a strong instrument and significantly related to the likelihood and restrictiveness of policy interventions. We then confirm the heterogenous impact of instrumented soft and hard interventions on credit. Our results are upheld for orthogonalized policy intervention indicators and in placebo tests for pandemic severity and interventions. The effects are sector-dependent, influenced by pre-pandemic bank and municipality characteristics and stronger for longer intervention duration and higher intervention speed. The evidence suggests clear policy implications.
Credit Constraints and Quantitative Easing (E5, E4)
This paper studies the effectiveness of quantitative easing (QE) in a framework that features credit-constrained agents. In particular, we assess the strength of the portfolio re-balancing channel of QE. Following asset purchases, a portfolio adjustment cost limits the ability of the unconstrained agents to arbitrage assets in their portfolios, thus creating a channel for QE to manifest in the real economy. The strength of this channel is, thus, dependent on the share of the unconstrained agents who intertemporally substitute in response to changes in relative asset prices. We show that a rise in the share of the constrained agents, who own no assets, dampens the strength of the portfolio re-balancing channel, thereby limiting the effectiveness of quantitative easing.
Creditor Rights, Collateral Reuse, and Credit Supply (G2, G5)
Utilizing a change to bankruptcy treatment of repo collateral, I provide causal evidence that strengthened creditor rights increase credit supply and financial instability by increasing the reuse of collateral. I use the 2000’s housing boom and bust as a laboratory and collect data linking dealers’ repledgeable collateral to their lending to mortgage companies. Exposed dealers increased their repledgeable collateral and credit provision to mortgage companies. Mortgage companies responded by increasing originations and pivoting toward non-traditional products. I estimate that the expansion in credit drove a 9% increase in originations and accounted for 38% of defaults on mortgages originated during 2005-2006, consistent with a financial accelerator. This paper changes our understanding of the size of the money creation generated in the repo markets.
Crime and (a Preference for) Punishment: The Effects of Drug Policy Reform on Policing Activity (K0, H0)
Every state imposes harsher punishments for drug offenses committed near schools. These “drug-free school zones” (DFSZs) aim to deter crime from entering schoolyards, but they can also give significant discretionary power to police, as they can expose offenders to additional risk for the same criminal activity. I exploit a sudden reduction in DFSZ size in order to examine offender and enforcement behavior and find an 18% decrease in drug arrests in “de-zoned” areas one year after the reform. There is no displacement of non-drug offenses and majority black neighborhoods have a larger decline in drug arrests. If offenders were significantly deterred by harsher penalties, as the law intended and Becker’s (1968) model predicts, there should be an increase in drug arrests. I therefore conclude that police respond to changes in punishment, where punishment severity and enforcement effort are complementary. Additionally, two concerns commonly associated with the “War on Drugs”view of weakening drug crime penalties, crime displacement and an increase in drug use, are not substantiated in this setting.
Crimes Against Women and Earthquakes (J0, I0)
We analyze the effects of a series of earthquakes that struck Mexico in September 2017 on crimes against women. We use a national municipal-level crime data from Mexico that reports domestic violence, sexual abuse, and rape. We track the effects of the September earthquakes using a difference-in-differences strategy and an event-study design. Using a difference-in-differences strategy, the results suggest that following the earthquakes, rape increases by 62%, domestic violence increases by 75%, and sexual abuse increases by 160%. Then, using an event-study design we observe two main patterns. First, domestic violence follows a U-inverted trend. Second, sexual harassment and rape increase and never returned to pre-earthquakes levels. To our knowledge, this paper is one of the few studies that explore the broader effects of earthquakes on crimes against women outside of domestic violence.
Cross-Border Institutions and the Globalization of Innovation (O3, F6)
We identify strong cross-border institutions as a driver for the globalization of innovation. Using 67 million patents from over 100 patent offices, we introduce novel measures of innovation diffusion and collaboration. Exploiting staggered bilateral investment treaties as shocks to cross-border property rights and contract enforcement, we show that signatory countries increase technology adoption and sourcing from each other; they also increase R&D collaborations. These interactions result in technological convergence. The effects are particularly strong for process innovation, and for countries that are technological laggards or have weak domestic institutions. Increased innovation contracting and exchange of capital are the key channels.
Crowding Out the Shadow: Effect of School Construction on Private Supplementary Education in Taiwan (I2, H5)
Scholars have long argued that the availability of schools (or the lack of them) is one of the drivers for the prevalence of private tutoring in East Asia. However, the official number of private tutoring businesses increased despite large increases in the number of high schools in Taiwan. This paper looks at the causal impact of school availability on private tutoring. I use a novel IV strategy and exploit variations across counties in high school construction to separate out the effect of high school construction on private tutoring through changes in availability. I find that an increase in the probability of getting into a public high school is associated with reductions in households' spending and participation in private tutoring.
Cybersecurity Hiring in Response to Data Breaches (M1, J2)
Do firms react to data breaches by investing in cybersecurity talent? We assemble a unique dataset on firm responses from the last decade, combining data breach information with detailed firm-level hiring data from online job postings. Using a difference-in-differences design, we find that firms indeed increase their hiring for cybersecurity workers. While this effect is statistically significant, the economic magnitude is small, which is consistent with firms' lack of incentives to improve their cybersecurity infrastructure. Further, we collect data from the MIT MediaCloud and Google Trends to measure media and public attention following breach events. We find that firms with greater media and search attention after a breach are three times as likely to post a cybersecurity job. This provides evidence of market failure in cybersecurity investment that can be attributed to information asymmetries, and suggests additional scrutiny can ameliorate this market failures. With an increase in both the value of data as well as the number of cyberattacks, our research provides important insight into how media coverage and public attention can provide proper incentives for firms to make substantive, instead of symbolic, IT investments to safeguard their customer data.
Cyclicality, Asymmetry, and Heterogeneity of the User Cost of Labor: New Evidence from Japan (E3, J3)
The user cost of labor (UCL) is a measure of the allocative wage and is essential for generating business cycle fluctuations in a wide class of macroeconomic models. Besides this appealing feature, an empirical challenge is that one needs a sequence of wages since hiring until separation to estimate the UCL. In this paper, we exploit a unique Japanese data that compiles wages of workers at each tenure, which allows us to estimate the UCL for a longer period at a more granular level than previous studies. Using the estimated series, we investigate the cyclical properties of the UCL. Our main findings are summarized as follows. First, the UCL is procyclical but the degree of cyclicality depends on the length of tenure that is considered in the estimation. Wages of workers at longer tenure are less procyclical, and taking these wages into account reduces the cyclicality of the UCL. Second, the responses of the UCL to exogenous shocks are asymmetric: they tends to be smaller to a contractionary shock than to an expansionary one. This result suggests difficulty in downward adjustments of the UCL. Third, we explore heterogeneity with respect to educational attainment, gender, and firm size. Among others, we find the UCL of non-college graduates are less procyclical than that of college graduates and non-college graduates are subject to larger employment risk over business cycles.
Dealing with Logs and Zeros in Regression Models (C1, A1)
Log-linear models are prevalent in empirical research. Yet, how to handle zeros in the dependent variable has remained obscure. This article clarifies this issue and develops a new family of estimators, called iterated Ordinary Least Squares (iOLS), which offer multiple advantages to address the "log of zero". We extend it to the endogenous regressors setting (i2SLS), and provide simple but effective solutions to address common issues like the inclusion of many fixed-effects. In addition, we develop specification tests to help researchers select between alternative estimators. Finally, our methods are illustrated through numerical simulations and replications of recent publications.
Debt Aversion: Theory and Experiment (D9, C9)
Borrowing and saving decisions are among the most important and economically significant choices people face in their lifetime. An unwillingness to save may have severe implications such as insufficient retirement savings. In the same way, borrowing too much or too little can have negative economic consequences. Debt aversion, defined as unwillingness to take on debt even if economically beneficial, has received increased attention by researchers lately, for its adverse effects on financial decision-making. In this paper, we propose a formal model of debt aversion, and use a novel experiment to elicit and to jointly estimate debt aversion with preferences over time, risk and losses. In the experiment participants can accept or reject different debt and saving contracts, defined over real monetary payments. Using participants' choices, we can identify whether they systematically prefer saving contracts over debt contracts, controlling for time preferences, present bias, risk aversion and loss aversion. To this end, we employ maximum likelihood estimations to structurally estimate the preference parameters of our model of debt aversion. We find that participants are on average debt averse, thus establishing debt aversion as dimension of individual preferences in its own right, that is distinct from other relevant preferences, for instance loss aversion. Further, testing the relation of debt aversion and individual characteristics, we find that debt aversion is negatively associated with cognitive ability, and positively associated with participants' savings. Moreover, we demonstrate robustness of debt aversion to a wide array of alternative modeling specification and influencing factors.
Deciphering Federal Reserve Communication via Text Analysis of Alternative FOMC Statements (E4, G1)
We apply a natural language processing algorithm to FOMC statements to construct
a new measure of monetary policy stance, including the tone and novelty of a policy
statement. We exploit cross-sectional variations across alternative FOMC statements
to identify the tone (for example, dovish or hawkish) and contrast the current and
previous FOMC statements released after Committee meetings to identify the novelty
of the announcement. We then use high-frequency bond prices to compute the surprise
component of the monetary policy stance. Our text-based estimates of monetary policy
surprises are not sensitive to the choice of bond maturities used in estimation, are
highly correlated with forward guidance shocks in the literature, and are associated
with lower stock returns after unexpected policy tightening. The key advantage of our
approach is that we are able to conduct a counterfactual policy evaluation by replacing
the released statement with an alternative statement, allowing us to perform a more
detailed investigation at the sentence and paragraph level.
Decomposing Gender Differences in Bankcard Credit Limits (J1, G5)
In this paper, we examine if there are gender differences in credit card limits by utilizing a unique panel data set linking mortgage applicant information with individual-level credit bureau data. We document that, on average, male borrowers have higher total bankcard limits than female borrowers, even after controlling for credit score, income, and demographic characteristics. Using a standard Kitagawa-Oaxaca-Blinder (KOB) threefold decomposition, we find that 87 percent of the difference is explained by differences in the effect of observed characteristics between male and female borrowers while approximately 10 percent of the difference in bankcard limits can be explained by differences in the levels of observed characteristics. Finally, we use a quantile decomposition strategy to analyze the gender gap along the entire bankcard credit limit distribution from 2006 to 2016. Our estimates show that gender differences in bankcard limits, along with the primary factors that drive this gap, have changed over time and vary across the distribution of credit limits.
Demonetization and Its Discontent: Political Strategy and Competition in Indian Elections (H7, H1)
On November 8, 2016, the Prime Minister of India announced the demonetization of notes that made 86 % of cash in circulation in India. I examine the impact of the prolonged cash shortage that followed on the political competition in the assembly elections within a year of the event. Using a difference-in-difference methodology comparing districts with high and low levels of population per bank, I show that demonetization had significant impacts on the political competition and the voter turnout in the elections that followed in districts more-acutely affected by the policy. Next, using the demonetization event and voter turnout in the past as instrumental variables for the number of candidates running for office and voter turnout, I show that Bhartiya Janta Party (BJP), the political party in power at the federal level that implemented demonetization, benefited from the decrease in competition and increase in voter turnout. We are unable to rule out a nation-wide change in voter preference in favor of BJP. But a comparison of elections within a year after demonetization but in different months suggest that the decrease in competition and the increase in voter turnout due to demonetization explains all the advantage BJP had in most-affected districts.
Determinants and Consequences of Poor Decisions in Health Insurance (D9, I1)
This paper aims at understanding decision patterns and welfare effects of poor decisions in health insurance conditional on a large set of sociodemographic characteristics. While giving consumers choice has the potential to improve welfare in principle, the prevalence of empirically observed choices that deviate from utility-maximizing behavior in health insurance markets questions the validity of such arguments. We exploit the highly regulated nature of contracts with only six distinct deductible levels and standardized covered services across contracts and providers in the Swiss mandatory health insurance market to identify optimal and non-optimal coverage levels at the individual level based on a range of standard and behavioral decision models. Using population representative survey and register data from 16,380 individuals, collected by the Swiss Federal government, we show that consumers lose up to USD 1,200 (USD 420 on average) annually due to non-optimal deductible choice. We identify high levels of heterogeneity, indicating that it is particularly the low-income share of the population demanding non-optimally high coverage and facing high financial losses and that the saliency of health issues (e.g., via the presence of long-term chronic diseases) increases the probability to choose optimal coverage levels. Our results highlight the heterogenous adverse welfare effects of choice in complex settings conditional on a choice menu including non-optimal options in general and have implications for policy in the Swiss mandatory health insurance scheme.
Diagnostic Expectations and Macroeconomic Volatility (E7, E3)
Diagnostic expectations have emerged as an important departure from rational expectations in macroeconomics and finance. We offer a first treatment of diagnostic expectations in linear macroeconomic models. We present a general proof of the existence of a rational expectations representation, which can be used to solve these models. Under some conditions, diagnostic expectations generate higher volatility than rational expectations. In a quantitative medium-scale DSGE model, we find that diagnostic expectations generate 38\% extra output volatility. Moreover, we discuss how the combination of diagnosticity with imperfect information can rationalize under- and over-reaction in macroeconomics.
Disclosure Law and External Audit Demand: Evidence from Latin America (G3, M4)
This article relies on difference-in-differences to empirically identify the effect of disclosure law changes on external audit demand by private firms in Latin America. Results indicate stronger disclosure law reduces the probability of external audit choice by medium-sized treated firms relative to their untreated counterparts. The finding supports agency theory’s prediction that disclosure law and audit choice are substitutes. It implies public policy in the region influences information asymmetry in medium-sized firms. Unintended consequences may occur if principals in these firms do not rely on the law to enforce disclosure
Discordant Relaxations of Misspecified Models (C1)
In many set identified models, it is difficult to obtain a tractable characterization of the identified set, therefore, empirical works often construct confidence region based on an outer set of the identified set. Because an outer set is always a superset of the identified set, this practice is often viewed as conservative yet valid. However, this paper shows that, when the model is refuted by the data, an nonempty outer set could deliver conflicting results with another outer set derived from the same underlying model structure, so that the results of outer sets could be misleading in the presence of misspecification. We provide a sufficient condition for the existence of discordant outer sets which covers models characterized by intersection bounds and the Artstein(1983) inequalities. Furthermore, we develop a method to salvage misspecified models. We consider all minimum relaxations of a refuted model which restore data-consistency. We find that the union of the identified sets of these minimum relaxations is misspecification-robust and it has a new and intuitive empirical interpretation. Although this paper primarily focuses on discrete relaxations, our new interpretation also applies to continuous relaxations.
Disentangling the Direct and Behaviour-Mediated Effects of Weather on COVID-19 (I1)
We investigate the causal impact of weather on the COVID-19 infection rate disentangling its direct (biological) impact from the indirect effect via the endogenous response of social activity. Combining mobile locations, weather data, and COVID-19 confirmed cases in a transformed two-way fixed effects model, we find that in the U.S. the effect of temperature in curbing the spread of the virus is substantially attenuated by individuals spending more time out of home in warmer days, especially when COVID-19 incidence is lower. Wind speed and precipitation have no relevant direct or indirect effect. As estimates are robust to an alternative definition of social activity based on the number of visits to indoor venues, we conclude that our results are not driven by individuals spending less time indoors on warmer days. Our estimates also suggest that school closures and lockdowns are effective in reducing infections.
Distribution-Free Assessment of Population Overlap in Observational Studies (C1)
The credibility of causal inference with observational studies relies crucially on the overlap of baseline covariates between different treated groups, which is also known as the positivity or common support assumption. The current empirical assessment of overlap is typically based on estimated propensity scores. This approach is meaningful only when the propensity score model is correctly specified, and in general, it has no formal statistical guarantee due to the lack of proper uncertainty quantification. In this work, we formally define a measure of population overlap inspired by the strict overlap condition (e.g. that propensity scores lie in [0.1, 0.9] almost surely), and develop a family of upper confidence bounds on this measure. We call them O-values. The O-values are valid in finite samples without any assumption on the data generating process, as long as the observations are independent and identically distributed. Technically, we construct the O-values based on a non-standard partial identification approach, with the uncertainty quantification handled by computable concentration inequalities.
Distributional Benefits of Government Spending at the ZLB (E6, E5)
Should we expect larger government spending multipliers from the COVID-19 fiscal packages given that monetary policy is constrained by the zero lower bound (ZLB)? We provide new evidence from quantile regressions and analyze them in a non-linear DGSE model.
Ramey and Zubairy (2018) showed that mean fiscal multipliers are larger during ZLB periods using 140 years of US data. In a related framework estimated with quantile local projections together with instruments for fiscal shocks, we find quantile fiscal multipliers to be larger than one during ZLB periods, but only conditional on being in the upper quantiles of future GDP growth. This means that, during ZLB episodes, fiscal stimulus increases upside risks to GDP more than it improves downside risks to GDP.
We then build a non-linear macro DSGE model able to replicate the historical distribution of GDP, with negative skewness both in its growth and gaps. As such, we also provide a framework to analyze the 5th percentile of the GDP growth distribution, i.e. Growth-at-Risk. Our simulations of periods of fiscal spending at the ZLB are consistent with the data: fiscal spending has a larger impact at the ZLB, especially in the upper tail of the future distribution of GDP with multipliers above one. The simulations further show that the larger multipliers during ZLB episodes are observed mostly when initial conditions are bad and when the economic recovery is protracted.
Our results suggest an insurance benefit of government spending: if the crisis is more damaging than expected and the recovery lasts longer, then fiscal multipliers are more likely to be around or above unity. This suggests that fiscal stimulus during the COVID-19 crisis is likely to be very effective with multipliers above one.
Distributional Welfare Effect of Inflation under Endogenous Search Model (E3, E5)
This paper examines the classic question of monetary economics: measuring the welfare cost of inflation. We extend the New Monetarist framework originated by Largo and Wright (2005) with price dispersion following Wang (2016) by adding heterogenous productivity which is the source for heterogenous search cost among buyers and affect their search intensity. Our model suggests that agents with different productivity choose different search intensity according to their opportunity cost of search. Low productivity workers search harder, as a consequence, there is positive externality towards high productivity workers since they benefit from sellers' posted price dispersion. Overall, inflation has a distributional welfare effect on heterogenous agents.
Diversity & Inclusion: A Tale of Two Economies (E3, E2)
With the improving health picture, analysts are predicting a quick bounce-back in economic activities/back to the “normal.” We present a new framework to characterize the pace of the labor market recovery by race and gender. Our study would help design policies that account for disparities beneath aggregate statistics and facilitate policymakers’ transition towards a framework that benefits all.
This study estimates the asymmetric economic damages from the COVID-19 pandemic for different segments of the population. That is, Black Americans were affected the most in terms of, relatively, higher and faster unemployment compared to other races. Furthermore, Black women were affected more than Black men. Moreover, the pace of recovery is slower for Black Americans compared to other major races, and again, Black women are standing at the bottom in the recovery phase. Additionally, unlike other races, more Black women than Black men are in the labor force, which intensifies the problem of uneven pace of recovery for Blacks.
A similar conclusion is estimated for the Great Recession (GR) where Black women suffered the most among races/gender. Additionally, the pace of recovery for Black women was the slowest among any race/gender. Therefore, the slower recovery in the Black labor market may have pulled the national recovery down/slower pace. Furthermore, we estimate that if the Black labor market would have followed the same pace as what the national market observed, then the pace of labor market recovery would have been faster than the actual recovery.
Policymakers should incorporate diversity & inclusion (D&I) instead of the current tradition of one-policy-fits-all. Accounting for the disparate impacts of economic downturns and the uneven pace of recoveries would help to achieve faster recoveries and place less burden policymakers’ traditional toolkit.
Divisia Monetary Aggregates and Monetary Policy in a Small Open Economy, the case of Singapore (E4, E5)
Since Barnett (1978) derived the user cost price of money, the theory of monetary aggregation has been developed and extended to a number of factors. Divisia monetary aggregates have been proved to be strictly preferred to the simple-sum counterparts. However, most central banks in the world, including that of Singapore, the Monetary Authority of Singapore (MAS) still reports its monetary aggregates as simple summations. We build a New Keynesian framework for a small and open economy with banking sector, and examine monetary policy rules targeting trade-weighted exchange rate index. Monetary policy in a small open economy is examined based on the framework of Gali & Monacelli (2005). We introduce the banking sector and Divisia monetary aggregates as in Belongia & Ireland (2014). The combination of these two framework allows us to explore the case of a monetary policy rule that targets the exchange rate rather than the conventional interest rate or money supply. In the calibration, we construct Divisia monetary aggregates for Singapore based on the available data set from Jan 1991 to Sep 2019. We choose Singapore as it is a typical small open economy in which the monetary authority (MAS) explicitly pegs the domestic currency on a basket of strong foreign currencies.
Do as I Say and Do as I Do: Paternalism and Preference Differences in Decision Making for Others (C9, G1)
Money managers justify the cost of advice by offering portfolio customisation for each client’s risk tolerance (Bernstein, 1992; Campbell and Viceira, 2002). However, instead of tailoring investments, managers often direct clients to invest according to their own risk preferences, rather than those of their client, which has caused significant underperformance in manager led funds (Linnainmaa et al, 2021). We study whether money managers overrule their client’s risk preferences paternalistically when investing for them. We also examine how differing risk preferences between managers and clients affects investments. In an online experiment, we test whether managers disregard their client’s risk preferences, by explicitly informing managers of their client’s preference before they invest on their behalf. In a Gneezy and Potters (1997) investment task, participants invest in a risky project first for themselves, then for another participant. When investing on their behalf, participants have complete information about their recipient’s risk preference, but no material stake in the decision. We use the strategy method to systematically vary the difference between decision-maker and recipient’s risk preferences within-subjects. We also elicit EQ and SVO to test whether those high in empathy adopt the recipient’s preferences more and whether those with more concern for recipients impose their own preferences out of paternalism. From preliminary results, we find that only 7% of managers project their preferences onto investments, whereas 13% follow their client’s preference exactly. However, overall managers do invest further from their client’s wishes, and closer to their own, the more their risk preferences differ. These findings add to the current policy debate about the merits of United States money managers assuming fiduciary duty for clients. Indeed, our results suggest that even if managers did act as fiduciaries, only a minority would follow the client’s preferences to the letter, and some may disregard them all together.
Do Children Perform Better in Religious Schools? Evidence from Population Data (I2, I0)
Religious schools enjoy a high academic reputation among the parents in many societies. Previous studies that assessed the effect of religious schools were conducted in countries where religious schools are private or where they charge fees and set admission criteria. As a result, the effect of religious schooling could not be separated from the effect of private schooling and selection. We contribute to the literature by using Dutch data that include the entire population of children born between 1998 and 2007. In the Netherlands, both public and religious schools are publicly funded, schooling is free of charge and admission is independent of the child’s religious or ideological character. Using a range of models including fixed effects models, coarsened exact matching, and treatment effect bounds, we compare school outcomes of children in religious versus public schools. Our results indicate that children in religious schools outperform children in public schools in primary education. The benefits of religious schooling were largest for children in Orthodox Protestant, Islamic and Hindu schools, which mostly attract children from a disadvantaged socioeconomic background. However, the influence of religious schooling fades out by the end of secondary education.
Do Corporates Set Pension Discount Rates Strategically? (G3, G0)
Corporations can choose to reduce the magnitude of their pension liabilities through their choice of pension discount rates. We document that the majority of U.S. firms set pension discount rates above benchmark rates. Since mandatory contributions to underfunded pensions constrain corporate investments, setting higher pension discount rates improve firm value of more productive firms. Consistent with this idea, we find higher discount rates help to improve investments of underfunded firms, particularly for financially constrained firms most vulnerable to lower rates. Imperfect elasticity of pension discount rates to market interest rates offers firms leeway to alleviate the constraints from defined benefit pension plans.
Do Cost Reminders Lead To Healthier Decisions? Experimental Evidence on the Use of Nudges to Reduce Tobacco Intake in Rural Bangladesh (I1, D9)
Two behavioral change interventions from a field experiment that nudged a rural-poor population in Bangladesh to reduce their tobacco intake are assessed. The first intervention reminded participants of their cumulative tobacco expenses using record-keeping logbooks, while the second intervention provided visual posters on the harmful health effects of tobacco consumption on the self and children. Although logbooks reduced smoking tobacco intake from increased salience of financial costs, male participants are found to substitute cigarettes with cheaper smokeless tobacco. In contrast, participants in the poster intervention made no such substitution. Educated male participants with children less than five years of age, and those who felt greater guilt and shame, further lowered their tobacco intake, suggesting altruism to be a key motivating factor. The study provides a cautionary tale that cost-saving behavioral interventions aiming to promote healthy decision-making should be designed keeping in mind the range of substitutes available on the market.
Do Institutional Investors Drive Female Critical Mass within a Firm’s Lifecycle and its Impact on Performance (G3, G3)
This paper examines the effect of having three or more female representatives on a board of directors (“critical mass”) and its implications on firm performance. Drawing on insights from institutional theory and agency theory, we examine whether institutional shareholders promote or discourage female critical mass to improve gender diversity. We document that higher institutional ownership reduces the likelihood of hiring boards with female critical mass, and this manifests predominantly among firms that have greater level of agency problems (i.e. free cash flow). Additionally, we find that these effects are prevalent in boards during CEO turnover time. We also find that, female critical mass boards are significantly positively related to performance. We employ an instrumental variable approach using firms’ addition to S&P500 index as an instrument to account for potential endogeneity in the examined relationship between female critical mass and institutional ownership. Based on our findings, we conclude that higher ownership by institutional investors negatively impacts on female critical mass, which in turn may worsen board effectiveness.
Do Languages Generate Future-Oriented Economic Behavior? Experimental Evidence for Causal Effects (D9, D1)
Studies show that the use of languages which grammatically associate the future and the present tends to correlate with more future-oriented economic behavior. We take an experimental approach to go beyond correlation, and to identify causality. We asked bilingual participants, people fluent in two languages which differ in the way they encode time, to make a future-oriented economic decision. The participants who were addressed in a language in which the present and the future are marked more distinctly tended to value future events less than participants addressed in a language in which the present and the future are similarly marked.
Do Major Government Customers Help U.S. Firms Escape Foreign Competition? (H5, F1)
This paper studies whether firms that generate a substantial share of their revenues from the government are more resilient to foreign competition. Using the United States granting Permanent Normal Trade Relations to China as an exogenous shock, we show that government contractors have better performance than their peer firms in industries that are more exposed to import shocks. The results are robust to the inclusion of a broad set of controls, instrumental variable analysis, matching analysis, and alternative proxies of government customer and import competition. There is also a stronger effect for firms with greater financial constraints and less corporate diversification. We further show that government procurements allow firms to maintain a higher level of investment and investment efficiency and enjoy lower cost of bank debt when facing increased import competition. Overall, this study provides further evidence on the effect of the government’s participation in product market.
Do Private Schools Increase Academic Achievement? Evidence from France (I2)
While the relative effectiveness of private and public education systems is the subject of a large research literature and is at the heart of public policies in many countries, empirical work on the subject in France is lacking. I use propensity score matching on a large French database to estimate the effect of enrollment in a private school on academic achievement as measured by ninth grade test scores in three school subjects (mathematics, French, and history-geography). I find that private school attendance has a large and significant effect on educational success. Boys’ (girls’) scores in private school were between 0.19 (0.13) and 0.23 (0.19) standard deviations (SD) higher on standardized tests in ninth grade. A series of checks confirm the robustness of these results. Moreover, the results show that it is boys with low test scores in sixth grade who benefit the most from the positive effects of enrollment in a private lower secondary school.
Do Tax Deferred Accounts Improve Lifecycle Savings? Experimental Evidence (C9, H3)
Tax deferred accounts (TDAs) are an increasingly popular method of saving for retirement, and have become common across many developed countries. Nevertheless it is unclear whether TDAs actually improve a household's lifecycle savings behavior and retirement preparedness because it is difficult to perform a counterfactual analysis. Households always have other means of savings so there is no guarantee that a TDA, with its inflexible restriction that funds cannot be drawn until retirement, will be attractive or improve lifecycle savings. In this paper we resort to laboratory experiments to address the question of whether TDAs improve lifecycle savings by comparing experimental treatments where subjects have access to TDAs with treatments where they do not, and we are also careful to consider the tax consequences of our different treatments as well. We find that the presence of TDAs substantially improves a household's lifecycle savings behavior, making them better prepared for retirement.
Do Women Political Leaders Enhance Government Financial Conditions? Evidence from U.S. Cities. (H7, G4)
Limiting the increase of public debt is a key policy issue in most economies, especially after the fiscal policy responses to the COVID-19 crisis. This paper analyzes the effect of electing female political leaders on the financing cost of local government debt. To address endogeneity of political leadership, we collect gender information on mayoral candidates and employ a regression discontinuity (RD) design to analyze 604 female-male contested mayoral elections in 336 U.S. cities during 1990-2014. Using municipal bond issuance data provided by Thomson Reuters Securities Data Company (SDC), we find that municipal bond yield spreads of cities narrowly electing women mayors are 33 basis points lower than that of cities electing male mayors. This effect is robust to specifications using different functional forms of the assignment variable, alternative bandwidth choices, using various measures of bond yield spreads, and controlling for bond characteristics. More importantly, we show that our results are not driven by party affiliation. To explore the potential mechanism behind female mayors lowering municipal financing cost, we show that the gender effect is more pronounced in cities with higher risk of financial distress. Moreover, we find that female mayors reduce government debt and improve the fiscal stability of cities they govern. Specifically, female mayors lead to lower debt per capita ratios and lower debt to revenue.
Does accounting for children lead to higher optimal income redistribution? Overlapping generations model approach. (H3, E2)
We study the optimal design of the tax-benefit system in a framework that accounts for income risk and endogenous fertility. Children in our set-up are public good due to PAYG social security. We calibrate the model to the US. We show that the relationship between the welfare effect and the redistribution scale via labor tax is hump-shaped. The current tax system in the case of the US offers too low redistribution. Implementing the optimal tax generates large welfare gains. Endogenous fertility channels additionally boost welfare gains because higher redistribution leads to higher fertility levels. We also show that the optimal tax scheme depends strongly on the family policy structure, and redistributive child-related transfer, like child allowance, can substitute for the redistribution build-in the labor income tax scheme. Keeping tax system structure in a current shape and increase expenditure on child-related transfers by 2% of GDP leads to similar welfare gains as implementing optimal labor tax.
Does Bank Competition Increase Bank Liquidity Creation? A State-Level Perspective (G2)
One purpose of regulations regarding bank competition is to encourage depressed local credit markets. Does enhanced competition through bank deregulation revive local economy? Exploiting staggered bank deregulation events in the United States, I document that state-level bank deregulation does not, on average, significantly affect state-level bank liquidity creation, while bank-level analyses demonstrate that enhanced bank competition decreases bank liquidity creation. In addition, I find that states and banks respond to state-level deregulation events differently. My results suggest that the policy, which is applied to all heterogeneous banks and states in the same way, does not fit all.
Does Civil Forfeiture Fight Crime? Evidence from New Mexico (K4, H7)
This study examines civil forfeiture’s impact on crime rates. Proponents of the policy, which allows law enforcement to take and permanently keep property without a criminal conviction, claim it is an essential crime-fighting tool, particularly in the War on Drugs. Critics challenge the crime-fighting efficacy of civil forfeiture and warn the policy violates individual liberties. Previous research has associated increases in civil forfeiture with higher rates of drug arrests, but to date, no study has examined the impact of a significant civil forfeiture reform. Using a five-year panel of monthly crime rates, I studied the impact on crime of legislation that eliminated civil forfeiture in New Mexico. This study did not find sufficient evidence to conclude civil forfeiture effectively fights crime. Specifically, when the policy was eliminated in New Mexico, crime rates did not worsen compared to control states.
Does Consumer Monitoring Reduce Corporate Tax Evasion Along the Supply Chain? Evidence from Mongolia (H2, H8)
This paper tracks the effects of consumer monitoring on firms’ tax evasion along the supply chain. To do so, I study a Mongolian government program, which incentivises consumers to report their purchases. First, I estimate the effect of the program on corporate income tax (CIT) and value-added tax (VAT), by comparing retailers who are directly affected, and wholesalers, who are only indirectly affected. I find that retailers increase their reported sales, but partly offset this by artificially inflating their costs on CIT returns. As a result, retailers’ CIT liabilities increase by 11%. In comparison, their VAT liabilities increase by 31% because VAT is less prone to such cost manipulation. Second, I find that the program also increases the VAT liabilities of upstream firms by about 15% when they are more likely to sell to (monitored) retailers, compared to the upstream firms that sell to firms that are not directly monitored. The program does not, however, affect the upstream firms’ reported CIT liabilities. My findings highlight the enforcement advantage of VAT compared to CIT and that consumer monitoring enhances the self-enforcing mechanism in VAT along the supply chain.
Does Fluctuation in Past Flows Generate Time-Varying Loss Aversion of Mutual Funds? (G2, G4)
This study investigates how past flows affect fund managers’ dynamic risk attitude. Using detailed mutual fund holdings within the US market, we measure fund managers’ time-varying loss aversion based on an asset allocation model which considers both their rational expectations and risk attitude. We find that mutual fund managers are prone to the disposition effect. That is, loss aversion decreases after lower past flows but increases after higher past flows. The positive loss aversion-flow relationship is more pronounced among fund managers with higher loss aversion in nature, poorer past performance, or more illiquid assets. In addition, we document a strong seasonal effect: during the last quarter of a calendar year, the positive relationship between loss aversion and past flows becomes weaker. However, further conditional on performance of the first three quarters, past winners become even more loss averse whereas past losers tend to be less loss averse for breaking even. We show that such behaviors from fund managers are consistent with time varying risk attitude and fund tournament phenomena.
Does IT Help Startups? Information Technology in Banking and Entrepreneurship (G2, E4)
This paper analyzes the increasingly important role of information technology (IT) in banking for entrepreneurship. We show that IT in banking spurs entrepreneurship. Job creation by young firms is stronger in US counties which are more exposed to IT-intensive banks through their historical geographical footprint, especially for firms in industries that rely more on external finance and have low startup capital. Entrepreneurs that use their homes as collateral benefit disproportionately more from IT in banking when house prices rise. We also show that IT adoption of banks reduces the role of distance between the headquarter and lending location in small business lending, suggesting that IT adoption can mitigate information frictions.
Does Paid Family Leave Save Infant Lives? Evidence from California (I1, J1)
One goal of the paid family leave program in the U.S. is to help working parents balance their
careers and family responsibilities and hence improve the well-being of their infants. A large
body of literature evaluates the effects of California’s Paid Family Leave program (CA-PFL) on
early childhood outcomes, but most studies have been based on the analyses of surviving infants.
If the CA-PFL reduces infant deaths, then such analyses would understate the program’s true
effects. Using the linked birth and infant death data in the U.S. with a difference-in-differences
framework, I find that the implementation of the CA-PFL reduced the post-neonatal mortality
rate by 0.135 (per 1,000 live births), or it saved approximately 339 infant lives in California from
2004 to 2008. The effects were driven by death from internal causes and there were larger effects
for boys than girls. These results are stable across a variety of robustness checks, and additional
examinations give little reason to believe that the results are induced by the endogeneity of
policy, simultaneously shocks, and fertility changes.
Does Real Earnings Management Adversely Affect Analyst Coverage and Forecasts? (M4, G1)
Cohen et al. (2008) provide evidence that, after the implementation of the Sarbanes-Oxley Act (hereafters, SOX) in 2002, firms tend to switch from accrual-based earnings management methods to real earnings management methods to manipulate their reported earnings. We investigate whether, in the post-SOX era, real earnings management adversely affects the coverage and forecasts of financial analysts who play a key role as information intermediaries between firms and investors in stock markets. We find that the extent of real earnings management is negatively associated with the number of analysts covering and forecasting earnings for firms. This association is less pronounced for firms that exhibit a high degree of accrual-based earnings management, thus reconciling with Cohen et al.’s finding that firms utilize accrual-based earnings management and real earnings management as substitutes in managing earnings. We find no evidence that real earnings management reduces the informativeness of, or increases the error in, analyst forecasts. This suggests that, given an analyst’s decision to cover firms that engage in real earnings management, s/he does not compromise on the quality of her/his forecasts. In aggregate, the reduced analyst coverage, albeit not in company with an increase in forecast error or a decrease in forecast informativeness, would potentially undermine the analysts’ overall information-intermediary role in stock markets, thereby deteriorate the overall information environments of firms, and weaken capital market efficiency. Our study thus calls for the importance of scrutinizing and curbing real earnings management.
Does Scarcity Reduce Cooperation? Evidence from Rural Tanzania (O1, C9)
Cooperation is essential to reap efficiency gains from specialization, not least in poor communities where economic transactions often are informal. But cooperation is difficult to sustain, especially under destitute conditions, since defecting from a cooperative equilibrium can yield safe, short-run benefits. In this study, we investigate how food scarcity affects the level of cooperation by leveraging quasi-experimental variation in economic conditions induced by the harvest in rural Tanzania. We document a significant increase in food consumption between the pre- and post-harvest period and show that lean-season scarcity depresses socially efficient but personally risky cooperation in a lab-in-the-field experiment.
Does the Marketing Experience of Firm Executives Promote Corporate Innovation? (M3, O3)
A vast literature examines how varied economic environments and firm characteristics promote corporate innovation via improving innovative skills, facilitating financing, reducing innovation risks, and enhancing monitoring on research and development (R&D) processes. The commercialization of innovative products and services is vital to the realization of their economic value and thus to success in innovation for a firm, yet little research attention has been paid to the factors that determine innovation via influencing the commercial potential of innovative products and services. Based on a sample of Chinese listed companies from 2009 to 2018, we examine whether the marketing experience of executives, which plausibly facilitates commercialization of innovation outputs, spurs corporate innovation. We find that firms having a higher portion of marketing-experienced executives exhibit a higher degree of innovation. The finding is robust to using propensity-score matching, coarsened-exact matching, firm-fixed-effects regression, two-stage least squares regression, the impact threshold for a confounding variable test, difference-in-differences regression analysis, a dynamic panel generalized method of moments estimator, and a placebo test to control for potential endogeneity, and is also amenable to using alternative measures of corporate innovation and of executive marketing experience. Furthermore, we find that the impacts of innovation on corporate performance and productivity are more positive for firms with a higher portion of marketing-experienced executives. Additional analysis reveals that the positive effect of executive marketing experience on innovation is stronger for high-growth firms, financially constrained firms, firms confronted with fierce product market competition, firms with a high portion of executives experienced in R&D, and firms that have high stock option holdings by executives. Our results suggest that the marketing experience of executives strengthens firm innovation as well as its contribution to improving corporate performance and productivity. This thereby highlights the benefits and importance of having experienced marketers in a firm’s top management team.
Dominant Currency Dynamics: Evidence on Dollar-Invoicing from UK Exporters (F4, F3)
How do the choices of individual firms contribute to the dominance of a currency in global trade? Using export transactions data from the UK over 2010-2016, we document strong evidence of two mechanisms that promote the use of a dominant currency: (1) prior experience: the probability that a firm invoices its exports to a new market in a dominant currency is increasing in the number of years the firm has used the dominant currency in its existing markets; (2) strategic complementarity: a firm is more likely to invoice its exports in the currency chosen by the majority of its competitors in a foreign destination market in order to stabilize its residual demand in that market. We show that the introduction of a managerial fixed cost of currency management into a model of invoicing currency choice yields dynamic paths of currency choice that match our empirical findings.
Downward Wage Rigidity In a Liquidity Trap (E3, E4)
A large strand of literature finds that there exists downward nominal wage rigidity in the economy – there is more resistance to nominal wage cuts than to raises. Since wage cuts are more likely to happen during economic downturns, a natural question to ask is how the asymmetry in rigidity affects the severity of recessions. In this paper, we focus on the most severe recessions – depressions that are coupled with deflation and a liquidity trap, such as the Great Depression and the 2008 financial crisis. Our finding is surprising: downward wage rigidity may lessen the severity of a depression.
Conventional wisdom suggests that in a recession, wage rigidity may cause a rise in real wages, which lowers labor demand and exacerbates the decline in output and employment. However, we find that this is only part of the story. A rise in real wages raises the marginal cost of firms and the purchasing power of consumers, both of which raise inflation expectations. When the economy is in a liquidity trap and the nominal interest rate hits the zero-lower-bound, higher inflation expectations lower the real interest rate and stimulates investment demand. This mechanism works against the conventional mechanism, and offers some relief to the economy. Our finding is based on analyses of a new Keynesian model, revised to incorporate downward wage rigidity and a liquidity trap.
This mechanism makes it necessary to revisit some policy discussions. As an example, we find that the effect of the so-called “forward guidance puzzle” — that the economy responds excessively to the Fed’s forward guidance policies in theoretical models — is substantially weakened within our framework. Our result suggests that the effect of forward guidance policies is most likely asymmetric, more powerful in boom times, but is much more subdued in a liquidity trap.
Driving, Dropouts, and Drive-throughs: Mobility Restrictions and Teen Outcomes (J2, R4)
This paper explores the consequences of teenage mobility restrictions on educational attainment and labor market outcomes. We provide evidence that graduated driver licensing (GDL) laws, originally intended to improve public safety, impact both high school completion and teen employment. Many teens use automobiles to commute to both school and to part-time employment. Thus, the effects of automobile-specific mobility restrictions are ex ante ambiguous. Combining variation in the timing and stringency of GDL law adoption with changes in compulsory schooling laws, difference-in-difference and triple-difference strategies show that restricting teen mobility reduces high school dropout rates and teen employment. These findings are consistent with a model in which teens use automobiles to access educational distractions (employment or even risky behaviors). We develop a discrete choice model that decomposes these effects into a direct component generated by changing educational access and an indirect component due to substitution between activities.
Dynamic Demand of Capital and Labor: Evidence from Chinese Industrial Firms (D2, J2)
This paper investigates how firms dynamically adjust their use of capital, labor, materials and energy using the firm-level data in Chinese industries from 1998-2007. First, we present several stylized facts on factor input adjustments, such as the mix of smooth and lumpy adjustments in capital and labor, and the interrelation in factor adjustments. Based on these evidence, we then propose a dynamic firm model that incorporates the key features: disruption of the production process and reallocation of internal resources when firms adjust capital and labor; fixed costs of installing capital and creating or discontinuing a job vacancy; a convex cost component, aimed at capturing smooth adjustments; and inter-relational cost associated with joint adjustments in capital and labor. The structural parameters in capital and labor adjustment costs are recovered via simulated moment matching method, so the proposed dynamic model is able to reproduce the stylized facts that are directly observable from the firm-level data. Our structural estimations show: (1) existence of joint adjustments is a result of congestion effects, which means that it is more costly for firms to adjust capital and labor at the same time than it is to adjust them separately; (2) for both capital and labors, the convex adjustment cost is the dominant one comparing to non-convex cost; (3) ignoring labor adjustment cost leads to a larger bias than ignoring capital adjustment cost. The main policy implication is that the policy focusing only labor (such as revision in labor law) will also affect firm's investment decision; likewise, the policy only on capital (like modification of capital depreciation in accounting) will impact firm's decision in labor as well.
Dynamic Games and Rational-Expectations Models of Macroeconomic Policies (E1, C3)
We consider a linear-quadratic differential game with two decision makers, which is interpreted as a model of the interactions between the government and the private sector. The open-loop Stackelberg equilibrium solution of this game is characterized analytically. On the other hand, we formulate a linear dynamic continuous-time model with rational expectations. We show that under some assumptions, the problem of determining optimal policies for a government with an economy given by the rational-expectations model is equivalent to the problem of determining the leader’s open-loop Stackelberg equilibrium strategy for the differential game. Consequences for the time inconsistency of optimal policies and for the problem of non-uniqueness of solutions of rational-expectations models are briefly discussed.
Dynamic Patent Portfolio Management (G3, G1)
I propose a tractable model integrating dynamic internal capital allocation with im- perfect patent protection, thereby endogenizing patent war and financing constraints. I emphasize the central importance of capital intangibility for corporate decisions when intangibles are insecure. The main results are: (1) relative to the first-best, im- perfect patent protection introduces the non-diversifiable patent risk as a new factor for internal capital misallocation; (2) the firm manages patent risk via patent portfolio management, litigation, and patent insurance; (3) the endogenous capital reallocation can diminish the impact of the imperfect patent protection; (4) firms on the verge of liquidation tend to use patent litigation to alleviate financial distress; (5) endogenous technology choice play an important rule in patent portfolio management, and (6) imperfect patent protection creates the wedge between average q and marginal q, yet this wedge will be diminished when the firm’s intangibility is high. Remarkably, this paper extends the Modern Portfolio Theory to the real investment case.
E-cigarettes and Smoking: Correlation, Causation, and Selection Bias (I1)
This study investigates the correlation between e-cigarette use (electronic nicotine delivery systems, ENDS) and smoking cessation, examines whether there is selection bias in cross-sectional studies of ENDS and cessation, and estimates the causal effect of ENDS on cessation.
ENDS may encourage or discourage cessation, depending on whether they are complements or substitutes for smoking. Ignoring the issue of selection-into-treatment (i.e., ENDS use) leads to bias in estimates of the causal effect of ENDS use on smoking. Many public health studies ignore the endogeneity problem but nonetheless conclude that ENDS hinder cessation.
I use data from Korea, 2014 to 2018, to establish that there is positive correlation between ENDS use and smoking. Then evidence for selection bias is presented, including changing characteristics of ENDS users and stronger correlation between ENDS and smoking after a negative government report on ENDS.
Causal effects are investigated with parametric (bivariate probit and copula models), control function (moment-based), and semiparametric models for endogenous treatment effects with binary treatment and outcome variables. The causal effect of ENDS on cessation is estimated to be positive: ENDS use is associated with a 16 percentage point higher probability of cessation for male smokers choosing to use ENDS. The preferred model, based on the rotated Clayton copula with Gaussian (probit) marginals, is selected by formal information criteria. The results vary across the many other models estimated, but in a majority of them the treatment effects are roughly similar, statistically significant, and evidence for the endogeneity of ENDS use is confirmed.
The results lead to important policy implications. Some public health and regulatory officials discourage ENDS use because they believe, based on mere correlation, that ENDS causally discourage cessation from tobacco. The present study indicates that, instead, ENDS perhaps should be encouraged as part of cessation efforts.
Early Childhood Human Capital Formation at Scale: Experimental Evidence from Bangladesh (H1, I1)
Can governments leverage existing public service delivery platforms to implement early childhood development interventions at scale? We study a randomized controlled trial of an early childhood stimulation program, executed simultaneously in three different locations in Bangladesh. This program targeted more than 18,000 children aged 3-18 months at baseline -- integrated into the nation's flagship public nutrition program, without any additional financial incentive for program delivery. With partial compliance and take-up rate of 50%, intent-to-treat estimates show short-run improvements in child outcomes along dimensions of cognitive (0.16 SD), language (0.22 SD), and socio-emotional development (0.12-0.15 SD). The program also led to improvements in nutritional outcomes, reducing wasting and underweight rates by 4.8 and 3.5 percentage points, respectively. The intervention led to positive intra-household spillovers on sibling's primary school attendance. Complementary increases in parental investments and the take-up of the pre-existing nutrition program are key mechanisms underlying the treatment effects. A cost-benefit analysis suggests an internal rate of return of 19.4%, driven by economies of scale and low program costs.
Early Female Marriage and Sex Differentials in Child Healthcare and Nutrition (I1, J1)
Preferential treatment of boys at early stage of life is an important issue in the son-preferring societies of the Indian Subcontinent. In this study, we examine to what extent this prevailing gender bias is associated with the practice of early female marriage. Using data from four Demographic and Health Surveys, we study the association between early marriage among Pakistani women and differential peri- and post-natal child healthcare, nutrition and child development outcomes. At issue is whether early marriage has a role to play in the gender discrimination that begins in the womb, and whether boys and girls are treated differently when parents allocate nutrition and healthcare resources. We find that early marriage is significantly associated with several healthcare, nutrition and child development outcomes. The sex of the child too is significant in some estimations, showing the prevalence of son preference. However, the effects of maternal marriage age on child outcomes are found to be not gender-specific. Whether or not a woman married before reaching the age of 18 does little to modify the male gender bias prevalent in the society. This finding underscores the strength of existing patriarchal social norms which still reward women with greater say for bearing sons.
Early-Life Experiences, Altruism, and Stock Price Crash Risk (G4, G3)
In this paper, we study the effect of chairs with nature disaster experience on firm-specific stock price crash risk. Using a Chinese sample, our empirical analysis indicates that firms led by a board chair experienced the famine significantly have lower crash risks. Such a result is robust to the estimations addressing endogeneity concerns, including tests using the heterogeneous in the excess death rate among different provinces of China and Wenchuan earthquake as experiments. Reinforcing bad news hoarding statement, we document that the experience is negatively associated with the information disclosure quality by investigating the impact of the famine experience on restatements of financial reporting, earnings management, and earnings smoothing. The results document the experience can improve the disclosure quality, and therefore reduce the firm’s crash risk. Further, we find evidence that board chairs’ nature disaster experience can make them do more corporate social responsibility, enhance firm’ donation and live near Confucianism centers. Taken together, this study advances our understanding how a board chair’s nature disaster experience influences firm-specific stock price crash risk and identifies the channel of altruistic attitudes associating board chair’s experience and decision-making.
Eco-Certification: Warm Glow or Cold Prickle? (D9, Q5)
Eco-certification programs, which allow firms to display an eco-label on a product if it meets some environmental quality standard, have become a prominent complement to, and sometimes substitute for, direct environmental regulations. This paper analyzes two common approaches to modeling why consumers are willing to pay extra for certified products: one approach has consumers experience positive utility, or “warm glow,” from benefiting the environment; the other has them experience negative utility, or “cold prickle,” from damaging the environment. The two approaches yield identical predictions, provided consumers’ reference point for measuring benefits or damages is treated as fixed.
Social psychologists Miller and Monin (2016) argue, however, that many puzzling behavioral phenomena involving moral choices can be unified by assuming that people divide such choices into two categories. In situations perceived as “moral opportunities,” choosing the moral option enhances one’s moral self-image, giving rise to warm glow, while not choosing the option leaves one’s self-image unchanged. In situations perceived as “moral tests,” however, choosing the option leaves one’s self-image unchanged, while failing to choose the moral option diminishes one’s moral self-image, giving rise to cold prickle or guilt. Miller and Monin suggest, moreover, that these moral frames are highly context-dependent, varying across people as well as time.
By modeling a market in which eco-certified goods compete with non-certified ones, I show formally that, when such moral framing is allowed for, and when the reference point relative to which consumers experience either warm glow or cold prickle is modeled as endogenous to other consumers’ choices in plausible ways, ambiguity of market, welfare, and environmental outcomes becomes pervasive. It arises when assessing the impact of introducing eco-certified goods, when determining optimal certification standards, and when comparing eco-label programs’ effects to those of policy alternatives.
Economic Impact of the Most Drastic Lockdown During COVID-19 Pandemic---the Experience of Hubei, China (H1, C2)
This paper uses the panel approach proposed by Hsiao, Ching, and Wan (2012) and Hsiao and Zhou (2019) to assess the evolution of economic consequences of the drastic lockdown policy in the epicenter of COVID-19---the Hubei Province of China during worldwide curbs on economic activity. We find that the drastic 76-day COVID-19 lockdown policy brought huge negative impacts on Hubei’s economy. In 2020:q1, the lockdown quarter, the treatment effect on GDP was about 37% of the counterfactual. However, the drastic lockdown also brought the spread of COVID-19 under control in little more than two months. After the government lifted the lockdown in early April, the economy quickly recovered with the exception of passenger transportation sector which rebounded not as quickly as the rest of the general economy.
To the best of our knowledge, this is the first study identifying the dynamic macroeconomic consequences of the COVID-19 lockdown policy in the epicenter of China in times of pandemic. The counterfactual are constructed in the absence of “lockdown”, in times of pandemic not in the absence of pandemic. To get a whole picture, we assess the consequences on total output and its sectoral components, its private components, and the transportation sector. Second, our method does not involve modeling how COVID-19 spread and how the COVID-19 and other factors have affected the evolution of GDP, investment, or retail sales. Our results do not depend on the specific model specifications assumed in some current studies. It appears that as the world is facing the complex situation of global anti-epidemic efforts, our estimated results could shed light on how to formulate effective public policy in times of pandemic crisis.
Effect of Transparency on Corruption: Evidence from USA (H7, H1)
Transparency international constructs index about the transparency of policies across the world. But there is no transparency index for the states of USA. In this paper we construct transparency index for different states from the information available on the website of the state governments. These include various parameters as whether people are allowed during decision making process, how much information reaches people etc. Then we rank the states according to the index constructed. We look at the relationship between transparency and corruption. For corruption we use corruptions convictions index (CCI), corruption perceptions index (CPI) and corruption reflections index (CRI). These indices have been constructed by Dincer and Johnston . Corruption convictions index (CCI) is based on the convictions of public officials and is constructed from the data of the Public Integrity Section of Department of Justice. Corruption Perceptions Index (CPI) is based on survey of perception of corruption in three branches of government viz. executive, legislative, and judicial. Corruption reflections index (CRI) is a state level version of the method used by Gentzkow, Glaser and Goldin (2006). CRI is based on the corruption stories covered in the Associated Press (AP) news wires which are electronically available via LexisNexis. Given all these indices we regress corruption and transparency. We take state weather, income, population as controls. Our initial results suggest that more transparent the laws of the state, the less corruption is there in that state.
Effective Training Through a Mobile App: Evidence from a Randomized Field Experiment (O3, Q1)
Farmers in developing countries usually lack access to vital resources and services that facilitate the adoption of new technology and better farming practices. Agricultural extension services, including technical training, are important methods to overcome these deficiencies and reduce poverty by providing information and transferring knowledge to farmers. We study the impact of providing technical training through information and communication technology, an easy-to-use mobile application (app), on farmers’ technical knowledge and welfare. We develop a mobile app that contains and disseminates technical training videos for each farming stage for grape farmers in rural China. In addition, we provided aspirational videos via the same app which demonstrated success stories of farmers who had adopted the techniques being taught through the app. We find that app training improves farmers’ knowledge about the new production technique by 0.58 SDs and improves the sweetness of grape by 0.28 SDs. While there is no effect on yield, we find that farmers with app training increases sales volume by 34% and revenue by 40%. Farmers with access to the aspiration videos show an increase in aspiration for producing more sweet grapes. Our mobile app provides a less labor-intensive alternative that may help many technical assistance services in developing countries.
Effectiveness of Relocation Incentives for High Skilled Professions (I2, H2)
Effects of Capital and Liquidity Requirements in a Macroeconomic Dynamic Model of Banking (G2)
This paper studies the quantitative impacts of Basel-style capital and liquidity requirements on bank lending, bank liquidity holdings and interbank trading activities. We develop a model in which banks are subject to business cycle variations, are financed by deposits and equity, and transform these liabilities into loans, liquid bonds and interbank lending. Banks are exposed to credit and liquidity risks, both made up by systematic and idiosyncratic components, where idiosyncratic liquidity risks can be coped with through the interbank market. We contribute to the literature by investigating the Basel-style capital and liquidity requirements from a macro-prudential perspective, i.e. evaluating their performance considering a system of banks, while the existing literature has merely focused on micro-prudential (single bank) aspects. We find that (1) banks’ capital is pro-cyclical while their liquidity is countercyclical; therefore, liquidity requirements are important in mitigating liquidity risks in economic expansions, (2) there are U-shaped relationships between social welfare and capital and liquidity requirements, indicating the existence of optimal capital and liquidity requirements, (3) introducing liquidity requirements reduces interbank rates and interbank trading volume, indicating that these liquidity requirements are effective in mitigating banks’ excessive reliance on the interbank market to manage liquidity shortages, as expected in Basel III, and (4) introducing liquidity requirements lowers social welfare but improves banks’ stability by reducing bank defaults and the volatility of bank lending during business cycles.
Effects of Childhood Peers on Personality Skills (I2, J1)
Despite large literature on peer effects, there is relatively little understanding of peer effects on personality skills, which are predictive of subsequent human capital development and a wide range of life cycle outcomes. We fill this gap by investigating whether, and how, childhood peers affect personality skill development of children. To identify peer effects, we use variation in the proportion of disadvantaged children in the classroom, generated by random classroom assignment and parental decision to migrate to urban area for employment. Using administrative data linked to an extensive survey with information on Big-5 personality traits, we find that the presence of disadvantaged peers significantly lowers personality skill development measured by conscientiousness, agreeableness, emotional stability, and social skill. Furthermore, the effects on personality skills are driven by the peers' average personality skills, rather than academic achievements, suggesting personality-to-personality peer effects channels.
Effects of Paid Family Leave on Unpaid Eldercare Providers' Labor and Health Outcomes (J1, I1)
As the size of the older population in the U.S. grows, informal care provided by family and friends remains an important source of long-term care. In addition to demographic transitions, COVID-19 has drastically increased eldercare needs by disproportionately impacting elderly individuals. However, informal caregiving is associated with a reduction in labor force attachment for caregivers of working age. As a measure to support employed caregivers, some states have implemented paid family leave (PFL) policies for increased flexibility of work for caregivers. This paper evaluates the impact of PFL in California and New Jersey on the labor and health outcomes of middle-aged individuals with parents or spouse in need of eldercare using rich panel data. I analyze heterogeneous effects by gender of the caregiver, intensity of eldercare, and type of care recipient. Preliminary findings suggest that PFL lowered the probability of women with an elderly parent or spouse being retired by 4 percentage points. For women with a spouse in need of eldercare, having access to paid leave lowered their probability of being retired by around 11.3 percentage points. There were no significant effects on work hours or wages. In terms of health outcomes, PFL increased men’s depression score when they had a parent in need of care and decreased their depression when they had a spouse in need of care. In addition, the results suggest that the policy increased women’s self-reported physical health when they had a spouse in need of IADL care. I investigate possible mechanisms by exploring the care recipients’ health outcomes, nursing home usage, financial outcomes, and retirement savings. This study contributes to policymakers designing leave policies to address increasing eldercare needs due to aging populations and the pandemic by discussing the differential impacts of paid leave policy on unpaid caregivers’ outcomes.
Emigration and Public Finances (H2, J6)
According to the World Migration Report 2020, the number of international migrants has increased over the last five decades, from 84 million in 1970 to 272 million in 2019. Among all the migrants, approximately 141 million live in Europe and North America, and around 80% emigrated from developing countries. The economic consequences of emigration for a developing country are mixed. On the one hand, emigrants generally embody a younger and higher educated part of the population. For example, 74% of immigrants are in the working-age range of 20 to 64, representing the loss of human capital for the origin countries. On the other hand, emigrants remit substantial funds to family members in their home country. International remittances have grown from an estimated 126 billion in 2000 to 689 billion in 2020.
This paper studies the aggregated impact of emigration on tax revenue in developing countries. Although there is evidence of the impact of remittances on tax revenue, the aggregated fiscal impact of emigration has not been studied. We try to fill this gap in the literature. Using cross-country data from 1990 to 2020, we follow the approaches of Asatryan et al. (2017) and Docquier et al. (2016) to identify the causal effect of emigration on tax revenue. In particular, we will use a set of instrumental variables that include a gravity-based instrument and a weather-related instrument for emigration. The results will depend on which effect is larger: the positive effect through remittances or the negative effect through human capital loss. If the impact of remittances can compensate the revenue loss associated with human capital outflow, then the net effect of emigration would be positive. The findings of this paper could be useful for policymakers to understand the economic consequences of emigration and design policies that increase these benefits.
Endogeneity in Modal Regression (C1, C5)
In this paper, we propose a control function approach to account for endogeneity in a parametric linear triangular simultaneous equations model for modal regression when the conditional mode of unobservable error term on the explanatory variable is nonzero. We adjust the endogeneity with the residuals from the conditional mode decomposition of the endogenous variable as controls in the structural equation, and develop a computationally attractive two-step estimation procedure with the conditional mode independence restriction. Particularly, in the first step, we construct the estimated modal residuals from the reduced-form linear modal regression for the endogenous variable. In the second step, we include the reduced-form residual nonparametrically as an additional variable and propose a three-stage estimation method for the resulting semiparametric partially linear modal regression which has not been fully investigated in the literature. The proposed estimators could be easily solved by virtue of a modified modal expectation-maximization (MEM) algorithm. Consistency and asymptotic properties of the estimators for both parametric and nonparametric parts are rigorously established under generic regularity conditions, where we demonstrate that the parametric estimators are root-nh^3-consistent (n is sample size and h is a bandwidth) and the estimation of the nonparametric component is oracle. Monte Carlo simulations are conducted to examine the finite sample performance of the proposed estimation procedures. Two applications to the real dataset of Return to Schooling and Colonial Origins of Comparative Development are presented to further illustrate the proposed estimators in practice. We in the end develop an adaptive lasso method to select instrumental variables and demonstrate the oracle property of the proposed penalized modal regression model.
Endogenous Product Adjustment and Exchange Rate Pass-Through (E3, F4)
We document how product quality responds to exchange rate movements and quantify the
extent to which these quality changes affect the aggregate pass-through into export prices.
We analyze the substantial sudden appreciation of the Swiss franc post-removal of the
1.20-CHF-per-euro lower bound in 2015 using transaction-level export data representing
a large share of total exports. We find that firms upgrade product quality after the
appreciation. Furthermore, they disproportionately remove lower-quality products from
product ranges. This quality upgrading and quality sorting effect accounts for approximately
one-third and one-tenth of total pass-through one year after the appreciation, respectively.
We cross-check our results with the microdata underlying the Swiss export price index,
which includes an adjustment factor for quality based on firms' reported product replacements,
and obtain similar results.
Endogenous Spatial Production Networks: Quantitative Implications for Trade and Productivity (F1, R1)
In modern economies, production is organized in large-scale complex networks of firms trading intermediate inputs with each other. Larger Indian firms selling inputs to other firms tend to have more customers, tend to be used more intensively by their customers, and tend to have larger customers. Motivated by these regularities, I propose a novel empirical model of trade featuring endogenous formation of input-output linkages between spatially distant firms. The empirical model consists of (a) a theoretical framework that accommodates first order features of firm-to-firm network data, (b) a framework for structural estimation based on maximum likelihood that is uninhibited by the scale of data, and (c) a procedure for counterfactual analysis that speaks to the effects of micro- and macro- shocks to the spatial network economy. In the model, firms with low production costs end up larger because they find more customers, are used more intensively by their customers and in turn their customers lower production costs and end up larger themselves. The model is estimated using micro-data on firm-to-firm sales between Indian firms located across 141 districts. The model's fit is good. The estimated model implies that a 10% decline in inter-state border frictions in India leads to welfare gains ranging between 1% and 8% across districts. Moreover, over half of the variation in changes in firms' sales to other firms can be explained by endogenous changes in the network structure.
Endogenous Technology, Scarring and Fiscal Policy (E3, E6)
This paper analyzes the macroeconomic effect of COVID-19 in a New Keynesian DSGE model with endogenous technology growth. I model the epidemic shock as a combination of lockdown policies, a shock to the labor supply and a demand shock triggered by the risk of virus contraction in the exposed sector. I show that due the relative strength of these channels inflation decelerates also over the medium-term. The epidemic induces a drop in investment in R&D and technology adoption which depresses technology growth and causes substantial supply-side scarring effects. These spillovers to the non-exposed, productivity-improving sector generates a deep and persistent crisis with permanent output losses. The zero lower bound constraint further intensifies the severity of the crisis and the scarring of the supply side because of the role of monetary policy in not only stabilizing demand but also the long-run trend path, emphasizing the role of fiscal policy in the context of the epidemic. Fiscal support to investment in technology in the non-exposed sector is effective in sustaining demand during the crisis without accelerating the virus spread, while at the same time alleviating the scarring effects caused by the epidemic.
Escaping Secular Stagnation with Unconventional Monetary Policy (E7, C9)
We design a new experimental framework to study policy interventions in secular stagnations and liquidity traps using an overlapping-generations environment where participants form expectations and make real economic decisions. We explore the ability of unconventional monetary policy to lead economies out of deflationary traps and away from a binding zero lower bound. We observe that participants are able to coordinate on high inflation full-employment equilibria. Permanent exogenous deleveraging shocks induce pessimistic expectations that precipitate persistently deflationary episodes. These shocks generate considerable consumption heterogeneity and also change subjects' forecasting heuristics to make them more backward-looking. We experimentally test policies aimed at re-anchoring expectations and the economy on the high inflation steady state. Permanently increasing the central bank's inflation target is insufficient to generate inflationary expectations. Eliminating the zero lower bound, on the other hand, is consistently effective at stimulating spending and generating the necessary inflation for the economies to escape the zero lower bound. Negative interest rates are more potent than raising the inflation target at shifting consumption to the present. Our findings suggest that inflation expectations and demand is better stimulated through realized wealth effects than coordination on rational expectations equilibria.
Estimating Effects of Dividend Tax Policy Changes (H2, G3)
At the beginning of 2021, we witnessed the high market volatility related to individual investors' activities, which reflected on specific stocks, including GME, AME, BB, etc. The events also lead to the block of trading from brokerage agencies. Compared with the U.S. market, the Chinese financial market has far more (appropriately 160 million in 2019) individual investors. To increase financial market stability, provide an incentive for listed firms to distribute dividends, and encourage long-term investment, on June 13, 2005, the Ministry of Finance and State Administration of Taxation of the People’s Republic of China jointly issued a document (Caishui 2005 No. 102) to lower the dividend tax rate from 20% to 10% for all investors. By conducting difference-in-differences and propensity score matching estimations with machine learning algorithms, I find that, after the 2005 policy change, lowering the dividend tax decreased the turnover rate and the number of trading volumes by over 18% and 68 million shares respectively. Therefore, this paper initiates the exploration of how policy variations affect investors' behavior.
Estimating Policy Functions in Payments Systems Using Reinforcement Learning (C7, E5)
This paper uses reinforcement learning (RL) to approximate the policy rules of banks participating in a high-value payments system. The objective of the agents is to learn a policy function for the choice of amount of liquidity provided to the system at the beginning of the day. Individual choices have complex strategic effects precluding a closed form solution of the optimal policy, except in simple cases. We show that in a simplified two-agent setting, agents using reinforcement learning do learn the optimal policy that minimizes the cost of processing their individual payments. We also show that in more complex settings, both agents learn to reduce their liquidity costs. Our results show the applicability of RL to estimate best-response functions in real-world strategic games.
Estimating Policy-Corrected Long-Term and Short-Term Tax Elasticities for the United States, Germany, and the United Kingdom (H2, E6)
We estimate the elasticities of the most important tax categories with the help of a new quarterly database of discretionary tax measures for the US, Germany, and the UK over the period 1980Q1 to 2018Q2. Employing Romer & Romer’s (2009) narrative approach, we construct a policy-neutral dataset based on revenue figures from governmental records. Using this quantitative information, we are able to subtract policy-induced changes, which are typically not considered in the extant literature. Furthermore, we estimate state-dependent elasticities and adopt a more intuitive approach to measuring SR asymmetries based on the phase of the business cycle. In addition to ‘booms’ and ‘recessions’, we define a 'neutral' business cycle situation. Such a setup avoids setting a slightly negative/positive deviation from the long-run trend equal to a large recession/boom. In this regard, our findings can be summarized as follows: (i) In Germany and the UK, long-term tax-to-base elasticities are generally higher than short-term elasticities, whereas results for the US are mixed. (ii) Short-term elasticities for base-to-output elasticities tend to be smaller than unity, whereas long-term elasticities are close to unity. (iii) Germany and UK tax-to-output elasticities in the short term are lower than long-term elasticities, with mixed results for the US. (iv) For tax-to-base elasticities, we find business cycle asymmetries only across countries but not within countries. (v) For base-to-output elasticities, our results suggest few asymmetries across countries and more asymmetries across tax types. (vi) Typically, the above conclusions do not hold for corporate income tax.
Estimating Shadow Policy Rates in a Small Open Economy and the Role of Foreign Factors (E4, E5)
The shadow monetary policy rate (SMPR) is particularly useful to measure the stance of the domestic monetary policy when the short-term interest rate is at its effective lower bound and unconventional policy measures are implemented. We present a methodology to estimate a SMPR for the case of a small open economy based on a dynamic factor model informed with a large dataset comprising monetary variables. It further allows to consider and separately identify the impact of foreign monetary conditions on domestic ones and, thereby, on the SMPR. We study the case of Chile using data that reflects monetary policy actions taken by the Fed and the Central Bank of Chile. Two results stand out. First, under large negative shocks (e.g. the Covid-19 shock or the Subprime crisis) expansive unconventional monetary policies makes the estimated SMPR decreasebelow zero. In normal times, in contrast, the shadow rate is identical to the MPR. Second, we find that the domestic SMPR is mainly driven by domestic factors in the short run and by foreign factors in the long run. This finding lends support to the classic trilemma.
Estimating the Economic Cost of COVID-19 (E3, E2)
With the development of COVID-19 vaccines, analysts are predicting a quick bounce back in economic activities (back to the “normal”) and thereby suggesting that damages from the pandemic are temporary in nature.
Our study presents a framework to estimate economic costs of the COVID-19 pandemic for the U.S. economy. We characterize whether the pandemic-related damages are short lived or long lasting. The potential damages are estimated in terms of losses in major variables, including employment, consumption and GDP growth rates. By accurately estimating near-term damages from the pandemic, our framework would assist policymakers in regard to effective decision making.
To estimate damages from the pandemic, we utilize the pre-COVID potential level of the target variables, GDP for example, as a benchmark and compare these estimates with those that are calculated during the COVID pandemic. The intuition behind this method is that the pre-COVID potential GDP level is estimated using expansion phase growth rates and with the assumption of no COVID pandemic resurgence in the near future. Consequently, these rates are higher than those that are calculated during the pandemic and thereby the gap between these two measures of potential GDP is utilized as a benchmark to estimate the economic cost of COVID-19.
Our study suggests damages from the pandemic are not short lived (potentially long lasting) as the level/trend of potential GDP (and other variables) has shifted downward. That is, the major sectors have shifted to a lower growth trajectory compared to the pre-COVID growth path. These findings have vital implications for policymakers. That is, comparing real GDP with the post-COVID potential GDP estimates would paint an illusion of a stronger recovery. Furthermore, the illusion of a stronger recovery may lead to a sooner-than-appropriate-time policy normalization. Policymakers would avoid policy errors by utilizing our framework in the post-COVID era.
Estimation of a Partially Linear Seemingly Unrelated Regressions Model: Application to a Translog Cost System (C1, D2)
This paper studies a partially linear seemingly unrelated regressions model to estimate a translog cost system that consists of a partially linear translog cost function and input share equations. A simple and feasible estimation procedure is proposed. We show that both our parametric and nonparametric component estimators are consistent, asymptotically normal, and more efficient relative to the single-equation counterparts. We highlight that the relative efficiency gain of the nonparametric estimator for a particular equation, based on the Cholesky decomposition, improves with its position in the system and is maximized when this equation is placed at the end. A model specification test for parametric functional forms is proposed, and how to correct the between- and within-equation heteroscedasticity is also discussed. An Italian banking data set is used to estimate the translog cost system, yielding policy implications for risk management in banking.
Evaluating Marginal Internalities: A New Approach (H3, I1)
Consumers of sin goods often do not take the full health costs of consumption into account. In such cases, they impose an internality on themselves. If a social planner wishes to impose a sin tax, she needs first to measure the money-metric of the marginal internality.
This paper exploits a relationship between the sin good demand's health insurance elasticity and the marginal internality to estimate the latter. It develops a model where the internality arises because of both a self-control problem and biased beliefs about potential health harms. We show that the health insurance elasticity's determinants consist of rational factors and the money-metric of the marginal internality. By using empirical measurements of both the rational factors and the elasticity, we can determine the underlying marginal internality.
The most reliable existing method for the measurement of marginal internalities is the so-called "counterfactual normative consumer" approach, applied by Allcott, Lockwood and Taubinsky (2019, QJE) to sugary drinks consumption. It is a direct method that uses surveys to measure consumers' nutritional knowledge and self-control.
This paper's approach does not rely on surveys and uses different data to measure the same statistic. Thus, it complements the "counterfactual normative consumer" approach. We calibrate the model to sugary drinks consumption and show that our results are within the range of estimates derived by Allcott et al. (2019).
Evidence of Seclusion’s Effect on Suicide: Implications of COVID-19 Economic Interventions and Relocation (I1, H8)
Prior to COVID-19, suicide in the United States was on an upward trajectory. However, preliminary data from the Centers for Disease Control and Prevention show the suicide rate dropped approximately 6.4% in 2020 amid the pandemic. Following the lifetime utility hypothesis of Hamermesh and Soss (1974), we provide evidence to explain why this occurred, showing why some states have historically experienced higher suicide rates and how changing factors amid the pandemic led to less suicide. The relationship between states having smaller population densities and higher rates of suicide is almost one-to-one, and this study is mostly concerned with investigating the effect of seclusion on suicide with population density acting as a proxy. We do this by estimating a log-log suicide mortality panel model using annual data on all 50 states and Washington, D.C. for the period 1999-2019. Our model includes a number of socio-economic, lifestyle, and environmental variables that previous empirical studies suspect influence suicide rates. We test the sensitivity of the estimated effect of population density on suicide with several different specifications to confirm robustness. Our results show that the estimate of population density (-0.77, SE: -1.52 — -0.01) is robust and isolation is the best predictor of a state’s suicide rate. Our analysis of control variables then shows that unemployment (0.10, SE: 0.01 — 0.18) has had the second-largest effect on suicide, historically. Government interventions to enhance unemployment insurance and the mass exodus from cities to rural areas are most responsible for the reduction in suicides amid the COVID-19 pandemic of 2020.
Evolution versus Creationism in the Classroom: The Lasting Effects of Science Education (I2, Z1)
Anti-scientific attitudes can impose substantial costs to public health, the environment,
and the economy. In this paper, we focus on the content of science education as one determinant of (anti-)scientific attitudes that is directly subject to the policymaker. Specifically, we isolate the causal effect of student’s exposure to teachings about evolution theory in science education on (i) their knowledge about evolution at the end of schooling, (ii) their attitudes on evolution in adulthood, and (iii) the probability that they work in life sciences.
In a generalized difference-in-differences framework, we exploit state-level reforms of the coverage of evolution in US State Science Education Standards as a source of arguably exogenous variation of students’ exposure to evolution teachings.
First, we link data on the evolution coverage in State Science Standards with the individual level National Assessment for Educational Progress (NAEP). We find that being exposed to a more comprehensive evolution coverage in high school increases students’ knowledge about evolution at the end of high
school. At the same time, we see that students’ knowledge in scientific areas other than
evolution (placebos) is not affected. Second, we link data on the evolution coverage in State Science Standards with the General Social Survey (GSS). We find that being exposed to a more comprehensive evolution coverage in high school increases individuals' approval of evolution in adulthood. At the same time, we find no effects on non-evolution scientific, religious, and political outcomes. Third, we link data on the evolution coverage in State Science Standards with the American Community Survey (ACS). We find that being exposed to a more comprehensive evolution coverage in high school increases individuals’ probability to work in life sciences in adulthood, but not in other occupational fields. In sum, these findings show that the content of science education has lasting effects on individuals.
Examining Patent Examiners: Present Bias, Procrastination and Task Performance (D9, O3)
We explore the unproductive procrastination behavior of patent examiners, probe whether present-biased preferences cause such behavior, and estimate the magnitude.
This paper focuses on patent examiners' task performance rather than task completion timing to assess present-biased time preferences. Previous economic studies using field data take task completion on or near deadline as evidence for present-bias-induced procrastination. By construct, we use variation in task performance to enhance the power of identification. Intuitively, if a delay is systematically accompanied by substandard performance, it will not be bolstered by option value theory but rather by present-bias theory because postponed decisions become counterproductive.
We set out a quasi-hyperbolic discounting model where a patent examiner is assigned a biweekly quota of patent application reviews and determines the level of effort by the deadline. We test the model's predictions using data on patent prosecution activities in the U.S. We also estimate the time preference parameters of the quasi hyperbolic discounting model. The panel structure of the patent examination data allows us to follow each patent examiner's performance profile. Since the present-bias factors are estimated as the number of patent examiners, the dimensionality problem occurs in the estimation process. We address this issue using a Bayesian inference approach with the Markov chain Monte Carlo (MCMC) method.
The estimation results provide strong evidence that present bias is widespread among patent examiners. Specifically, more than half the patent examiners have a present-bias factor of less than one. We draw two policy implications from our analysis. First, given that attrition rates are significantly higher for less present-biased patent examiners than for more present-biased patent examiners, the employee retention policy should target the former group. Second, reducing the patent examination quota can improve patent examination quality and timeliness. For a pharmaceutical patent, this reform could yield substantial consumer benefits.
Exit Expectations, Time Inconsistency and Optimal Design of Currency Unions (E5, E4)
In currency unions, monetary policies are collectively determined by the member countries. For countries lacking commitment, belonging a currency union can improve the credibility of monetary policy by losing monetary autonomy. Focusing the purely monetary aspect of currency unions, this paper studies the optimal design of a currency union and the effect of rising exit expectations on it. Our main results are two folds. First, a higher exit expectations lets the public anticipate a higher probability of domestic discretionary policymaking in which the authority has incentive to generate surprise inflation. As a result, we have a higher expected inflation along with a lower growth rate in a rational expectation equilibrium. Second, from an advanced country's perspective, a higher exit expectations increases the optimal share of developing countries within the currency union. In other words, a higher exit probability requires a higher share of member countries with dissimilar economic backgrounds within the currency area.
Export by Cohort (F1, L2)
We find a ‘cohort effect’ among new exporters in the same market: firms entering later tend to sell more at the same age. Better sales performance of late entrants indicates improvement in measured firm characteristics. To identify such changes, we rely on structural models of new exporter dynamics at the market level, among which two competing theories stand out: demand learning and customer base accumulation. It turns out that the correlation between sales growth at entry and cohort is a very useful moment for model selection. A major class of learning models features mechanisms through which sales growth is driven solely by a mean-reverting process. Such mechanisms are shown to predict only negative correlations, inconsistent with the data. In contrast, we show that investment mechanisms, such as customer base expansion through advertising, are essential to generating the right correlation. Guided by the qualitative analysis, we build a simple customer base accumulation model with advertising and structurally estimate it. We then validate the model’s capability to replicate the empirical cross-cohort lifecycle trajectories. Our estimates suggest new exporters entering one cohort later on average gain 0.2% in measured productivity and start with a 7% larger customer base.
Female Mayors and Violence Against Women: Evidence from the U.S. (J1, H7)
Women are overrepresented in some crimes. In the United States, more than one in three women experience sexual violence involving physical contact during their lifetimes. Yet, only one out of four sexual assaults are reported to the police. This study examines the effect of female leadership in local government on violence against women. Regression discontinuity estimates show that elections of female mayors decrease violence specifically targeted toward women but leave other violence unaffected. The effect is strong only for local elected leaders and it is persistent throughout her term, while there is no effect for non-local leaders. Moreover, the study explores behavioral responses by victims. Evidence suggests that female victims are more likely to report violence against them after female mayors take office. Importantly, female victories are followed by greater police responsiveness to violence against women. There are no such effects for violence against men. These findings survive various robustness checks. The evidence accords with deterrence that refers to the behavioral reduction in crime due to offender anticipation of punishment.
Financial Autonomy in Small Open Economies (E4, G1)
This paper studies term structures of government bonds and estimates the bonds risk premia to explain stabilized cross-border capital flows in small open economies without effectively independent monetary policy. In Hong Kong (HK), the local currency is hardly pegged against the US dollars, and the monetary authority chooses open financial accounts at the cost of losing monetary independence. HK's Exchange Fund Bills and Notes (EFBN) are estimated to have a similar term structure with the US Treasury Bonds in the sample from 1996Q4 to 2020Q2. The EFBN risk premia, however, deviated occasionally from the US counterpart, suggesting divergent market expectations on HK and US assets. A vector-autoregressive model is employed to explain the EFBN risk premia with HK macroeconomic variables. The result shows that HK domestic price and real output are unable to predict the risk premia, speaking to the absence of policy aim at internal stability in HK. These findings also imply that, in addition to shocks from hard currency and the local economy, the risk premia in small open economies are to be explained by factors related to regional financial markets.
Financial Infrastructure and Micro-enterprise Performance: Evidence from India (G2, O2)
How does financial infrastructure affect micro-enterprise performance? We study this question in the context of India's bank branch expansion policy initiated in 2006. The policy classified regions as ``underbanked'' if their ex-ante branch density was less than the national average branch density, and encouraged private banks to open branches in these underbanked areas. We exploit this threshold-specific treatment assignment rule in the spirit of a regression discontinuity design to identify the causal impact of branch expansion on micro-enterprise performance, using data from a nationally representative survey on unincorporated micro and small enterprises. Focusing on regions located within a narrow window around the discontinuity threshold, our results show that enterprises in underbanked regions witnessed a significant increase in both revenues and value-addition in the aftermath of the policy intervention. The treatment effects are economically significant – an additional private bank branch opening is associated with a 7% increase in enterprise revenues and a 9% increase in value-addition. There was however no change in formal or informal credit for these enterprises along either the extensive or the intensive margins, indicating that the positive impact of branch expansion on enterprise revenues and profitability occurred through spillovers from increased local demand and industry linkages with registered enterprises. Our findings lend support to the view that banks can serve as vehicles for pro-poor growth.
Financial or Non-financial Shocks: Rivals That Play Together (G0, E3)
This paper estimates a model using Bayesian methods and data from the US (1990Q1- 2019Q2) to explore how the financial sector contributes to business cycles through banks’ asset reallocation channel and the quality of capital adequacy constraint. The paper shows that the contribution of financial and non-financial shocks has varied before, during and after the recent financial crisis; housing demand and asset price shocks are the main contributors and the credit shocks are the most persistent. In addition, the paper presents the application of macroprudential tools, along with their impact on the economy in general, and on welfare in particular. The findings illustrate that the tools which control household borrowing ability, such as loan-to-value or debt-to-income ratios, do not impact welfare significantly.
However, the impact of policies on the leveraged sector is substantial. The paper proposes macroprudential policies that allow policy makers to stabilize the economy without changing welfare. Such policies, however, should be timely, targeted and temporary, otherwise they may cause disruptions.
Financial Sanctions and the Transformation of the Global Banking Network (F5, F3)
Since the 2010s, targeted and secondary financial sanctions have become the most powerful and widely used tool of the United States’ economic statecraft. While it contributes to containing money laundering and terrorist financing activities, it is unclear how banks might be affected by these measures and how they will respond? This paper creates an original dataset of over 3,000 active and defunct banking branches and subsidiaries of 39 global systematically important banks (G-SIBs). The dataset traces changes in these banks’ ownership and status between 1990 and 2020. We find that financial sanctions have re-shaped the global banking network. In particular, large European banking groups are hit heavily by these sanctions. Rising compliance cost has forced them to close their branches and subsidiaries in high risk and low yield regions such as the Caribbean, Africa, and the Middle East. When these banks retreat, regional banks and Chinese banks are rising and quickly taking over their positions. Interestingly, American banks have also strengthened their positions in the global banking network during the past ten years. This paper is primarily a descriptive study. It aims to present how financial sanctions have re-shaped the global banking network, and it explores potential new research directions on the banking industry.
Financially Lost Generations: The Liability Burden of Coming of Age During Recessions (G5, J0)
We analyze how entering the labor market during a recession affects personal financial liabilities using a large panel representative of the US population with credit history. Adverse initial labor market conditions deteriorate credit scores, decrease the probability of getting a mortgage, and increase bankruptcy probability. The effects we estimate are sizable and persistent, even after controlling for individuals' incomes.
Financing Mechanism of Green Fund: Behavioral of Government, Social Capital and Enterprise (G4, H3)
In order to adapt to climate change, the Green Climate Fund(GCF) helps developing countries build green projects, but developing countries are still experiencing huge financial gaps. In response to carbon emission target, China has established the National Green Development Fund in 2020, and local governments have established green funds from 2017 to 2021.We summarize that the common model of green funds is that fiscal funds or government-guided funds are used as parent funds, and sub-funds are set up to leverage the participation of social capital. We used assumption of investors' bounded rationality, constructed dynamic evolutionary game models built with three participants, which include the government, social capital and enterprises. Use dynamic evolutionary game model and numerical simulation analysis method to discuss the equilibrium strategy and influence mechanism of attracting social capital investment.
We found that the green investment of social capital is affected by the government's attitude, incentive policies, and risk sharing. Enterprises should make correct use of green funds, and accept the supervision of fund management companies to ensure the environmental effect of green funds. In order to mobilize private, institutional, and commercial financing for climate change mitigation and adaptation, the government should give full play to its leverage in the green fund, focusing on government guidance and market operation, to build greener projects and adapt to climate change.
Firm Digital Adoption during COVID-19 (O3, L2)
Is COVID-19 exacerbating the digital divide across firms? Anecdotally, the pandemic forced many firms to adopt new technologies, as consumers shifted online and bricks-and-mortar retail was disrupted. This paper examines the extent to which COVID led to firm adoption of digital technologies across 24 advanced and emerging economies. The analysis combines novel data on the high-frequency diffusion of website technologies, with firm data on their pre-COVID characteristics. The econometric impact of COVID is identified using weekly differences in the timing of COVID across countries. We find that COVID led to a rapid diffusion of many technologies among firms: e-commerce, online payments, general analytics and A/B testing. For e-commerce, a more basic technology, we find similar trends across firms in different country income levels, but adoption of A/B testing, a sophisticated technology, is much more pronounced for firms in advanced than emerging economies. Similarly, across firm types, we find smaller firms are more likely to adopt e-commerce during COVID, whereas large, productive firms with advanced software and cloud in place pre-pandemic, were far more likely to adopt A/B testing. While COVID has led to broad firm adoption of more basic technologies, the rapid diffusion of sophisticated technologies seems to be more confined to a minority of advanced firms and countries, resulting in heterogeneous post-COVID recovery.
Firm Heterogeneity and Capital Markets (E2, G3)
Capital markets play an increasingly important role for firms’ acquisition of funds in the most recent years. I identify external equity financing shocks by applying a novel approach, so called granular instrumental variable, to infer a shock sequence without having to rely on specifying a theoretical model. The paper analyzes the heterogeneous responses of firms investment, sales, and debt choices along several dimension of the firm distribution. I find financially constrained firms, measured as young non-dividend-paying firms, benefit relative more when equity financing conditions are endogenously improved.
After controlling for the cylicality of equity issuance along the firms size dimension, I find smaller firms increase their sales by more relative to large firms.
Firm-Bank Relationships: A Cross-Country Comparison (G2, E5)
We document the structure of firm-bank relationships across the Euro area countries and present new stylised facts using novel data from the recent credit registry of the Eurosystem. The main findings are the following. First, the number of firm-bank relationships varies significantly across firm size and across countries in the EA. Even though the number of relationships is increasing with firm size in all countries, the absolute number of relationships differs across countries. The median firm in Italy, Spain, and Portugal has at least one more bank relationship than a firm in Ireland, the Netherlands, Germany and Austria, controlling for the firm size and industry. Second, the median firm in all countries appears to have a main bank that lends at least half of the firm's total credit. Still, there is a significant heterogeneity across countries and firms size on the credit concentration, with firms in Spain, Italy, and Portugal to depend less on their main bank compared to firms in Ireland and the Netherlands, where the share of the main bank is almost equal to one. Lastly, differences in the credit contracts are also identified. In Spain, Portugal, Italy, and France debt contracts with shorter maturities such as trade receivables correspond to a larger share of the total credit, while in Germany and the Netherlands most of the credit is allocated in contracts with longer maturities, such as collateralized loans and revolving credit. Such differences appear to be correlated with the structure of the firm-bank relationships and at the same time could have important implications for investment decisions in long-run projects. Overall, the paper highlights the noteworthy heterogeneity in the firm-bank relationships across the Euro area countries that is linked with the growth of the corporate sector, but also with the firms' exposure to a bank shock.
Firm-level Political Risk and Debt Choice (G2, G3)
In this study, we examine the effect of firm-level political risk on debt financing choices. Using a U.S. sample, our investigation indicates that firms with higher political risk significantly prefer issuing private debt to public debt. Such a finding is robust to the tests addressing endogeneity concerns, including a test using the disagreement among the congress representatives as an instrumental variable and a difference-in-difference test using the beginning of Iraq War as the experiment. We identify several reasons of our findings and provide supportive empirical evidence. Firstly, our results are more pronounced when a borrower has higher default probability, lower asset liquidity, and higher financial statement transparency in a period of high policy uncertainty. These findings are consistent with the notions that private lenders are advantageous because they can efficiently manage the reorganization process when the borrowers are financially distressed, and because they can continue updating their information regarding a borrower when the policy uncertainty rises. Next, we find evidence that there exists an implicit contract between a borrower and its relationship bank, following which a borrower accepts less favorable terms during normal times in exchange for the lender’s support during difficult times. Taken together, this study advances our understanding of how political risk influences financing decisions and highlights the cross-sectional heterogeneity in the level of political risk facing individual firms.
Fiscal Stimulus and Firms: Profitability during the Global Financial Crisis (E6, H6)
This paper examines how fiscal stimulus impacted corporate profitability during the global financial crisis. Using Worldscope data for more than 66,600 non-financial firms in 52 advanced and emerging economies between 2008 and 2010, it finds that a decrease in the structural fiscal balance has a significant effect on the profitability of the sampled firms. A decline of one percent of potential GDP in the structural balance is associated with a 1.2 percentage point increase in the profit (relative to total assets) of the panel average firm. The effect has a greater statistical significance in advanced economies compared to emerging ones, and in Europe and Asia than in other regions. It differs across industries, being felt most significantly in manufacturing and construction industries and having key implications for the fiscal response to shocks at industry- and firm levels.
Follow-Thy-Neighbor? Spillovers of Asset Purchases within the Real Sector (D2, E5)
Market participants are interconnected and events which affect some entities can disseminate through the economy due to agglomeration spillovers, local aggregate demand effects or competition between firms. By exploiting the first large sovereign bond purchase program of the ECB, the securities market program (SMP), this paper investigates whether asset purchases induce spillover effects between treated and non-treated firms. I compare firms linked to banks which held SMP eligible assets to firms linked to banks which did not hold SMP eligible assets and take spillover effects between and within the two groups into account. When ignoring spillovers in the assessment of real effects, there is no difference between treated and control firms. In contrast, when taking spillovers between firms into account, directly treated firms invest less and induce negative spillovers on firms operating in the same industry and region. The paper shows the importance to consider spillover effects when assessing the real effects of unconventional monetary policy: when neglected, spillovers can cover up direct effects.
Forecasting with Panel Augmented Macroeconomic Data using Recurrent Neural Networks and AutoML (E3, C6)
A fundamental problem in macroeconomics is the external validity of time series models. Typical macroeconomic data only has approximately 300 quarterly time steps, which implies the best reduced-form models are linear. Structural models often require priors and to heavily borrow from microeconomics research to identify parameters. We propose a simple data augmentation: training reduced-form and structural models with a panel of countries, instead of the single country of interest. This augmentation improves the GDP RMSE forecasting performance of AR(2), VAR(1), and VAR(4) models over 1 to 5 horizons by an average of 15 percent, compared to a single country baseline. This also seems to make models more policy invariant, as country out-of-sample forecasts, where data on every country but the country of interest is used, outperforms, achieving roughly half the RMSE performance improvements of using the entire panel of countries. Furthermore, we document similar generalization gains for DSGE models like the Smets-Wouters model, suggesting that the external validity of estimation and calibration could benefit from matching moments on a panel of countries rather than just the United States or other baseline data sets. This procedure is easy to implement and can immediately improve the external validity of most macroeconomic models. Finally, as the effective data size has increased, we test more non-parametric machine learning models on this data set. We use both a recurrent neural network and AutoML, which performs a horse-race of many machine learning models and reports the best performing model on the validation set. We show that before the data augmentation, these methods perform poorly on test sets, but with the augmented data, the AutoML, a rough proxy for excellent non-parametric machine learning performance, outperforms all baseline economic models in 4 of the 5 horizons. This performance is confirmed with robustness checks.
Foreign Aid through Domestic Tax Reforms? Evidence from Multinational Firm Presence in Developing Countries (H2, O1)
This paper studies the effect of developed country corporate tax cuts on economic activity in developing countries. We exploit a major U.K. corporate tax reform as an exogenous shock to foreign investment in Africa. Difference-in-differences estimates show that multinational U.K. firms increase their subsidiary presence in sub-Saharan Africa by 15-20% following the 2009 U.K. tax cut, and these results obtain at both the extensive and intensive margin. Results are robust to measuring effects relative to other multinational firms from countries with similar colonial histories (France and U.K.) or from counties with the same language and similar capital markets (U.S. and U.K.). Exploiting location-specific survey data, we document that these increases in U.K. subsidiary presence are associated with increased employment rates of African citizens within 5 kilometers of the local U.K.-owned subsidiary. Our findings imply that, beyond the common goal of motivating domestic investment and employment, corporate tax cuts in developed countries have economic impact in other less developed countries.
Foreign Reserves, Fiscal Capacity and Lender of Last Resort (F3, O1)
Why do emerging markets accumulate foreign reserves for precautionary purposes while advanced economies do not? In this paper, I construct a theoretical model, similar to Farhi and Tirole (2012), to show that it is optimal to accumulate reserves in equilibrium when a government has under-developed fiscal capacity - the ability to extract resources from its citizens. According to the liquidity literature led by Holmstrom and Tirole, governments have an advantage relative to the private sector to provide liquidity during crises because of their power of taxation. However, some governments don't have the ability to tax the whole economy, due to informality for example, and, thus, compensate by hoarding liquidity. Therefore, reserves are useful to implement lender of last resort policies because they increase an economy’s pledgeability when there is a lack of fiscal capacity. In terms of policy, this paper implies that overcoming currency mismatch or original sin, avoiding sudden stop episodes, and being fiscally credible, without improving fiscal capacity, may not be sufficient for low income and emerging economies to eliminate the need to accumulate foreign reserves.
Fossil Fuel Equity Bias (G4, Q5)
Public markets represent an important channel for raising additional funds for green investments. In this paper, we develop a theoretical framework to describe potential behavioral biases in equity markets that can lead to underinvestment in green assets. Investors’ preference for holding fossil fuel based assets relative to emerging green assets might stem from the existence of an information asymmetry between the two types of assets with respect their expected payoffs and risks. The effort and cost for gathering information on newly emerging technologies are not the same for the already established ones. We rationalize an underdiversified portfolio through a combined learning-investment model. Building upon the literature of investment under incomplete information, we derive a (Merton type) continuous-time stochastic learning-investment model, where investors chose both the optimal degree of attention to private information and portfolio weights. Attention to information is incorporated as an explicit cost, and directly affects the utility of the investor. High information costs reduce the ability of the risk-averse investor to learn about the underlying asset, which would allow her to reduce the uncertainty about the equity’s expected mean-returns. Hence, with a high level of uncertainty and high information costs, investors are inclined to favor more familiar assets. To achieve the targets aimed at limiting the negative effects of climate change there is a need for a greater involvement of the financial sector. Climate policies, on top of the current fundamentals, can be designed to reduce the uncertainty evolving around the performance of green assets.
Free-riding in a Stochastic Partnership with Joint Liability (C7, L2)
We introduce a stochastic stopping game with payoff externality to analyze the free-riding problem in jointly liable partnerships. We find that players may suffer from a "curse of productivity," namely, a more lucrative partnership may backfire by stimulating free-riding behavior. With more than two players, free-riding will likely trigger a "Domino effect," which we base on to characterize the set of group sizes where coordination is possible. We also examine the "No-First-Defect" commitment as a remedy to the free-riding problem and show that it can result in a Pareto improvement by avoiding pre-emption. Our results shed light on how to alleviate the free-riding problem in jointly liable partnerships.
FX Hedging, Currency Choice, and the International Role of the U.S. Dollar (F3, E3)
Our paper studies how foreign-exchange (FX) hedging affects firms’ currency choice, and exchange-rate pass-through. We develop a theoretical model that features dynamic currency choice with FX hedging and incomplete pass-through. Our model predicts that FX hedging favours foreign currency pricing when the benefits from FX hedging are high. We test and quantify these theoretical results by using French firm-product-level data on exports to extra-EU countries and their FX derivatives positions. We find that firms belonging to industries more reliant on FX hedging and with a high share of imports in foreign currency are more likely to use local (destination) currency pricing (up to 10% more). This is especially the case for the US dollar as the probability increases up to 25%. Finally, we document that FX hedging firms using local currency pricing have a lower exchange-rate pass-through into export prices even two years after the shock.
Gain without Pain? Non-Tariff Measures, Plants’ Productivity and Markups (F1, L2)
This paper studies how productivity and markups respond to non-tariff measures (NTMs). We build a novel time-varying dataset on all NTMs applied to imported prod- ucts by Indonesia. Using price and quantity information, we disentangle the impact of NTMs on plants’ technical efficiency and markups. We find that on average, NTMs generate fewer distortions than import tariffs. However, while specific NTMs increase the quality of the products on which they are applied, others act as barriers to trade similar to import tariffs. These results suggest that in order to gauge their impacts and guide policy-making, NTMs should not be bundled together in empirical analyses.
Gender and Moral Hazard in Microfinance (G4, Z1)
This study examines whether social context and norms lead to gender differences in moral hazard behavior among microfinance borrowers by conducting microfinance field experiments in comparable matrilineal and patrilineal societies in India. This paper also takes the first step to analyze the strategic interaction of the constituent components of default which is not explored in the prior literature in microfinance experiments. The previous experiments in microfinance have focused primarily on only one of the two moral hazards (ex-ante moral hazard and ex-post moral hazard) channels through which default occurs while abstracting from the other (Abbink et al. 2006, Kono 2006, Gine et al. 2010). Our experiment can thus be viewed as a generalized version of the previous experiments on moral hazard in microfinance. Our experimental design allows us to isolate the two moral hazard channels by incorporating three treatments which are, the ex-ante game treatment where subjects’ only make decisions over project choice, the ex-post game treatment where subjects’ only make decisions over repayment choice and the full game treatment where subjects’ make decisions over both project and repayment choice. Thus, decomposing the moral hazard channels, we find that women become more prone to ex-post moral hazard when there is no possibility of ex-ante moral hazard and vice versa, but we find no such effect for men. Our most important finding is that in patrilineal societies, women are significantly less prone to exhibit both types of moral hazard behavior than men, but we observe a reversal of gender effect in the matrilineal society. Our results thus suggest that gender difference in moral hazard is driven by the social context, norms and gender roles in different societies.
Gender Bias in Education Expenditure in Iraq: Evidence from Household Socio-Economic Survey (I2, O1)
This paper examines the nature, scope, and plausible determinants of gender discrimination in allocating household resources by examining the education expenditure on boys' and girls' education in Iraq. Using data from the household expenditure survey conducted in 2007, as expected, we find a strong male bias in household resource allocation in Iraq. However, there exist considerable variations in this bias depending on the age of the child, income level of the household, rural-urban divide, and the regions of Iraq. These results suggest that parents' allocation of resources for education expenditure for boys and girls is motivated by the economic interest of the households. This suggests that changing the incentive structure such as targeted employment opportunities for women, better access to childcare and health care, and better physical and social infrastructure would help attain gender equality.
Gender Differences In Reaction To Enforcement Mechanisms: A Large-Scale Natural Field Experiment (C9, G2)
We followed 58,345 borrowers from a peer-to-peer lending platform to study how women and men react to enforcement mechanisms differently. In the experiment, borrowers were randomized into treatments where they received different text messages urging for timely repayment if they had loans due the next day. Compared to a reminder message, the messages inducing social pressures and financial incentives reduced the overdue rate for both genders. However, females were more responsive to messages producing social pressures, while males were more responsive to financial incentives. The results imply the potential importance of a gender-dependent mechanism to enhance compliance.
Getting on the Wagon: Decreasing Alcohol Abuse to Improve Rural Well-being (O1, I3)
A major global social and economic problem is the widespread abuse of alcohol. In western Kenya, a highly populated area with significant levels of rural poverty, illicit sale and consumption of locally brewed spirits is causing significant negative societal impacts: harming health, decreasing agricultural productivity, and leading to intra-household conflict. We implemented a randomized control trial across a wide area of western Kenya in which treatment group households received three days of group training along with a longer period of individual and household-level counseling with the objective to decrease substance abuse. Findings show that training attendance and alcohol counseling significantly decreased alcohol usage relative a control group as measured by both self and spouse reported consumption levels. The study also shows that treated households increased their time spent in agricultural work and decreased reporting of intimate partner violence (IPV) by the spouse relative to the control group. Experimental auctions for agricultural inputs also show those exposed to alcohol counseling have greater willingness to pay for productivity-enhancing agricultural inputs than those in the control group and were more willing to share information about experimental auction winnings with their spouse. A cost-benefit analysis shows that simple alcohol counseling can lead to significant net benefits through increased agricultural productivity and decreased IPV.
Giving a Hand to Those in Need: The Politics of the Paycheck Protection Program (H8, G3)
The Paycheck Protection Program (PPP) is a large relief package by any standard. A natural question is: did the funds flow to where they should? We answer this from the lens of special interest politics. Using a sample of private and public firms, we present first evidence that lobbying is associated with 60 percent larger PPP loans, controlling for firm characteristics and location and industry fixed effects. This is consistent with the notion that lobbying firms have experience in navigating administrative and policy complexity, and can therefore take advantage of a government aid program. We also find that, in areas with greater conservative presence, PPP loans were less responsive to lobbying. This points to an interesting interaction between political ideology and special interests: where the prevailing ideology is inclined toward providing government support (i.e., less conservative), more lobbying can help secure more aid. Further, PPP loans were more responsive to lobbying in areas that were hit harder by the pandemic. This is in line with a benign interpretation: access to policymakers established through lobbing can help facilitate allocation of public funds to where they are needed the most. Finally, we present evidence of feedback effects: firms that receive larger loans in the first round ramp up their lobbying spending. This again could be interpreted in a benign way: as the pandemic continues to ravage the economy, firms that received PPP funds step up their efforts to share information with policymakers on the need to maintain public support to the private sector. A less benign interpretation would be that these recipients realize the advantage lobbying provides and increase lobbying efforts to secure a piece of the pie in further rounds of the PPP. Our analysis shed light on these potential dynamics, although it cannot distinguish between information vs rent-seeking explanations.
Global Confidence, Uncertainty, and Business Cycles (E2, F1)
This paper investigates the role of global confidence cycles, measured as the common factor across a wide range of survey-based business or consumer confidence indicators on global macroeconomic fluctuations over 1970-2019. I estimate a factor-augmented vector autoregression model, where global confidence shocks are identified through recursive restriction. I report three main results. First, the global confidence cycle, in particular that of consumer confidence, has played a key role in global business cycle fluctuations, explaining over a third and over a half, respectively, of total variation in global industrial production and employment over 1985-2019. Second, the results suggest that while global business confidence shocks are in nature demand-driven, global consumer confidence seems to reflect macroeconomic shocks in both demand and supply sides, in line with “animal spirit” and “news” views on the relation between confidence and economic activities. Third, the empirical results suggest that the plunges in economic confidence caused by the COVID-19 pandemic have led to around 25 percent decline in global industrial production by the end of 2020. The results are robust to the ordering of variables or alternative identifying schemes that include external instrument identification.
Global Prices, Trade Protection, and Internal Migration: Evidence from Indonesia (F6, O1)
I study how regions respond to price shocks in the presence of internal migration. I examine Indonesia in the 2000s as it faced a commodity boom for its export commodities and initiated its trade protection on rice, a staple food for its population. I use the variation in the potential shares of palm oil and rice sectors across district economies to measure local exposure to shocks. I find that the commodity boom increased the real expenditure per capita of palm oil-producing districts. These districts also received more migration, providing evidence that palm oil price shocks were no longer localized and internal migration spreads the windfall. However, these relatively higher levels of real expenditure per capita do not last after the commodity boom ends in 2014. Meanwhile, the trade protection on rice did not materialize as higher purchasing power to rice-producing districts. I estimate the overall welfare gains in Indonesia between 2005 and 2010. Gains from migration accounts for one-third of these gains.
Global Share Repurchases Over the Business Cycle (G3, E3)
Using a dataset that covers details of repurchase cases across the globe, we comprehensively analyze how repurchase motives, details, and consequences vary over the business cycle. We find that in economic recession, firms buy back stocks to bring up undervalued stock price or to enhance market liquidity, and their repurchase announcements are accompanied by higher short-term stock returns and a lower completion rate. In expansion periods, firms repurchase to distribute excess cash; their announcements are followed by less inefficient investment, higher long-term returns and a higher completion rate.
Global Value Chains and the Dynamics of UK Inflation (E3, F1)
This paper investigates the relationship between the involvement of the UK economy in global markets and its inflation dynamics. We first document that the UK Phillips curve started to flatten in the early 1980s with a significant change in the slope in the late 1990s and show that this change is strongly associated to trade openness. We then use disaggregate data to show that the interaction between the sectoral and source-country dimensions crucially affect the slope of the UK Phillips curve: we uncover robust evidence that UK industries with higher shares of intermediate imports from Emerging Market Economies have flatter Phillips curves. Interestingly, while integration with China is an important force behind this result, integration with other Emerging Market Economies also plays a significant role in putting downward pressure on UK inflation. Our empirical findings are consistent with a two-country, multi-sector model where firms use intermediate goods from home and foreign industries in addition to the labor in production.
Going Universal. The Impact of Free School Lunches on Child Bodyweight Outcomes (I1, H4)
In recent years, several countries have switched from means-tested to universal provision of free school meals to increase meal uptake, reduce stigma and combat childhood obesity. Evidence on the success of such policies is lacking. We study the impact on young children's bodyweight of a switch to universal provision of nutritious free school meals in England, using unique high-quality data on children’s height and weight collected by trained nurses in schools. We exploit identifying variation in the timing of weight measurements in a difference-in-difference framework which allows us to show how bodyweight outcomes evolve as dosage of free meals increases throughout the school year.
We find that exposure to high quality universal free lunches increases healthy weight prevalence and reduces obesity prevalence and BMI by the end of the first year of school. The effect is driven by substitution of low-quality home-produced lunches with school meals among children not eligible under means-testing, while we find no evidence of income or parental labour supply effects. This suggests universal provision can improve the diets of relatively well-off pupils.
We contribute to the literature by evaluating a policy that is a relevant policy option in the large number of countries that already run high quality, means-tested free meal policies. We also contribute to the wider literature on the relative advantages of universalism versus targeting or means-testing and on the role of in-kind transfers in promoting child welfare. Our results suggest that parents from a wide range of backgrounds face time and/or information constraints in preparing their children’s diets which mean that there is a role for universal food provision in school to reduce childhood obesity.
Grandfathers and Grandsons: Social Security Expansion and Child Health in China (I1, D1)
We examine the multigenerational impacts of a nationwide social pension program in China, the New Rural Pension Scheme (NRPS). NRPS was rolled out in full scale since 2012 and enrollees age over 60 in rural areas are eligible to receive at least 70 CNY non-contributory monthly pension. We use household age eligibility change over time and across households to identify the impacts of NRPS on child health. We find NRPS substantially increases grandchild's weight, but not height. Child BMI z score increases by 0.9-1.1. We find a novel gender pattern that the impacts are likely to be driven by grandfather's pension receipt, and are more salient in boys subsample. Grandsons living with NRPS eligible grandfathers are more likely to be overweight/obese. We explore the mechanism and find the impacts are more plausible to be driven by a mixture of income effects and son preference.
Growing like Germany: Local Public Debt, Local Banks, Low Private Investment (G2, E6)
Using a firm-bank panel of more than 1m German firms over 2010-2016, we document that local public bank lending to municipalities crowds out private investment. Our results show how crowding-out can happen in a developed economy characterized by low interest rates and fiscal austerity. Our mechanism relies on two structural features of Germany's banking landscape: First, the geographical segmentation of credit markets for small and medium firms (SME) which are dominated by local banks. Secondly, a special statutory mandate requiring local public banks to lend to municipalities. With yields on local government debt declining to all-time lows, local public banks tried to alleviate stress on their balance sheets by using their local market power to charge higher rates on their SME customers. This crowded out firm investment. Perversely, fiscal consolidation at the state and federal levels contributed to this effect by putting pressure on the budgets of municipal governments which increasingly borrowed from local public banks. Crowding-out lowered aggregate private investment by around 30-40 bio euros per year (or 1 percent of GDP). Thus, we identify a novel channel through which low interest rates can adversely affect bank lending and firm performance. Our results also illustrate how segmented credit markets can amplify negative multiplier effects from fiscal austerity.
Hadwiger Separability, or: Turing Meets Von Neumann and Morgenstern (D9, D8)
It can be time-intensive to make optimal choices. This is especially true in settings where people make many related choices at once. To ensure that decisions are made in a reasonable amount of time, people often follow heuristics that isolate -- or separate -- some of their choices from others. A sizable empirical literature on choice bracketing and related phenomena provides evidence for this kind of behavior.
This paper incorporates time constraints into decision theory, and uses the resulting framework to better understand how and why people separate choices.
I combine classic approaches in decision theory and computational complexity theory to obtain a model of a time-constrained agent. There are three ingredients to my model: choice under risk, time constraints, and high-dimensionality. I begin with a standard model of choice under risk. Then I introduce time constraints: for any given menu, the agent must express her choice within polynomial time. If a choice correspondence can be executed by an algorithm that respects this time constraint, it is called "computationally tractable". Finally, I specialize the model to focus on high-dimensional problems where the agent makes many related choices at once.
Theorem 1 shows that a time-constrained agent who satisfies the expected utility axioms must have a Hadwiger separable Bernoulli utility function. Hadwiger separability relaxes additive separability by allowing for complementarities and substitutions, but limiting their frequency. Note that, like many results in computational complexity, this theorem relies on conjectures that well-known problems are computationally hard (e.g. P\neq NP).
Theorem 2 provides a partial converse to theorem 1.
Theorem 3 shows that time-constrained agents may be better off violating the expected utility axioms, even if they intrinsically want to maximize expected utility. My result quantifies "better off".
Finally, I explore applications to demand estimation with many products/characteristics.
Happiness and Willingness to Compete: A Pre-Registered Online Experiment (C9, J0)
Willingness to Compete (WTC) has emerged as a key concept in economics. I conduct an online experiment on 645 respondents with real-effort tasks eliciting WTC for different levels of happiness. This represents the first time an online experiment systematically tracks and manipulates happiness in the context of WTC. Despite the study’s power, it can only detect minimal effects of happiness. As expected, male subjects show higher WTC whereas unemployment seems to correlate with lower WTC.
Healthcare Reform: Equity versus Efficiency (I1, I3)
In this paper we conduct an ex-ante evaluation of a health insurance reform in Chile that has an increase in equity as its main objective. Currently, the private system offers actuarially fair insurance policies (the efficiency component). New legislation, however, mandates that premiums should be non-discriminatory with respect to age and gender in the private health insurance system (the equity component), while maintaining premiums the public insurance system, which is determined solely by the individual’s income, unchanged. We develop a heterogeneous agent model of choice of both consumption-saving and private-versus-public health insurance, which we estimate using data on insurance claims in the private system and from Chile’s Household Budget Survey, Employment Survey, National Socioeconomic Characterization Survey, and the Survey of Household Finances. We discuss different microeconomic and macroeconomic outcomes such as health-risk taking, consumption-saving and welfare. The counterfactual analysis findings arise from the resulting risk characteristics of the new pooling equilibrium that would prevail in the private system after the reform. Namely, there are liquidity effects (changes in premiums) in the short term that would favor of high-risk individuals (who would pay less for private insurance, encouraging some to move towards the private system), while the opposite outcome takes place for low-risk individuals who stay in the private system or are forced to move to the public system, which generally has lower coverage. We also find an average positive but heterogenous welfare effect of the reform, which includes a reduced need for subsidizing the public system due to the drop in the risk-skimming behavior traditionally shown by private insurers. This result is an example of a non-discriminatory policy change that can be supported from a traditional economic point of view.
Heterogeneous Effects of Conditional Cash Transfers and Extended School Day Program on Academic Achievement: Evidence from Dominican Republic (I2, O3)
There is abundant evidence documenting the impact of CCTs on educational outcomes focused on enrollment, attendance, dropout, and graduation; but instances that use and find impacts on learning outcomes are still scant. Furthermore, social policy impact studies usually evaluate effects on educational outcomes excluding interactions with education policies. Exploiting the expansion of the extended school day program that reached a coverage of more than 50% of total public enrollment in the Dominican Republic by 2019, and the administration of nationwide diagnostics tests during 2017- 2019 by the Ministry of the Education on three student cohorts (N = 503,828), we implement a regression discontinuity design taking advantage of the conditional cash transfer allocation rules and merging it with the beneficiary registry (Sistema Unico de Beneficiarios - SIUBEN) and households transfers records, providing us a valid comparison framework. We also employ recently developed ML quasi-experimental estimators to study heterogenous effects. Preliminary results indicate that there are no differences in academic achievement of students from beneficiary and non-beneficiary households, but we find heterogeneous positives effects in beneficiaries who have been included in the extended school day program.
Heterogeneous Returns to Medical Innovations (I1, J2)
Recent literature in applied economics has studied the impact of several single medical innovations on health-related economic outcomes. Although these studies rely on credible identification strategies, the scope of the treatments analyzed is very limited. On the other hand, there is an extensive, albeit descriptive, literature studying the efficiency of medical technology in a broader perspective. This paper is setting up a quasi-experiment to estimate both total and heterogeneous impacts of medical innovations on economic outcomes for a broad set of 90 health conditions.
We estimate the impact of medical innovation on economic outcomes as an innovation-induced reduction in income loss due to the onset of a specific disease by using a difference-in-difference-in-differences approach. Rich administrative panel data for Sweden (1978–2006) for 930,000 individuals combined with the annual data on new molecular entities and patents granted in healthcare allow us to mimic this experiment. The control group of individuals belongs to the same cohorts as the treated and experience a hospital admission due to the same cause (across 90 conditions) in the future. Across all conditions, we observe that they follow the same pre-trends if hospitalized a few years apart.
Turning to results, we find a large and positive effect of medical innovations on economic outcomes that is the first contribution to the state-of-the-art. The average effect from a 1 SD increase in medical innovations on real family income is 8.6%. Regarding sources, the main effects on disposable family income are nested in the innovation-induced increase in own and spouse’s disposable income, wages, and a reduction in sickness and unemployment payments. This average effect entails significant heterogeneity (our second contribution), which main sources are chronicity and severity of disease. The effects by year at admission are relatively constant and positive suggesting constant returns to health rather than decreasing.
Hiding Filthy Lucre in Plain Sight: Theory and Identification of Business-Based Money Laundering (K4, F3)
Proceeds from illicit activities percolate into the legal economy through several channels. We exploit international regulations targeting money laundering via the financial sector to identify the flows of “dirty money” into legitimate establishments: business-based money laundering (BBML). Our variant of the monopolistic competition model embeds a drug cartel that channels illicit proceeds into an offshore financial investment and into BBML. Tighter regulations in one channel increase the flow in the other. We use a research design that links U.S. county business activity to the evolution of anti-money-laundering regulations in Caribbean jurisdictions to provide the first empirical evidence of the phenomenon.
High Frequency Trading and Price Discovery in the Foreign Exchange Market (F3, G1)
We study the effect of high-frequency trading (HFT) on the intraday trading of currencies at the platform of Electronic Broking Services (EBS). We use the realized variance-based information share to measure the price discovery efficacy and find that the contribution of HFT is time-varying and is highest during the overlapping trading hours of London and New York, after dividing the 24-hour trading of FX markets into Asian, European, London-NY overlapping and U.S. sessions. Our new evidence shows that the liquidity supply from HFT accounts for the improvement of price discovery attributed to HFT, while the arrival of new information around macroeconomic announcements does not weaken the contribution of HFT to price discovery.
Historical Anti-Semitism and Firm Access to Finance (G2, G3)
This paper analyzes the effects of anti-Jewish pogroms in the historical “Pale of Settlement” on the access to informal finance of present-day firms in East Europe. While historical anti-Semitic pogroms are associated with higher general trust among the present-day residents, this runs in parallel with a lower trust in finance, with this financial-antipathy mechanism dominating the net effect. Firms located in regions with a higher historical intensity of pogroms tend to experience higher financial constraints, accompanied by a reduced access to external credit, and a lower propensity to invest. These effects diminish with distance from pogrom activity, which further indicates a cultural link to local historical anti-Semitism. Our results are not driven by the experiences of the Holocaust or a Communist regime. Our causal interpretation of the long-lasting effects on financial development of historical discrimination against Jews is supported by the results of identification strategies based on instrumental variables and regression discontinuity design.
Holy Cow! Religious Violence, Informal Cattle Trade and Externalities in India (O1, Q1)
We study how religious violence breaks down informal markets, with economic costs for large sections in India. Coinciding with the rise in Hindu nationalism, India recorded an exponential increase in violent attacks on persons suspected of trading cattle (mostly Muslims) for slaughter by cow-vigilante groups in the last decade. We show that this violence led to a drop in cattle trade and that rural households’ inability to sell unproductive cattle led to increased cattle abandonment. We construct a novel dataset using a quadrimestral representative panel of Indian households, a state-level panel of road accidents, media-reports of vigilante violence, and historical data on Hindu-Muslim conflicts in India. Exploiting the temporal and spatial variation of violence, we causally estimate its impact on the cattle trade market with an event study design. We show that violence led to a 20% decline in cattle holdings among households in the affected regions. Next, we causally identify the effect of violence on the number of stray cattle using a proxy measure: the number of road accidents due to stray cattle. We construct a Bartik instrument for vigilante violence using historical Hindu-Muslim conflicts (1950 - 2000) to find an alarming 400-600% increase in road accidents due to stray cattle in affected regions. The results are consistent with predictions from our theoretical model. Finally, we examine externalities on farmers by conducting a primary household survey in Rajasthan, the largest state, with spatial variation in the violence. We find that farmers self-report large crop damages from stray cattle and precautionary costs in regions experiencing violence.
Hospital Ownership and Exposure to Bankruptcy Risk: Evidence from the COVID-19 Impact (G0, I1)
This paper assesses the impact of COVID-19 on the financial stability of hospitals distinguishing by ownership type. We model the COVID-19 impact on the liquidity and solvency metrics of hospitals and identify operational, structural and cyclical factors affecting the bankruptcy risk across counties and census tracts. Most of the typical financial performance data are low frequency. To provide more timely indicators of hospital's financial stress in response to the COVID-19 pandemic, we use high-frequency daily visits to healthcare facilities tracked by smartphones to predict operational indicators that are historically associated with hospital financial distress. We calibrate a model that allows us to estimate a stable relationship between mobility indicators and operational indicators and use it to assess the financial impact of the pandemic. We find that the pandemic affected hospitals in 2020 differently according to their ownership structure. In particular, investor-owned hospitals experienced significantly more financial distress than others. We predict that 27.55% of all hospitals will become financially distress after 2020, up 0.69 percentage points from 2019. In contrast, 37.80% of investor owned hospitals are predicted to become financially distressed, up 5.86 percentage points from 2019. Since investor-owned hospitals are the main providers of specialty treatment such as psychiatric and acute long-term care, the increased financial distress among these hospitals is likely to result in long-term effects on the patients seeking the specialty treatment.
Household Constraints and Financial Decisions (G5, G2)
We propose debt-wealth constraints to separate all households into constrained and unconstrained groups. A constrained household is subject to debt payment dues with limited wealth. However, an unconstrained family has cash reserves. Since cash reserves can absorb investment losses, only unconstrained households can exercise lifelong mean-variance optimizations on risky investments. Constrained (unconstrained) households are conservative (aggressive) in their risk tolerance. We test differential decisions through fixed income mutual fund data. When funds deliver low returns, conservative investors run, but aggressive investors stay engaged. Fund flows to funds with high-risk exposure are significantly higher than flows to funds with low-risk exposure.
Household Savings and Monetary Policy under Individual and Aggregate Stochastic Volatility (E5, E3)
We study a heterogeneous-agent model with sticky-prices in which total factor productivity and individual productivity are subject to stochastic volatility shocks. Agents save through liquid bonds and illiquid capital and shares. To construct equilibrium, we use a deep learning algorithm. Our method preserves non-linearities, which is essential for understanding portfolio choices. With rich heterogeneity at the household level, we are able to quantify the impact of uncertainty across the income and wealth distribution. We find that persistent high levels of uncertainty increase wealth inequality, and that in response to a contractionary monetary policy shock, illiquid wealth inequality decreases and liquid wealth inequality increases.
Housing, the Credit Market and Unconventional Monetary Policies: From the Sovereign Crisis to the Great Lockdown (E5, E6)
This paper develops a two-country model of a monetary union to evaluate the interaction between housing, the credit market and unconventional monetary policies first during the ECB’s Asset Purchase Programmes (APP) from 2015 until 2020, then, over the ECB’s Pandemic Emergency Purchase Programme (PEPP) in 2020. In this paper the 2020 lockdown is presented as a negative signal from macroeconomic fundamentals which causes labor to grind to a halt. The model is calibrated for the Euro Area and incorporates heterogeneous households, portfolio balance effects, a credit market susceptible to default, and nominal and real rigidities. The model features the housing accelerator and the post-crisis house price double-dip. The findings illustrate the way in which macro-housing channels lead to self-reinforcing loops, affecting the portfolio re-balancing channel as the main way for asset purchases to influence the economy. The results show that asset purchasing performs better during a crisis, particularly if it is conducted for an appropriate extent of time. The findings illustrate that the PEPP should be extended until the covid-19 crisis phase is over and that it alone is not sufficient to accelerate the recovery; more actions, namely targeted fiscal policy, are required. Finally, the APP, PEPP and lockdown are assessed through a welfare analysis.
How Competition Shapes Peer Effects: Evidence from a University in China (I2, J0)
Competition is widely used to increase effort and performance. In many domains, however, performance depends not only on individual effort but also on cooperation between agents. In these cases, competition may decrease individual performance since it may lower cooperation between agents as the chance of winning a competition decreases in the success of peers. Education is a natural setting in which help from others enhances individual performance. Using administrative data from a university in China, this paper examines how competition changes peer effects and peer interactions. Exploiting randomly assigned roommates, we first show that high ability students have moderately detrimental effects on the performance of high ability roommates. More importantly, we provide novel evidence that this negative peer effect increases significantly along various dimensions of competition intensity within a dorm room. Our survey reveals that competition discourages help and interaction among roommates which may drive our findings.
How Do Firms Adjust When Trade Stops? (F1, D2)
Recent trends in deglobalization and trade wars beg the questions whether firm-adjustments that have been observed over the decades of trade liberalization will be reversed in a symmetric fashion.
Indeed, when faced with strong negative demand shocks, firms are likely to adjust on a number of dimensions. Such adjustments might interact with each other and moreover involve substantial degree of heterogeneity as different sectors in the economy are likely to be subject to non-uniform adjustment costs and expectations of the demand shock permanence.
Empirically, many international trade barriers that lead to substantial negative demand shocks are likely to be correlated with the broader economic adjustments. For instance, they may be linked to changes in domestic worker wage expectations and labor supply. Technological shocks may also trigger alterations to trade agreements but are also likely to lead to demand changes directly or through the production function recompositions.
We explore a unique event when due to political reasons, unrelated to the underlying economic conditions, the exporters lost access to a major export market. In particular, in 2014 Russia introduced sanctions on imports from Europe and this caused abrupt negative trade shock to food production and food wholesaler sectors in Lithuania.
We are guided by a theory where firms are forward-looking and face nonconvexities in the labor market along with convex capital adjustment costs and costly revenue-increasing prospects. By us- ing a unique firm-level dataset, consisting of all exporters in the country, we find that part-time employment is used as the first shock absorber, followed by investment and full-time employment. The sectoral adjustments differ as wholesalers act as intermediaries in redirecting food products to new geographic markets and thus experiencing a positive demand shock for their services.
How Do Firms Invest? The Role of Astrological Consultations on Investment Expectations and Decisions (E2, E7)
Investment is a key determinant of a nation’s progress. A substantial share of national investment comes from the private sector, where firms need to invest both to keep up with technological progress as well as to develop innovations that will give them a competitive advantage in the market. When planning their investments, firms are faced with substantial uncertainty about future market conditions and need to form expectations about the profitability of such plans. How do firms form such expectations?
In this paper we attempt to address the question by running an experiment on New Zealand firms about their investment plans and decisions. We survey 587 firms about their planned one-year-ahead investment expenditures. We find that firms use astrological predictions when forming their investment expectations and making decisions. We treated a random selection of these firms, 212 in total, with exogenous predictions about their future profits. The predictions are obtained from professional astrologers, where the firm’s returns are predicted to be high (>5%), moderate (1%-5%), or low (<1%). We then conduct a follow-up survey, approximately one year after the first survey, asking how much the firm actually invested in the previous year.
We find that the provision of astrological information led to statistically significant changes in firms’ investment decisions. On average, firms receiving the highest return prediction over-invested about 1 percentage point more than the control group. The effect of the treatment was stronger on firms receiving the lowest return predictions. Namely, on average, firms receiving the prediction of a return below 1% tended to under-invest about 2.7 percentage points more than the firms that did not receive any astrological prediction. These estimates are robust to the inclusion of subsector fixed effects and individual-specific controls such as the firm’s size and age, and the respondent’s age, ethnicity and level of education.
How Do High-Skilled International Students Impact Domestic Students? Evidence from the Post OPT STEM Extension Period (F6, I2)
In this paper, I look at how international students impact domestic students’ academic performances in the US higher education. I focus on the post-2008 period when there was a unique shock of international student enrollment due to the OPT-Stem extension. Using administrative data from a large public university, I find that international students have a slight negative affect on domestic students’ grades and graduation rates in the Engineering and Sciences. However, there is also positive returns to education from interacting with international students conditional on graduating in Engineering. In essence, there is a small cost of international students in terms of increased competition for higher grades but these same forces bring positive externality in learning and eventually higher returns to education.
How Do State Infertility Insurance Mandates Affect Labor and Marriage Outcomes? (I1, J1)
Approximately 1 in 8 couples have trouble getting pregnant or sustaining pregnancy in the United States. Despite the recent technological development and increased usage of infertility treatments, medical treatment for infertility problems remains extremely costly. To address a perceived need for coverage, to date 19 states have enacted some form of the infertility insurance mandate. In this paper, I investigate the impact of state infertility insurance mandates on labor-market outcomes and marriage rates of women between the ages of 28 and 42, who are most likely to be affected by the mandates. Using a triple-difference approach applied to data from March Current Population Survey, I find that infertility insurance mandates have reduced the employment rate of mostly affected women, whereas their wages are unaffected. It brings concerns whether such mandates lead to labor market inefficiencies that work against the interest of the group they are meant to benefit. Mandates also bring positive impacts on the stability of marriage and the probability of new marriages, with larger effects on white women and well-educated women.
How Does Maternity Leave Allowance Affect Fertility and Career Decisions? (J1, J2)
The level of compensation during maternity leave varies significantly across countries, yet previous research provides limited insights on its consequences. In this paper, I assess how the generosity of maternity leave allowance affects first-time mothers' subsequent fertility decisions and career trajectory. I exploit the fact that the allowance is capped in Belgium so that women with pre-leave earnings above the maximum threshold face drastically lower replacement rates. Using a regression kink design, as well as a rich set of administrative data on mothers from 2002 to 2015, I find that subsequent fertility increases with the level of benefits: for each additional euro in daily allowance the probability of having a second child increases by 0.6 percentage point. Subsequently, I explore the consequences for their career and show that mothers who receive higher benefits are more likely to leave salaried employment for self-employment. I demonstrate that the transition to self-employment does not affect their earnings in the long run, suggesting that the career changes might reflect non pecuniary preferences. In fact, heterogeneity analysis reveals that those working in sectors with poor work-family balance are more likely to become self-employed.
How Parental Educational Investments Respond to Changes in Ability Belief: An Application of Big Data Techniques in Education (I2)
How parents make their parental educational investment decision is a basic concern for development economists. This paper proposes a novel parental educational investment decision model to explore the situations under which the correlation between ability belief and parental educational investment becomes negative and proves the model predictions with a field experiment hosted in China.
In the decision model, discontinuity is introduced to parents' utility function. Because of the jump in the utility function, parental ability belief can negatively affect parental investment under certain situations. The model predictions are proved empirically using a field experiment.
In the experiment, I introduced exogenous changes into parental ability beliefs by eliminating existing information friction: parents are lack of expertise to predict future performance accurately. I used well-trained machine learning algorithms to predict students’ future performance and delivered this prediction to parents randomly selected.
The reduced form results show that parents are generally overestimating their child’s ability and the treatment significantly attenuates this bias. The treatment is also used as an instrumental variable to estimate the causal effects of ability belief on parental educational investment. the results prove that the correlation is mostly positive, but the sign flips to negative when parents believe their child hasn’t reached their requirement. The empirical results are consistent with the model predictions. Third, IV estimates also show that parental educational investment driven by the intervention is effective.
Human Capital Effects of Spending Time with Parents: Evidence from a Swedish Childcare Access Reform (I1, I2)
Parental time investments and early childhood education are important determinants of children’s human capital development. Differential parental investments are also believed to contribute to higher parity siblings’ human capital disadvantage. In this paper, we study the effects of increased one-on-one time with a parent during infancy on the human capital formation of children, using Swedish administrative data on health and educational outcomes. To this end, we use an exogenous increase in opportunities for one-on-one time induced by a nation wide reform that mandated municipalities to offer childcare access for infants' older siblings, while parents were on parental leave to care for their infants. We estimate the intention-to-treat effect in a triple-differences setting, comparing infants with and without siblings, pre- and post reform, exploiting that some municipalities were affected by the the reform mandate, while others already offered childcare. Survey data on childcare enrollement allows us to establish that the reform had a significant impact, doubling the childcare enrollment of older siblings. This suggests that families made use of the new opportunity for one-on-one time. We find no robust overall effects on the children's 6th grade test scores, but we find evidence of improved test scores for children a the lower end of the test score distribution. These effects are driven mostly by boys and children of low SES mothers. Exploring potential pathways, we find significant reductions in boys' behavioral problems as measured by drug prescriptions and doctor's visits. However, we do not find any effects on maternal stress, earnings or family separations. Results thus suggest that increased opportunities for one-on-one time improved the early cognitive and emotional development of children, in particular for educationally disadvantaged children.
Identification and Estimation of a Dynamic Multi-Object Auction Model (C5, C7)
In this paper I develop an empirical model of bidding and entry behaviour in repeated simultaneous first-price auctions. The model is motivated by the fact that auctions rarely take place in isolation; they are often repeated over time, and multiple heterogeneous lots are regularly auctioned simultaneously. Incorrect modelling of bidders as myopic or as having additive preferences over lots can lead to inaccurate counterfactuals and welfare conclusions. However, such a complex model poses a computational problem for estimation, predominantly due to the high-dimensionality of multi-object auctions.
I prove non-parametric identication of primitives in this model, and introduce a computationally feasible procedure to estimate this type of game. The procedure makes use of an intermediate step in which the econometrician estimates the bidder’s value function, essentially estimating the game as though it were static. The model primitives are then `backed out’ from the results of this intermediate step. I then use a simulation study to consider the properties of the proposed estimation procedure; I find that the procedure performs at least as well as standard approaches to estimating dynamic games, but takes only a small fraction of the time.
Identifying Always-the-Same-Rating Reviewers in a One-Sided-Review System Using Big Data Analytics (M3, D8)
In one-sided review systems (e.g., Amazon), product reviews can be written without providing personal information. Consequently, consumers, sellers, and firms have no direct way to evaluate the credibility of the reviewers, limiting the utility of unfiltered Big Data from one-sided review systems. It is therefore imperative to identify and remove potential misleading reviews before estimating and predicting product quality from online reviews. This study identifies ‘Always-the-Same-Rating’ reviewers (ASRs), reviewers that always give the same star rating for all reviewed products. ASRs can generate a self-selection bias, lessening the informativeness of average measures of product quality (e.g., average star ratings). This study identifies ASRs in twenty-nine product categories by analyzing individual 230 million reviews for 15 million products written by 102 million reviewers on Amazon. In detail, ASRs can be divided into two subgroups, ‘Always-the-same-rating reviewers in All categories’ (AiAs) and ‘Always-the-same-rating reviewers in a category’ (AiCs). Surprisingly, the vast majority (99.997%) of AiCs are found to be AiAs; 98.203% of AiA reviewers always give a five-star rating for all reviewed products. The digital music category, in particular, shows a relatively high share and volume of ASRs among all categories, making an ideal focal category for further empirical analysis of ASRs. This study empirically demonstrates that star rating, the usefulness of reviews, length of headline and review, and holiday are potential indicators of reviews written by ASRs. In addition, deep learning models effectively identify reviews written by ASRs, and the positive weighted convolutional neural network (CNN) on top of Bidirectional Encoder Representations from Transformers (BERT) embedding shows higher performance than the unweighted one. Combining text and non-textual data shows a higher predictive performance than the text-only case. The approaches developed in this study will be useful for mitigating the effects of potential self-selection bias in online product reviews and promotional reviews.
Identifying the Determinants of Corporate Cash Holdings: The Precautionary Motive Redux (G0, E5)
Building on a rich micro-based dataset covering most European countries over 2000-2017, this paper proposes a simple identification strategy to isolate key determinants of corporate cash holdings. In firms that use cash for precaution, liquid assets fall in rising collateralizable assets because internal and external funds have become substitutes in a low-interest rate environment. These credit-constrained firms tend to be low-productivity firms. In firms that use cash against a strategic/insurance-related motive, the relationship is positive because rising collateral goes hand in hand with higher retained earnings to ring-fence complementary long-term investment. These unconstrained firms tend to be high-productivity firms. Results are in contradiction with standard theories of corporate liquidity demand (Holmstrom and Tirole 1998), where only constrained firms accumulate cash at the time of investment. The conjecture which is brought to the test is that the recent low-interest-rate environment emptied the opportunity cost of holding cash such that now internal and external finance are substitutes in high levels of pledgeable collateral and the precautionary motive for holding cash less compelling.
Identifying the Heterogeneous Impact of Highly Anticipated Events: Evidence from the Tax Cuts and Jobs Act (G1, C3)
We develop a method for estimating the stock market impact of aggregate events. Based
on using data on both stock and options prices, our technique accounts for two important
sources of bias present in traditional methods. First, our method takes into account market
anticipation, without the need for information on specific firm characteristics. Many event
studies only measure a fraction of an event’s full value effect, so the measured market reaction at event resolution can be misleading, particularly in the case of a very high degree of market anticipation. Second, our method is robust to the possibility of the event being good news for some firms and bad for others, without prior specification of this heterogeneity.
We apply the method to the passage of the Tax Cuts and Jobs Act (TCJA), which exhibits both
anticipation and heterogeneity. We estimate the market anticipated the probability of passage
to be as high as 95% 30 days before the event. The full value impact of the TCJA is found to
be 12.36%, compared to 0.68% when market anticipation is ignored. The firm-level impact of
the TCJA is considerably heterogeneous, with large and innovative firms with high growth
prospects being the largest winners.
Identity Display, Group Selection and Cooperation: A Public Goods Experiments under the Chaoshan and Hakka Culture in China (C9, D7)
Culture can deeply affect economic behaviors, economists and sociologyists have expounded the relevant views for a long time. Under the background of high-level globalization, it is important to understand the relationship between cultural differences and social preferences. This paper uses the method of laboratory experiment, and compares the behavioral differences in cooperation between Chaoshan and Hakka culture in Guangdong, China. As a multi-national country, China always holds the cultural concept of "harmony in diversity". In order to promote multi-national integration and common development, this paper contributes to find an effective coordination mechanism of cooperation in the context of cultural diversity, and therefore introduces two mechanisms of identity display and group selection into the game of public goods. The experiment finds that both identity display and group selection have positive and effective effects on promoting cooperation. When social cooperation is carried out within groups with the same cultural identity, it is of great significance to display culture identity in the cooperation. While when group selection is ensured to be realized, adding a segment of group selection may further promote social cooperation. In general, under the diverse cultural background in China, we can achieve a higher level of social cooperation with the help of culture identity. This research helps to improve the system construction, as well as promote social cooperation and common development in multi-cultural areas.
Immigration and the Comparative Advantage of Nations (F2, J6)
How does immigration affect export performance? To answer this question we propose a unified framework allowing to disentangle various mechanisms put forth in previous literature. These include the role of networks in reducing bilateral transaction costs as well as productivity shifts arising from migration-induced knowledge diffusion or from increased workforce diversity. While we find evidence supporting all three channels (at both the intensive and the extensive margins of trade), our framework allows to gauge their relative importance. We then focus on the diversity channel and find stronger results in sectors with more complex production processes and more intensive in teamwork cooperation. This is consistent with theoretical models linking the distribution of skills to the comparative advantage of nations. Overall the results are robust to using a theoretically grounded IV approach combining three variations on the shift share methodology.
Immigration, Migrant-Intensive Jobs, and Sectoral Shift (J6, F1)
We test the hypothesis that a local influx of immigrants crowds in or out employment of native-born workers in more relative to less immigrant-intensive nontradable jobs, but has no such effect across tradable occupations that was found in Burstein et al. (2020) from the US perspective. We also investigate if there is a sectoral shift within the tradable industries due to a change in labour composition in the tradable sector. We use Canadian Census Data from 1981 until 2016 from the Canadina Public Use Micro Files (PUMF).
The preliminary result shows that immigrants do not cause crowding in nor crowding out in either tradable and nontradable industries for highly educated workers while there is an expansion of native employment in both tradable and nontradable industries for low educated workers from 1981 to 2001. For the estimates from 2001 to 2016 there are no statistically significant effects of immigration on native employment in either education class nor tradable and nontradable industries from 2001 to 2016. These results are different compared to the results found from the US perspective.
Impact of COVID-19 on Residential Real Estate (R1, R3)
We study how stay-at-home and shelter-in-place orders, imposed in the U.S. due to the COVID-19 outbreak, affect pricing and time on the market in the residential real estate markets. We exploit county-level variation in the intervention dates to study the effect of these interventions on outcomes (prices, sales, and vacancies) both across and within housing markets. We use the ZTRAX transaction-level data and regression discontinuity design to show that stay-at-home orders reduced house prices by a significant 4 percent within 2 miles of the county border. We control for square footage, age, transaction month and year fixed effects, and boundary fixed effects. We explain it with a search and matching model, in which interventions reduce the frequency of meetings of buyers and sellers during the lockdowns, which reduces prices.
Impact of COVID-19 on Small Businesses in Big Cities- New Evidence from San Francisco (I1, R1)
Covid-19 has been devastating for small businesses across the nation, particularly in dense urban areas. A literature review of immediate impacts of COVID on small businesses in the US revealed that the knowledge generated using nationally representative data does not allow us to adequately capture local impacts in big cities, specially in high cost of living areas that have seen an exodus of residents and office workers due to shelter-in-place. Small businesses that supported this population has seen a drop-in clientele. Apart from public health guidelines, there have been demand-driven reasons for the hugely negative impacts on small businesses. This has had catastrophic impacts on small business owners and their families and has the potential to change the vibrancy of big cities.
To gather data to further inform the extent of impacts and influence relevant policy-making in San Francisco, we used a community-based participatory research model in which university researchers collaborated with the city’s Small Business Commission, supervisors and other business stakeholders to come up with a survey instrument that would adequately capture local struggles, barriers and opportunities. The survey was translated into 8 other languages to capture data from small business owners with attention to equity and social justice issues and was administered between January and February of 2021. Data was collected through merchants associations, commission and office of workforce lists and is a representative 1% sample.
Preliminary results from the survey indicate that the pandemic has had disproportionate impact on small businesses, particularly arts, restaurants, personal services, gyms and salons, which might change the city for a long time to come. This study will help inform policymakers, not only in San Francisco but in similar cities across the nation, potential ways to support and preserve our communities through this crisis.
Impacts of the Covid-19 crisis: Evidence from 2 million UK SMEs (G3, G2)
The Covid-19 pandemic materially reduced UK economic activity and small and medium sized businesses (SMEs) were hardest hit. At the start of 2020 before the pandemic took hold, UK SMEs accounted for around 60% of employment and half of total revenues. In this paper, we use a novel near real-time dataset on the universe of UK SME accounts with major banks to document the impact of the Covid-19 crisis, in terms of turnover and cash flow declines. We use linear regressions and machine learning models to study the path of these variables for SMEs of different sizes, across different regions and sectors of the economy, to document substantial heterogeneity across firms. The dataset comprises monthly information on 2 million SMEs that have current accounts or debt with 9 major banks - roughly 5 billion data points. Our main findings are: the Covid public health interventions coincided with a 30% reduction in turnover year on year for the average SME; there was significant heterogeneity across SMEs, with the biggest reductions for younger SMEs in consumer-facing sectors in Scotland and London; but cash flows did not decline on average and there was much less heterogeneity across SMEs, as a result of substantial government support. Using a probit model, we also document significant heterogeneity, with regards to firm age, turnover and location, in the usage of government guaranteed scheme loans. Our analysis provides a framework to monitor SME developments in the coming months as the corporate sector recovers from the pandemic.
Import Competition, Foreign Inputs, and Employment: Evidence from the Colombian Liberalization (J0, F1)
We study how import competition and foreign inputs coming from high income
countries affect employment in emerging economies. Abundant evidence shows that import competition from low-wage countries decreases manufacturing employment in high-wage countries, but far less is known about the employment effects of foreign inputs, in general, and the effects of import competition on developing economies, in particular. We exploit exogenous tariff reductions
that raised Colombian imports from the United States, and therefore increased
import competition and decreased the prices of foreign inputs. Using highly detailed
administrative data, we build an input-output matrix that links foreign competition and
inputs at the industry level, and analyse changes in employment using household surveys
and matched employer-employee records. We derive four main conclusions that contrast with previous results found in high-wage countries. First, the positive effects of the
reduction in the prices of foreign inputs offset the negative effects of the increase in competition.
Second, decreases in manufacturing employment are explained by substitution with foreign
inputs, rather than by import competition. Third, low and high skilled employment
are equally affected by competition and inputs from the United States. Fourth, isolated
Colombian regions experience proportionally larger manufacturing employment loses due
to increases in the amount of imports from the United States. Our results show that
international trade do not create winners in developing countries and losers in developed
countries, but, instead, creates winners and losers within countries.
Incentive Pay Prior to CEO Turnover When Effort Choices Have Lasting Effects (G3, J3)
We present a modified principal-agent model to identify a link between the anticipated likelihood of future CEO turnover and the optimal sensitivity of incentive pay to firm performance when CEO effort choices have lasting effects on firm performance. In such a model, an increase in the anticipated likelihood of turnover reduces the impact of future performance contracts on current CEO effort and induces a compensatory increase in the optimal sensitivity of current CEO compensation to current firm performance. We test this prediction empirically using a sample of over 3,000 US firms constructed from the ExecuComp and CompuStat databases. Using an executive-specific fixed effects model, we find that two proxy variables for an increase in the anticipated likelihood of turnover do lead to greater sensitivity of current CEO incentive pay to changes in current firm performance. These two variables identify CEOs in the final two years prior to their departure from the position, but only when the departure reflects a planned succession, and when these CEOs have reached retirement age.
Indirect Network Effects and Policy Implication: Empirical Analysis of the Chinese Electric Vehicle Market (H2, D6)
Governments can accelerate technology adoption by directly subsidizing the technology or subsidizing
the adoption of its complements when indirect network effects exist. The optimal policy choices
depend on the scale of the indirect network effects, relative to the direct policy effects. This paper estimates the mutual indirect network effects between electric vehicles (EVs) and charging stations, and assesses the effectiveness of EV purchase subsidies and charger subsidies on EV adoption and their efficiency, applying EV sales and charger number data in China. Although the indirect network effect of chargers on EV adoption is significant, our findings suggest that EV purchase subsidies are 34.4% more effective than charger subsidies of equal-size spending in promoting EV adoption, at a lower cost of efficiency loss. Moreover, these two subsidies are different in their effects on the distributions of EV sales and consumer welfare: the recent changes in EV subsidies are in favor of high-range vehicles and their buyers while charger subsidies are in favor of low-range vehicles and their buyers.
Individual Compliance with Disease-Preventive Behavior: Experimental Evidence (I1, C9)
The rise and spread of COVID-19 has made individual compliance with disease-preventive behaviors such as wearing face coverings, social distancing, staying home when sick, hand washing, and vaccination critically important to reducing the spread and saving lives. From an economic perspective, these disease-preventive behaviors vary in terms of the private and external benefits they provide. Because of this, individual compliance may vary across behaviors, and interventions aimed at increasing compliance may be more successful for some behaviors than others. The experiments in this paper characterize a simplified version of the disease-prevention environment. Individuals face a possible loss, representative of getting sick, and are given the option to pay a small cost to decrease either their own probability of loss, others’ probabilities of loss, or both, depending on the treatment. Additionally, each decision environment is tested in three different frames, positive, negative, and neutral. In the positive frame, the positive externality created by paying the cost is emphasized in environments with an externality. In the negative frame, the negative externality created by not paying the cost is emphasized in environments with an externality. An additional treatment includes a “moral nudge”, in which subjects read the University’s mission statement to “advance the common good” prior to participating in the experiment. Initial results suggest that in baseline treatments, individual behavior is well predicted by the Nash equilibrium, which is sub-optimal for environments with an externality. Individuals do, however, choose to pay the cost more often in treatments with a positive frame on the externality, and even more often in treatments with a negative frame on the externality. Additional sessions are needed to confirm these initial results and to further investigate the “moral nudge”.
Individual Prevention and Organized Screening: A Reflection on Data of the Access of Early Detection of Breast Cancer in Emilia-Romagna, and in Bologna in Particular, After the Reorganization of the Offering (I1, K3)
The data (2002-2016) from the Regional Health Service of Emilia-Romagna on the access to the mammographic services in the Local Health Authorities (AUSLs) of the region (data accessed as generalized civic access) allowed (poster at AEA Meeting 2020) to identify the different choices (YES screening; NO screening) for the early detection of breast cancer undertaken by the women in Bologna and in the other AUSLs in Emilia-Romagna after the solutions adopted after 2010 to deal with the problems of waiting lists and the control of spending for the services of early detection of breast cancer, redirecting the services toward the screening of public health(poster at AEA Meeting 2017).
To better understand what happened in the early detection of breast cancer in Bologna after 2010, women residing in the Bologna AUSL who underwent at least one mammogram in 2010 in spontaneous access or in scheduled screening were taken into consideration and the choices they made in the following years up to 2016, the last available year, were analyzed (paper at WEAI meeting March 2021).
Thus in 2016 it appears that 58,1% of women who had had one or more mammograms through spontaneous access in 2010 progressively did not have any more mammograms (at least within the National Health Service), neither in spontaneous access nor in scheduled screening.
Now in this new paper we examine in depth the different behavioral patterns of women with respect to early detection of breast cancer in Emilia-Romagna, and in particular in Bologna, and we study their different reactions to changes in the offering. We, therefore, want to compare our results with the new debate taking place in Italy (National Screening Observatory, Italian Group for Mammographic Screening, Surveillance PASSI of the Istituto Superiore di Sanità) on the early detection of breast cancer and on the organized screening.
Inflation Gap Persistence, Indeterminacy, and Monetary Policy (E3, E5)
Empirical studies have documented that the persistence of the gap between inflation and its trend declined after the Volcker disinflation. Previous research into the source of the decline has offered competing views while sidestepping the possibility of equilibrium indeterminacy. This paper examines the source by estimating a medium-scale DSGE model using a Bayesian method that allows for indeterminacy. The estimated model shows that the Fed's change from a passive to an active policy response to the inflation gap or a decrease in firms' probability of price change can fully account for the decline in inflation gap persistence by ruling out indeterminacy that induces persistent dynamics of the economy.
Inflationary Household Uncertainty Shocks (E3, D8)
I construct a novel measure of household uncertainty based on survey data for European countries. I show that household uncertainty shocks are not universally like negative demand shocks. Notably, household uncertainty shocks are largely inflationary in Europe. These results lend support to a pricing bias mechanism as an important transmission channel. A comparison of results across countries suggest that demographics and factors related to average markups along with monetary policy play a role in the transmission of household uncertainty to inflation. I develop an Overlapping Generations New Keynesian model with age-dependent Deep Habits to rationalize these results.
Information Acquisition and Price Setting under Uncertainty: New Survey Evidence (E3, E0)
What makes prices sticky? While it is commonly understood that prices adjust only sluggishly to changes in economic conditions, the source of sluggish price adjustment is under-explored empirically. In this paper, we argue that sluggish information updating drives price stickiness. To this end, we use a panel dataset that contains information on both firm-level expectations and price adjustments and document the following facts: (1) there is a positive correlation between whether a firm updates its expectations and whether it adjusts prices; (2) firms update expectations more frequently and make less correlated forecast errors in downturns; (3) firms adjust prices more frequently in downturns which include both upward and downward adjustments. We then extend a Ss price-setting model with second moment shocks to allow for endogenous information acquisition by the firm. The model predicts that firms acquire information more intensively during periods of high volatility, also adjusting expectations and prices more often. It is countercyclical volatility, interacted with menu costs and information rigidity, that drives our results. This implies that the flexibility of the aggregate price level is counter-cyclical, making monetary policy less effective in recessions.
Information Acquisition and the Pre-Announcement Drift (G1, E4)
We present a dynamic Grossman-Stiglitz model with endogenous information acquisition to explain the pre-FOMC announcement drift. Because FOMC announcements reveal substantial information about the economy, investors' incentives to acquire information are particularly strong in days ahead of the announcements. Information acquisition partially resolves the uncertainty for uninformed traders, and under generalized risk sensitive preferences (Ai and Bansal (2018)), lowers the discount rate and results in a stock market run-up. Because our theory does not rely on the leakage of information, it can simultaneously explain the low realized volatility during the pre-FOMC announcement period and the lack of a positive correlation between pre- and post-announcement returns.
Information Driven Volatility (G1, G0)
Modern asset pricing theory predicts an unambiguously positive relation between volatility and expected returns. Empirically, however, realized volatility often predicts expected returns with a negative sign, as exemplified by the volatility-managed portfolios of Moreira and Muir (2017). We show that information-driven volatility induces negatively correlation between past realized volatility and future volatility and future expected returns. We develop a simple asset pricing model based on this intuition and demonstrate that our model can account for several volatility-related asset pricing puzzles such as the return on volatility managed portfolios, the “variance risk premium” return predictability (Bollerslev et al. (2009)), and the predictability of returns by implied volatility reduction on macroeconomic announcement days.
Information Integration, Coordination Failures, and Quality of Prescribing (H5, H7)
Organizations aim to improve the coordination of interdependent decisions to achieve more desirable outcomes. The difficulty for improving coordination is that information is incomplete and dispersed among decision-makers. Health care is a prominent example: a patient’s care delivery is spread across multiple physicians, and each physician has different knowledge of the patient’s health and medical history.
We analyze a public policy of health information integration in Finland. The country was one of the first ones to adopt a nationwide system for electronic prescribing (e-prescribing). Our identification approach is based on the staggered adoption of e-prescribing across all the municipalities between 2010 and 2014. Compared to individual providers’ incompatible and incomplete information systems, e-prescribing systems provide more comprehensive information on prescriptions across different physicians involved in a patient’s care. The adoption of e-prescribing serves as a plausibly exogenous shock to the information sets of physicians, being directly relevant to their prescribing decisions and coordination.
Using our prescription-level data, we find that the adoption of e-prescribing has no statistically significant effect on the overall probability of co-prescribing harmful drug combinations. We also evaluate regional heterogeneity in the effects because there is considerable evidence of an urban-rural gap in health care provision and outcomes. Similar to the average effect, there is no statistically significant effect on the probability of harmful co-prescribing in urban regions. However, in rural regions, the measure of low-quality prescribing reduces substantially, by approximately 35 percent. The improvement in the quality of prescribing in rural regions is driven by unspecialized physicians (generalists) and by interacting prescriptions from different physicians, rather than from the same physician. Despite the underwhelming results on average, our results for rural regions still support the view that information integration has the potential to improve coordination and mitigate the harms of fragmentation in health care.
Information Pools and Insider Trading: A Snapshot of America’s Financial Elite. (G2, G3)
We document abnormal correlations between the performance of hedge funds’ managers with an elite socio-economic background. In particular, Columbia, Harvard, University of Pennsylvania, Stanford, and NYU alumni are highly correlated among themselves. We take steps toward linking this phenomenon to a shared information pool with a quasi-natural experiment: the 2009 Galleon Capital insider trading scandal. The difference-in-difference analysis shows a significant reduction in returns of the elite managers following the scandal. Finally, we present evidences suggesting that investors recognize this pool’s value, as funds with access to elite information are associated with 55% higher assets under management at launch.
Institutional Minority Shareholders and Investment in Privately Held Firms (G3)
Institutional investors, who are traditionally public market players, have increasingly invested in private firms in the past decade. Exploiting comprehensive ownership and financial data on privately held firms in the United Kingdom, we study the impact of institutional investors on the investment behaviour of private family firms. We find that private family firms with institutional minority shareholders exhibit higher investment levels and are more responsive to changes in investment opportunities than their peers wholly owned by families. The positive impact of institutional investors on the investment behaviour of private family firms is concentrated in investment in intangible assets rather than physical assets. Supporting the view that institutional shareholders promote investment by alleviating financial constraints of private family firms, we further document that the effect is more pronounced in firms that are more dependent on external finance, and firms with institutional minority shareholders are less financially constrained. Overall, we provide novel evidence on one type of benefit from non-traditional investors’ investment in private firms, namely alleviating financial constraints, which facilitate higher investment and investment efficiency, especially in intangible assets.
Inter-firm Patent Litigation and Innovation Competition (G3, K4)
Using novel inter-firm patent litigation data, we show a significant interplay between intellectual property rights' boundaries and product market dynamics. Instrumenting a firm's patent litigation propensity with the passage of China's National Intellectual Property Strategy reform, we find that patent litigation reduces defendant firms' innovation activity and fosters more exploitative innovations. The effects strengthen with product market overlap between litigants. We further find that patent litigation intensifies product market competition among close rivals and results in lower and more disperse innovation activities within industry, implying an industry structure where Schumpeterian effect of competition is more likely.
Interest Rates and Asset Prices under Financial Liberalization (F3, F4)
This paper constructs a general equilibrium model and studies how the exchange rate regime, the capital account liberalization, and the financial development influence the equilibrium interest rates and asset prices as the financial markets integrate globally and deepen with larger borrowing capacities. With floating exchange rates, a more liberalized capital account leads to a higher real interest rate in equilibrium, whereas financial development lowers the equilibrium real interest rate. When the exchange rates are fixed, both financial development and capital account liberalization increase the equilibrium real interest rate, and unlike the interest rate levels under the flexible exchange rate regime, the equilibrium real interest rates are negative.
To capture the non-stationarity of the asset prices and the exchange rate, this paper assumes that the asset prices and the exchange rate follow the geometric Brownian motion processes and computes the stochastic steady state numerically. Through the comparative static analysis, this paper finds that the financial development and the capital account liberalization reduce the volatilities of domestic equity price and domestic bond price. The domestic currency appreciates at the stochastic steady state with a deeper financial market and a more liberalized capital account.
This paper also provides implications for the monetary policy when facing the volatile asset prices and exchange rates under the financial globalization. It finds that the design of monetary policy is closely associated with accurate estimations of the financial volatilities. Proper reactions to the risks make central banks less constrained by the Mundell-Fleming trilemma and leave space for the policy rate adjustments under both exchange rate regimes. In particular, monetary policy should be countercyclical to the volatilities of domestic equity and bond prices, and pro-cyclical to the volatilities of foreign bond price and the exchange rate.
Intergenerational Mobility in Switzerland: Evidence from Large Administrative Datasets (H0, J0)
Using large administrative datasets from Switzerland, we study the intergenerational transmission of economic status along four dimensions: income, wealth, education, and occupation status. Thanks to the linkage of i) full population census data with ii) social security earnings records dating back to 1981, iii) individual income and wealth tax data, and iv) large population surveys, we obtain precise measures of our outcomes around age 30 for all the cohorts born 1967--1982 and their parents.
Our results show that income mobility is particularly high in international comparison, exceeding the rates observed in the U.S. or even in Sweden---while at the same time educational and occupational mobility are low. However, we find that over time absolute upward income mobility has declined for cohorts born after 1975.
We further shed light on employment probabilities of women and mothers and the prevalence of entrepreneurship, given parental working histories. Mother’s occupation and education are less correlated with the respective child outcomes than father’s. This suggests that family culture matters for intergenerational transmission of economic status. Even though mothers typically spend more time with their children than fathers, their background has less of an impact on children’s economic status later in life.
We also study heterogeneity across subgroups, such as immigrants, and in regions within Switzerland. As we observe language spoken at home, we exploit the fact that Switzerland has three different language regions to understand how potential language barriers in school or local labor markets affect mobility patterns.
Internal Migration and House Prices in Australia (R1, R3)
Australia is one of the most mobile countries in the world through internal migration, which is an overlooked part of population change. This paper provides an exciting episode that has not been examined previously and explores whether, and to what extent, internal migration affects house prices across Australia. In this regard, we use the Statistical Areas Level 3 (SA3) disaggregated data set that represents the regional breakdown of Australia to study the impact of internal migration inflows on local housing prices. To address the potential endogeneity problem due to simultaneous causality between migration flow and house price changes, we confirm our results with two-stage least squares (2SLS) by a manually constructed instrumental variable that matches the shift-share instrument used in the immigration literature. We find a strong evidence that there is a local economic impact of internal migration in Australian cities; internal migration pushes up the demand for housing in migration-receiving areas and results in house price increases. According to 2SLS models, internal migration that amounts to 1% of the initial local area population is associated with point estimates of 0.62% to 0.81% increase in house prices. Put differently, as of 2019, an annual increase in the stock of migrants equal to 1% of an SA3 region's initial population leads to $4,290 to $5,605 annual increase in house prices. We have also suggestive evidence that the estimates of the OLS specification are downward biased indicating that migrants tend to move towards regional cities and/or towns in which house prices are more affordable, conditional on the local controls and the time effects. Considering that house price changes are an essential source of human capital accumulation and local economic development, internal migration and its influence on housing prices play a crucial role in fostering the sustainable development specifically in regional Australia.
International Reserves, Debt Currency Composition & Sovereign Default (E6)
We study the joint decision of accumulating international reserves and issuing debt in both foreign and local currency using a quantitative sovereign default model. In this environment the government lacks commitment to repay its debt obligations, issues long-term debt denominated in both foreign and local currency and accumulates international reserves. In addition, the government can inflate away its local currency debt using discretionary depreciation. However, inflation is costly and the government will balance this trade-off. The addition of the simultaneous decision of accumulating reserves and issuing local & foreign currency debt allows us to: i) extend the canonical sovereign default model which thus far only considered two out of three of our assets, ii) shed light on the sustainability benefits of optimal currency depreciation and iii) reconcile the canonical quantitative default model with the empirical observation that emerging economies’ governments have increased both their reserves and their local currency debt levels in the last decade. Furthermore, our analysis can also shed light on the optimal government response to a shock that hit the sovereign during exceptional financial stress episodes such as the global financial crisis or the COVID-19 pandemic.
International Spillovers of New Monetary Policy (E5, E6)
We study international spillovers of conventional and new monetary policies of a large
economy to a small open economy (SOE). Building on Sims and Wu (2020), we employ a medium-scale New Keynesian model that features all the major types of new monetary policies and the conventional monetary
policy in a unified framework. We extend their model to an open economy setting. We use our model as a measurement device to quantify the spillovers and study the economic mechanisms behind them. In our quantitative application, Canada is the SOE and the US is the large economy. Our results show that there is little difference in the spillover effects of conventional and new monetary policies on the GDP of the SOE. However, the effects on various components of GDP (consumption, investment and net exports) differ by policy.
We also simulate counterfactual monetary policy scenarios for the US and Canada around the Great Recession of 2008. Three main conclusions emerge from these simulations: (1) If the Fed had not engaged in quantitative easing (QE), the US recession in the wake of the 2008 financial crisis would have been deeper but Canada would have had better economic outcomes; (2) there are diminishing returns to QE in terms of its effects on both the US and Canadian real variables; and (3) had the Bank of Canada followed the Fed and engaged in QE of its own during the Great Recession, the real economic outcomes would have been better for Canada.
Intertemporal Relationships between Criminality and Victimhood: Evidence from Population-Level Panel Data (K4, C3)
In the economics of crime, it is a well-established stylized fact that those who commit crimes are more likely to be victims of crime. Similarly, victims of crime are more likely to be criminals. We explore the simultaneous nature of this relationship using a census of all police investigations in New Zealand between 2014 and 2020. We first revisit this hypothesis by pooling data over time and using recursive bivariate Probit methods, finding evidence of a weak, but fully simultaneous relationship between criminality and victimhood. We next explore this hypothesis by examining the intertemporal relationships between criminality and victimhood using a monthly panel of individuals chosen randomly from the New Zealand population. Panel fixed effects models reveal that previous victimization (offending) is only positively linked to current offending (victimization) in the months occurring immediately before victimization (offending). This suggests that the overlap between victimhood and criminality is driven primarily by criminal incidents occurring close together in time. The detailed nature of New Zealand Police records allows us to further explore intertemporal relationships by incident type, including violent crimes, property crimes, intimate partner violence, and offenses involving weapons.
Investment Risk-taking and Benefit Adequacy under Automatic Balancing Mechanism in Public Pension System (H5, G1)
The global trend of aging populations is a cause for concern regarding the financial conditions of public pensions. Policymakers worldwide are contemplating the introduction of policies to increase the sustainability of public pensions. The Japanese government has repeatedly revised the public pension system to increase the chances of achieving financial stability. The new systems introduced by the government include a fixed contribution rate and an automatic balancing mechanism, which reduce the benefit level according to the long-term demographic and financial conditions of the system. The public pension in Japan has a large pension reserve fund for investing in stocks and bonds worldwide. However, the risk-taking of the reserve fund along with the adequacy of the benefit levels has not been intensively discussed in the literature and as a practice. Therefore, by using prediction methods for the financial verification of public pensions developed by the government, and by applying stochastic simulation methodologies, we investigate the relation between investment risk-taking on the reserve fund and future benefit levels. This study contributes to the literature by focusing on the effects of investment risk-taking on the reserve fund in terms of the sustainability of the public pension system, particularly considering several potential pension system revisions under current policy discussion, which have not been addressed by previous studies. We find that the distribution of benefit levels shifts downward when the reserve fund invests more in bonds than in the term of the current allocation levels. Furthermore, while the benefit levels increase on average, they are faced with greater downward risk when the reserve fund invests in more stocks. These results indicate that lowering investment risk does not correspond to a low risk of benefit levels because investment returns may be unsatisfactory and the automatic balancing mechanism continues for a longer duration.
Is Gender Destiny? Gender Bias and Intergenerational Educational Mobility in India (I2, J6)
Many recent studies provide evidence of gender bias against girls in India, for example, in health, education expenditure, breast feeding, and sex selection. In contrast, the gender gap in schooling has narrowed substantially over the decades. Does gender convergence in schooling attainment imply that the girls in the younger generation in India enjoy equal educational opportunities as the boys?
To analyze this question, we study intergenerational schooling persistence addressing both empirical and theoretical challenges. We incorporate gender bias against girls in the family, school and labor market in a Becker-Tomes model and derive mobility and investment equations that can be taken to data. Parents may underestimate a girl's ability, expect lower returns, and have “pure son preference”. The model delivers the widely used linear conditional expectation function (CEF) for mobility under constant returns but generates strong predictions: parental bias cannot cause gender gap in relative mobility. With diminishing returns, the CEF is concave, and parental bias affects both relative and absolute mobility.
Since coresidency causes severe underestimation of the gender gap, we use data from India Human Development Survey that includes nonresident children and parents. Evidence rejects the linear mobility CEF in favor of a concave relation (both rural and urban). The daughters of uneducated fathers face lower relative and absolute mobility irrespective of rural/urban location. We find gender equality in absolute mobility for the children of college educated fathers in urban areas, but not in villages. Theoretical insights help understand the mechanisms, suggesting underestimation of academic ability and unfavorable school environment for girls. Rural parents exhibit pure son preference. Differences in the incidence of unwanted girls and the impact of parental nonfinancial inputs explain the rural-urban differences. The standard linear model misses important heterogeneity and yields misleading conclusions such as no son preference in rural India.
Is Grass Greener in the Gray Zone? Innovation and Entrepreneurship in the Cannabis Market (O3, I1)
50 countries and 34 US states have legalized some types of cannabis by 2019. Despite recognition of cannabis’ medical benefit as non-addictive pain medicine, there are heated debates on regulations regarding cannabis. Most studies focus on demand-side health outcomes or product availability but with limited understanding of the upstream, supply-side innovation and entrepreneurship in this market. This paper combines data from legislative documents, clinical trials, Crunchbase, patents, and trademarks to construct multi-dimensional supply-side outcomes in the global cannabis market. We collect detailed cannabis-related policies cross-country and cross-US state, as well as hurdles in tax, financial, legal, branding, and licensing for cannabis businesses. Our global sample covers about 50 countries; our US sample includes all US states during 2000-2018.
We use difference-in-difference models utilizing variation in medical and recreational cannabis legalization, controlling for confounding factors. In the US sample, we find a mild increase in clinical trial initiation following medical cannabis legalization, which quickly faded and re-ignited with an almost one-fold increase once recreational cannabis dispensaries are operational. Startup formation responds positively to medical cannabis laws, but the trend reverses when recreational cannabis laws are operational. Patents and trademark applications increase following medical cannabis legalization, with sizable increases in cannabis and increased obfuscation in hemp-derived CBD - consistent with branding strategies in this still quasi-illegal market. Results are stronger in the global sample – more precision and larger in magnitudes. The strong global increase in business formation, patenting and trademark filing suggests cross-country arbitrage opportunities and the value of intangible assets.
Ongoing work uses Bacon-decomposition to disentangle policy. Furthermore, we use textual analysis and manual curation to categorize business activities (e.g., types of products vs. services) and scientific complexity. We combine quantitative analysis with topic modeling on 300+ episodes from a leading cannabis business podcast to disentangle the mechanisms.
Is This Time Different for Monetary Policy? (E5, E3)
The improving health picture, alongside the robust response from policymakers, has materially changed expectations on the outlook. With the stage set for a rapid economic rebound, analysts are wondering if this time is different for monetary policy. Particularly, is this the end of the declining Fed Funds Rate (FFR) trend? That is, in each business cycle over the past 30 years, every peak in the FFR is lower than the past cycle. In addition, the FOMC offered larger incentives and for a longer duration in a recession relative to the past cycle. Therefore, each recession drained the FOMC’s resources and left the Committee with “less ammunition” to fight the next recession.
Does the current cycle indicate a break in the past declining FFR trend? Does the FOMC deviate from the past tradition of offering larger incentives for a longer duration?
This study develops a new framework to estimate the likely path of the FFR, particularly, the likely peak value for the FFR in the near future. Using the historical relationship between the FFR and the 10-year Treasury yields, our framework suggests that the FFR may follow the past declining trend in the near future. That is, the near-term peak in the FFR may be lower than the past cycle’s peak value.
Another application of our framework suggests that the estimated duration of monetary stimulus for the current cycle may be shorter than the past cycle. However, the estimated magnitude of the current monetary stimulus has already crossed the previous cycle’s bar.
Therefore, our work suggests, in terms of the past declining FFR trend and size of the monetary incentives, this time may not be different for monetary policy. Duration of the offered incentives may be different (shorter) for the present cycle.
Is Women’s Competitiveness Expressed through Their Husband’s Income? (J0, D1)
We test for the influence of heterosexual individual’s competitiveness on their partner’s income using a recently validated measure of competitiveness incorporated in 2017 within a large representative sample survey with income data from 2015-2021. First, we show that in aggregate, the past (before 2017) and future (after 2017) income levels of men and women increase with their own competitiveness when we do not control for contemporaneous (2017) income. When we control for contemporaneous income to eliminate the potential influence of past success on surveyed competitiveness, we find that only the future income of single men increases on own competitiveness, but not that of cohabiting men or women regardless of marital status. Remarkably, the change in coupled men’s future income increases only on their partner’s competitiveness, not their own. By contrast, the change in women’s future income level decreases on their partner’s competitiveness. We investigate the potential channels for our results with a reduced sample of years (2015-2019) for which we have work hours data. Men’s own work hours do increase on men’s own competitiveness. However, the longer hours apparently do not increase their income. Inconsistent with standard models of household specialization contributing to men’s labor market productivity, men’s work hours do not increase on women’s competitiveness. These findings suggest that women’s competitiveness increases their partner’s income by increasing their aggressiveness in wage bargaining.
Job Mobility Within and Across Occupations (J6, J3)
This paper assesses the impact of occupational mobility on life cycle wage inequality. I develop a model of job mobility which attributes differential returns to occupations to occupationally heterogeneous labor market frictions, compensating differentials, and non-pecuniary job switching costs. I estimate the structural model on linked Hungarian administrative data and use it to quantify the relative importance of each of these mechanisms. High-skill occupations offer higher wages and more stable employment; in turn, low-skill occupations feature higher non-wage amenities but larger non-pecuniary costs of switching to high-skill jobs. As a result, workers who start their careers in the bottom 10 percent of the wage distribution in a high-skill occupation surpass those who start in the top 5 percent of a low-skill occupation in 5 years. I find that occupationally heterogeneous labor market frictions are the key drivers of these ex ante wage profiles. These results indicate that occupational heterogeneity in the sources of wage inequality is instrumental to fully account for life cycle wage dynamics.
Just Do IT? An Assessment of Inflation Targeting in a Global Comparative Case Study (E5, E3)
The literature on the effects of inflation targeting (IT) remains open to question. Putting aside panel regression analyses, the literature has mostly used difference-in-differences estimators to evaluate the causal effects of IT. This paper addresses the concerns with those techniques using synthetic control methods for causal inference in a (large) sample of advanced economies (AEs) and emerging market and developing economies (EMDEs). First, we find that IT was relatively effective in reducing the inflation rate in advanced economies that pursued lower inflation, with falls lower than 1 percent over the first five years of the post-intervention period. Among EMDEs, the average reduction in inflation fluctuates around 2 percent per year. That said, these gains are statistically significant only in a few AEs (Canada, UK) and EMDEs (Colombia, Philippines, Poland, and South Africa). Interestingly, IT contributed to fight deflation in Japan and raise its inflation rate compared with its estimated counterfactual. Second, IT countries were able to cushion the external shocks related to the high commodity prices and the global financial crisis. During the 2007-09 period, the gains in lower inflation are almost negligible in AEs. In contrast, EMDEs achieved inflation rates about 3 percent lower than the average rate of the comparison group. Third, we find that the IT effectiveness---measured by the dynamic treatment effect---is statistically associated with proxies of the degree of central bank independence. Among AEs, policy formulation attributions seem to play an important role. In contrast, the central bank's ability on lending to the public sector appears as a statistically relevant covariate in EMDEs.
Kid-Cession: The Impact of the COVID-19 Pandemic on Parents’ Labor Force Participation and Welfare (J1, G5)
Prior work has shown that the COVID-19 pandemic has especially harmed the labor market outcomes of women with children. A likely explanation is the closure of schools and childcare resources due to the pandemic, along with differentials in time spent on child care between men and women. In this paper, we use administrative banking data and geographic data on school closures to calculate several measures of the disproportionate impact of the pandemic on women with children. First, we document differential trends in labor income, bank balances, and spending across households’ family structure (e.g., children vs. no children), gender, and whether schools were in person or virtual in the fall of 2020. Second, we use industry-specific unemployment rates—based on UI and PUA receipt into a customer’s checking account—to detect involuntary job separations (i.e., layoffs) vs. voluntary separations (i.e., quitting to take care of children). We then document resulting effects on household consumption caused by voluntary separations.
Land Prices and the Persistent Effects of Wealth Inequality (O4, E2)
We show how, in theory, wealth inequality can have long-run effects on an economy’s output. We develop and solve an endogenous growth model in which land enters both the preferences of households (via housing) and the production function for research and development.
When preferences over consumption and housing are non-homothetic, the cross-sectional distribution of wealth affects the fraction of national income devoted to housing relative to all other goods, reducing the demand for new products. This affects the profits accruing to producers of new varieties, which affects the incentives for firms to innovate, and so permanently affects output. The model therefore presents a simple mechanism linking together features of the United States economy observed over the last four decades: declining economic growth and firm entry, rising inequality, and the increasing role of housing in national income.
Land-Use Regulation and Economic Development: Evidence from the Farmland Red Line Policy in China (R1, R5)
Many countries have land-use regulations to preserve farmland from urban sprawl, such as greenbelts, urban growth boundary, and agricultural zoning. In this paper, I show that such regulations can distort economic activity across sectors and locations at a substantial cost to aggregate welfare in developing countries during urbanization. I study a major policy restricting farm-to-urban land conversion in China - the Farmland Red Line Policy - to provide causal evidence on the impact of land-use regulation on local development measured by GDP and population growth. The policy imposes an additional cost on urban land development, which depends on exogenous local geographical features. I show that a greater additional cost driven by the geographical features significantly reduces urban land supply, lowers GDP, and decreases population. To understand the aggregate impact of the policy, I develop a quantitative spatial equilibrium model that features endogenous land-use decisions. According to the model, the policy causes an excess supply of farmland and an under-supply of urban land, and the extent of such land misallocation varies across locations due to their local geographical features. In the constrained equilibrium, the spatial and sectoral mobility of workers implies that land misallocation leads to labor misallocation. The calibrated model reveals that the welfare of workers would have been 6% higher in 2010 if the policy had not been implemented. Moreover, a cap-and-trade system that achieved the same aggregate level of farmland would have been far less costly in terms of welfare. The results suggest that fast-growing economies in developing countries need to design land-use policies carefully, as the welfare costs of poorly designed policies can be substantial.
Latinx High School Mathematics Achievement in Gentrifying School Zones (I2, H4)
This project investigates high school mathematics achievement of Latinx students in schools that are located in gentrifying areas. As the rate of gentrification has increased in many inner cities that have traditionally been the home of Black and Latinx people across the country, it is important to investigate the impact that these demographic shifts have on student outcomes. As Latinos become a larger portion of the student population in K-12 public schools, it is important that we increase the enrollment of these groups not only to ensure that they are doing well economically but the country as well.
School Attendance Boundary data are merged onto the HSLS:09 data. HSLS:09 followed a nationally representative cohort of children from ninth grade through three years after their expected high school graduation year. This study focuses on students who are present in the first three waves of the HSLS data set who attended public high school and student high school transcript data was available for all four years. These selection criteria result in 10,810 students representing 767 high schools across the United States.
Multilevel models with random intercepts are used to estimate the relationship between gentrification, school diversity, and Latinx mathematics academic achievement and account for the nested nature of the data. This study shows that in the majority of gentrified secondary public schools became more diverse which attributed to the lower mathematics achievement for Latinx students. This study provides evidence that gentrification may be negatively impacting students who are not displaced because of educational redlining in the short term. This study not only adds to the current literature, but lays the groundwork for future research looking at the impact of gentrification on public secondary schools across the country.
Learning and the Anatomy of the Profitability Premium (G1, E2)
I introduce imperfect information and learning for unobservable long-run productivity into a dynamic asset pricing model and provide an explanation for the profitability premium. Firms with high profitability have greater information precision and face greater exposure to updated long-run productivity shocks through the learning mechanism. Deviating from the existing models without learning, my framework provides a unified explanation for a wide set of empirical facts: firms with high cash-based operating profitability (1) have higher information precision and capital allocation efficiency; (2) are more exposed to aggregate productivity shocks and, hence, earn higher expected returns; and (3) exhibit shorter cash-flow duration.
Leasing as a Mitigation Channel of Capital Misallocation (G1, E0)
Leased capital accounts for about 20\% of the total productive physical assets used by U.S. public listed firms, and this proportion is even higher among small and financially constrained firms - over 40\%. In this paper, we argue that leasing is an important alternative way of capital reallocation, complementary to directly purchasing capital from the reallocation market, and it significantly mitigates the credit-constraint-induced capital misallocation. However, in the existing literature, leased capital is an ``unmeasured'' capital in quantifying capital misallocation. Empirically, we show that neglecting leased capital and overlooking its mitigation effect lead to significant overestimations of both the capital misallocation (Hsieh and Klenow, 2009) and the cyclicality of capital reallocation (Eisfeldt and Rampini, 2006). Theoretically, we develop a general equilibrium model with heterogeneous firms, collateral constraint and an explicit buy versus lease decision to demonstrate this novel mechanism: the possibility for firms to rent capital when they are financially constrained mitigates capital misallocation.
Leasing as a Risk-Sharing Mechanism (G1, E2)
This paper argues leasing is a risk-sharing mechanism: risk-tolerant lessors (capital owners) provide insurance to financially constrained risk-averse lessees (capital borrowers) against systematic capital price fluctuations. We provide strong empirical evidence to support this novel risk premium channel. Among financially constrained stocks, firms with a high leased capital ratio earn average returns 7.35% lower than firms with a low leased capital ratio, which we call it the negative leased capital premium. We develop a general equilibrium model with heterogeneous firms and financial frictions to quantify this channel. Our study also provides a caveat to the recent leasing accounting change of IFRS 16: lease induced liability and financial debt should not be treated equally on firms' balance sheet, as their implications for firms' equity risks and cost of equity are opposite.
Leveraging on Human Capital: Labor Rigidities and Sorting over the Business Cycle (E2, E3)
This paper analyzes the scarring effects on workers and hysteresis effects on the whole real economy of business cycle fluctuations. We introduce a structural model of the labor market that features worker and rm heterogeneity, whereby workers accumulate human capital and search on the job. Wages are set through an optimal dynamic contract, with downward wage rigidity arising endogenously through limited commitment on the firm side. In this setting aggregate fluctuations alter the sorting between workers and firms and distort incentives to accumulate human capital. We show that contractual rigidities, together with limits to the intensity of investment in human capital, generate long term costs of business cycle fluctuations. Scarring effects for workers arise in absence of demand externalities or informational frictions, as a direct result of physical constraints to investment and limited commitment by workers. Once inefficiently separated, workers that look for employment in bad times direct their search towards less productive firms, a fact which has long lasting consequences for their working career. The consequence is an ensuing hysteresis in firms' productivity distribution, which tends to persist long after aggregate productivity reverted back to trend. Using administrative data on the universe of Italian labor contracts provided by the social security administration (INPS), we provide empirical evidence of these mechanisms.
Leveraging the ‘Greenback’: How Macro-Financial Policy Can Stimulate Green R&D for the Net Zero Target (O3, E6)
Whether it is the ECB president or a FED governor highlighting the importance of central banks in actively participating to mitigate climate change, or the increasing evidence of how climate change has an impact on macro-financial aggregates (e.g. natural rate, risk premium, among other indicators), macroprudential policy and monetary policy could help smooth the economic transition and achieve the sought after Paris Agreement net-zero CO2 emissions target. In this paper, we explore empirical evidence on the role that green innovation plays in successfully reducing emission intensity (emissions per output) using time series data from the European Trading Scheme carbon market and EU data on green patents. We find that research and development (R&D) plays a significant role in achieving the economic de-coupling of emissions to output (i.e. reducing emission intensity). To investigate the role financial policy could play in stimulating green R&D, we build a general equilibrium model based on previous work that explores the role of banking loans in boosting R&D, on one hand, and work that examines the role of subsidies for facilitating green innovation on the other. Our economy includes both a banking system and endogenous green innovation, which allows for increasingly efficient abatement. From a long-term growth perspective, we find that efficient abatement technology (i.e. greener technologies) would help achieve CO2 emissions reduction targets. However, the net-zero Paris Agreement target requires increasingly higher levels of abatement technologies. We also find that green innovation when coupled with a carbon tax facilitates the attainment of the net-zero target while overtime helping to notably decrease the carbon price. Finally, from a business cycle perspective, we compare macroprudential policy to fiscal subsidy. We show that a weighted macroprudential policy incentivizing banks to loan more to green innovators is complementary to subsidizing green R&D and could achieve more efficient outcomes.
Liquidity and Monetary Transmission: A Quasi-Experimental Approach (E5, G2)
In the face of lower real interest rates, central bank balance sheets are likely to remain larger relative to pre-crisis levels, resulting in greater banking system liquidity. However, there is little evidence on the impact of higher liquidity on credit supply and the monetary transmission mechanism in the ‘new normal'. We exploit a novel dataset on bank liquidity positions arising from a unique regulatory regime and combine it with a highly-detailed, loan-level administrative dataset on UK mortgages. Using the design of quantitative easing auctions as an instrument for liquidity to address endogeneity, we find that more liquid banks charge slightly higher mortgage interest rates, and pass on significantly less changes in risk-free rates. We explain this through bank behaviour that attempts to preserve net interest margins in the face of holding low-yielding liquidity. Consistent with this, we find excess liquidity leads to reaching-for-yield responses in banks' mortgage risk-taking. Additionally, the results shed light on the optimal mix between (un)conventional monetary policy tools. Policies that boost bank net interest margins are more likely to help the transmission of risk-free rates to lending rates.
Liquidity Coinsurance in Syndicated Loan Markets (G2)
We study the capacity of the banking system to provide liquidity to the corporate sector
in times of stress and how changes in this capacity affect corporate liquidity management. We
show that the contractual arrangements among banks in loan syndicates co-insure liquidity risks
of credit line drawdowns and generate a network of interbank exposures. We develop a simple
model and simulate the liquidity and insurance capacity of the banking network. We find that
the liquidity capacity of large banks has significantly increased following the introduction of liquidity regulation, and that the liquidity co-insurance function in loan syndicates is economically
important. We also find that borrowers with higher reliance on credit lines in their liquidity
management have become more likely to obtain credit lines from syndicates with higher liquidity. The assortative matching on liquidity characteristics has strengthened the role of banks as
liquidity providers to the corporate sector.
Liquidity Traps in a Monetary Union (F3, F4)
The closed economy macro literature has shown that a liquidity trap can result from the self-fulfilling expectation that future inflation and output will be low (Benhabib et al. (2001)). This paper investigates expectations-driven liquidity traps in a two-country New Keynesian model of a monetary union. In the model here, a rise in government purchases in an individual country has a weak effect on GDP in the rest of the union. The results here cast doubt on the view that, in the current era of ultra-low interest rates, a rise in fiscal spending by Euro Area (EA) core countries would significantly boost GDP in the EA periphery (e.g. Blanchard et al. (2016)).
The paper is available here:
Author web page: http://www.robertkollmann.com
Lone Stars or Constellations? The Impact of Performance Pay on the Distribution of Academics (M5, O3)
The effects of performance pay on effort and sorting have been studied extensively, but the effect performance pay may have on workforce composition in general, and the degree of matching assortativeness in particular, is much less understood. Yet if complementarities in worker skill exist, we would expect performance pay to increase positive matching assortativeness and this may greatly affect total productivity, particularly in industries with strong complementarities, such as academia (Kremer (1993), Legros and Newman (2002)). This paper studies the effect of performance-related pay on the distribution of academics across universities in a hedonic coalition formation model and provides empirical evidence that performance pay increases the clustering of similarly productive academics, giving rise to a less homogeneous distribution of academic stars across universities. To do so, the paper exploits the introduction of performance pay in German academia as a natural experiment and employs a newly constructed data set that encompasses the affiliations and publication records of the universe of academics in the country. Using plausibly exogenous variation in the productivity of new hires by exploiting mandated professorial retirement rules as an instrument, I find evidence of positive spillovers or complementarities between co-located academics. These are larger in fields with larger co-author teams and when relatively more productive academics are co-located with relatively less productive academics. The latter suggests the academic production function is submodular. I then test whether the introduction of performance-related pay increases clustering of similarly productive academics in a difference-in-differences framework, using the strength of complementarities in academic fields as continuous measure of treatment intensity. I find that performance pay increases positive matching assortativeness two- to three-fold. With a submodular production function, however, such an increased clustering of similarly skilled workers is not optimal, and a more equal distribution of academic stars could increase total research output.
Longevity versus Income Inequality: What Shapes the U.S. Wealth Distribution? (E1, D3)
We develop a rich general equilibrium overlapping generations model to quantify the role of trends in labor income inequality, longevity, capital income inequality and taxation on the changes in the distribution of wealth in the U.S. Unlike earlier research, which emphasizes the role of labor income and capital income shocks, we show that the key drivers of changes in the wealth distribution lie in the demographic processes. We decompose changes in wealth inequality to within (birth) cohort and between (birth) cohort components and show that the between component, driven predominantly by increased life span after retirement is a key driver of increased wealth inequality in the US. Our setup permits counterfactual simulations with progressive income and wealth taxes and shows that the trend in longevity cannot be feasibly attenuated with taxation.
Lying Aversion and Vague Communication: An Experimental Study (C9, D9)
An agent may benefit from misleading the audience's belief about the state of the world. While a more blatantly misleading message may be more effective than a vaguely misleading message, however, they may affect one's internal cost of dishonesty and social identity of honesty differently. Thus, a sophisticated agent must balance the degree of truthfulness and vagueness of the message.
We explore the extent to which these two types of lying costs affect people's sophisticated use of vague messages in communication and vice versa in a simple experimental setting. To this end, we introduce a novel experiment design that isolates the internal cost of lying and the social identity cost of appearing dishonest. In particular, the design varies the relevance of the social identity concern across anonymous and non-anonymous treatments. Our setup extends the framework of Fischbacher and Föllmi-heusi (2013) by implementing a richer message space with vagueness.
The experimental data shows that subjects employ more vague messages in treatments in which the social identity concern is relevant. In treatments in which the social identity concern is irrelevant, on the other hand, we find that most subjects exploit vagueness so as to be consistent with the truth, yet at the same time leveraging the imprecision to their own benefit in a more undisguised manner. We also find a smaller but non-trivial-sized group of truth-tellers who do not take advantage of vague messages even against potential opportunities for monetary gain. The result opens a new set of questions on the motivations behind the preferences for truth-telling.
Macroeconomic Dynamics of the Russian Federation: Econometric Model - 2020 (E1, C3)
The article presents the results of the new version of the author's econometric model of the Russian economy. This version consists of 24 equations and 66 identities that describe the relationships between 90 variables (14 exogenous and 76 endogenous).
The new version implemented the following innovations: a) the share of cash in the money mass and mandatory reserve ratio were included; b) the parameters were re-estimated due to new quarterly data from 1999 to 2019 (by OLS and ML – ARCH); c) the specification of the equations was improved due to new estimates.
The elasticities in equations and impulse multipliers estimated show the continuing stagnation of total factor productivity, the strong dependence of the Russian economy on demographic and international factors, and low efficiency of fiscal and monetary policies.
The forecast for 2020 - 2023 obtained on the basis of the model in the base case shows the GDP growth rate equal to -0.5% per year, while inflation will be 3-4% per year. Nevertheless, a noticeable acceleration of growth rates is possible at high (3% per year) growth rates of the world economy.
These results do not take into account the impact of the COVID-19 pandemic (which will change these results for the worse in both the global economy and the Russian economy). But they can provide a useful picture of the development of the Russian economy after the consequences of the pandemic is overcome.
Macroeconomic Effects of the COVID-19 Public Health Crisis (C6, E6)
This paper proposes a dynamic cascade model to investigate the systemic risk posed by sector-level disruption in input. We then use this model to study the effect of the challenge presented by COVID-19 on the U.S. economy. We construct a weighted digraph G = (V; E; W) using the industry-by-industry total requirements table for 2018, provided by the Bureau of Economic Analysis. In this graph, the nodes Vi represent sector level industries, an edge Eij represents a commodity flow from industry i to industry j, and a weight Wij captures this commodity's value. We impose an initial shock that disrupts one or more industries' production capacity, and we calculate the propagation of production shortage with a modified Cobb-Douglas production function. The initial shock is modeled based on the spike in unemployment between March and April 2020, as reported by the Bureau of Labor Statistics. The industries within the network are assigned a resilience r that determines an industry's ability to absorb input losses. If the input loss rate exceeds the resilience r, the industry fails, and its outputs go to zero. We observe a critical resilience rc, such that below this critical value, the network experiences a catastrophic cascade resulting in total network collapse.
Macroeconomic Statistics based on Surveys:Circumventing Subjectivity in Housing Sales and Rent Data (E5, G1)
The current market value of the residential housing stock constitutes a crucial input variable in macro- and microeconomics and the social sciences in general as well as for macro-economic monitoring and financial supervision. Market values are, however, only observable upon selling a property or establishing a new rent contract. As a substitute, owner-estimated and renter-reported values collected in country-representative surveys are regularly used as if they were true market prices. This article makes use of the Luxembourg Household Finance and Consumption Survey (HFCS), which we amended by questions about important price-determining housing characteristics for all households' main residences. We link each HFCS observation to a large collection of market sales and rent data via hedonic models thus yielding imputed current market prices and rents for each dwelling described in the survey. Thus, we add objective values to otherwise subjectively reported survey data. We characterise survey participants who tend to under- or over-report, respectively and find a strong correlations with tenure length, tenure type (renting versus owner-occupying), type of dwelling and household income and wealth. Such correlations question the appropriateness of relying on self-reported housing market and housing wealth data as they translate into mismeasured macro-economic indicators.
Macroprudential Policies in a New Keynesian Model with Investment Funds (G2, E4)
The relevance of the investment fund sector in the euro area is increasing and the real
economy is becoming progressively more reliant on fund intermediation. We find
that a one percent outflow from euro area bond funds reduces industrial production
by 0.3% after five months. This paper builds a New Keynesian dynamic stochastic
general equilibrium model with banks and funds that allocate credit to non-financial
corporates. Banks grant loans and issue safe deposits, while funds hold corporate
bonds and cash, and issue redeemable shares that expose them to redemption risk.
In the market solution, funds hold too little cash and, when hit by a redemption
shock, are forced to liquidate corporate bonds. This brings about large swings
in credit intermediation, output and consumption. We show that counter-cyclical
regulatory minimum cash buffers mitigate the effects of adverse financial shocks to
the macro economy.
Macroprudential Policy and Credit Spreads (E4, E3)
After the 2008 global financial crisis, there is a focus on macroprudential and banking policies to create a sound financial system in which financial problems are not spilled over to the real economy. Macroprudential authorities need to use indicators to assess the sustainability of credit growth and the level of system-wide risk and take the right policy-making decisions. We propose a countercyclical macroprudential rule, which responds not only to the credit growth but also to credit spreads. First, we empirically test the validity of this additional variable by providing evidence on the correlation of credit spreads with credit booms. Then, we explicitly introduce this variable into a Dynamic General Equilibrium (DSGE) model. We use our model to determine to which extent, having macroprudential measures responding to credit spreads may be welfare improving. Our results show the optimal weight that the rule should attach to this indicator so that we can give relevant policy recommendations to this respect.
Making News Salient (G0)
CEOs have incentives to communicate with their investors after news releases if the market misinterprets the news. I examine how CEOs communicate with the market through their trading patterns. I find that CEOs are more likely to purchase shares after positive and negative news releases, suggesting that they want to confirm their positive news if the market underreacts to it and want to mitigate the market overreaction to their negative news by purchasing shares. These patterns vary conditional on the information environment, institutional ownership, and news categories. My results suggest that CEOs can make the news salient via their trading pattern.
Managerial Attention, Employee Attrition, and Productivity: Evidence from a Field Experiment (M5, J2)
What is the causal impact of managerial attention on employee attrition, productivity, and well-being? How should firms strategically allocate managerial attention among workers? We formulate a theory that illustrates how different attention allocation strategies influence workers’ updated beliefs about the manager’s type, and, in turn, employee performance. To test the theoretical predictions of the model, we conduct a 6-month randomized control trial at a leading multi-national spa chain with 157 stores and more than 10,000 workers in China. In the experiment, managers are given a weekly list of employees with whom they are required to have a standardized, private conversation. We compare the random allocation method, where attention allocation is uncorrelated with any employee characteristics, to the directed allocation method, where managers focus on employees with more negative emotions and therefore higher attrition probabilities. We document significant causal effects of managerial attention on employee attrition and well-being. Consistent with the theory, we find that random allocation of managerial attention is more effective than directed allocation in reducing turnover.
Managing COVID Policy Uncertainty: A Behavioural Monetary Policy Model (E7, E6)
The COVID pandemic has triggered unprecedented macroeconomic shocks. Increasing fiscal deficits and large fluctuations in cash hoarding, alongside long-term trends towards use of electronic transactions and cybercurrencies, have magnified policy uncertainty. In COVID's aftermath, large deficits, low interest rates, and cost-push inflationary pressures limit central banks' ability to control monetary policy settings and the monetary transmission mechanism. In developing new macroeconomic policy insights, this paper builds on Baddeley (2021) in reconciling behavioral and rational expectations approaches to time inconsistency (Strotz 1955; Kydland and Prescott 1977, Barro and Gordon 1983; Laibson 1997; Angeletos et al. 2001; O’Donoghue and Rabin 2015; Angeletos and Huo 2021) applied to rules, discretion and reputation in monetary policy when policy-makers' discount functions are quasi-hyperbolic versus exponential. Theoretically, increases in policy uncertainty triggered by COVID shocks blunt monetary policy tools, undermine co-ordination of fiscal and monetary policy and limit central banks’ ability to target inflation, with the weakening of monetary policy increasing in policy-makers' propensity for present bias. These theoretical hypotheses are examined econometrically using monetary policy uncertainty indexes, e.g. see Husted et al (2017). The paper concludes with a discussion of macroeconomic policy implications for the COVID age.
Angeletos, Huo (2021). Myopia and Anchoring, AER 111(4):1166–1200
Angeletos et al (2001). The Hyperbolic Consumption Model, JEP 15(3):47-68.
Baddeley (2021). ‘Behavioral Monetary Policy: New Perspectives on Time Inconsistency, submitted to JPolEc, draft https://arxiv.org/abs/1907.07858.
Barro, Gordon (1983). Rules, discretion and reputation in a model of monetary policy, JMonetEc 12(1):101-121.
Husted et al (2017). Monetary Policy Uncertainty. FRB International Finance Discussion Papers 1215.
Kydland, Prescott (1977). Rules Rather than Discretion: The Inconsistency of Optimal Plans, JPolEc 85(3):473-492.
Laibson (1997). Golden eggs and hyperbolic discounting, QJE 112(2):443-477.
O’Donoghue, Rabin (2015). Present Bias: Lessons Learned and to be Learned, AER 105(2):273-279.
Strotz (1955). Myopia and Inconsistency in Dynamic Utility Maximization, RevEconStud
Mandating Public Annuity Purchase and Banning Gender-Based Pricing May Unintentionally Lead to Advantageous Selection (H5, G5)
A well-known solution to adverse selection in insurance markets is to mandate that
everyone buy insurance. This paper revisits this solution when gender-based pricing is
banned in a mandatory public annuity program with partial waiver, both of which are
observed in some countries. We consider a simple model with these two policy features
and the assumptions of positive health-wealth correlation and gender gaps in health and
wealth. We introduce a measure of the severity of adverse selection and decompose this
measure into the within-group and between-group effects when the gender-neutral
pricing is adopted. A surprising result is that the severity of adverse selection may be
zero and may even be negative (meaning that advantageous selection is present) if the
between-group effect is stronger. Our analysis suggests that advantageous selection
may arise from the interaction of gender-neutral pricing and the exemption clause of
the mandatory public annuity program. This provides an alternative mechanism to the
idea emphasized in models with multidimensional private information.
Marginal Tax Changes with Risky Investment (E6, H2)
Using an estimated life-cycle model, we quantify the role of heterogeneity in wealth returns for the response of income to marginal tax changes. In our economy, agents who are sufficiently productive can obtain higher returns by choosing to be entrepreneurs. Return heterogeneity amplifies the responsiveness of total income to marginal tax changes along the entire income distribution with the top 1 percent displaying the highest elasticities. Return heterogeneity increases the incentives to invest for the richest, high-return entrepreneurs, thus amplifying their income responses to marginal tax changes. This reallocation of capital increases aggregate productivity, generating a larger boost in equilibrium wages. This in turn strengthens the income response of the bottom 90 percent, but nevertheless, their response is smaller than at the top.
Mark My Words: The Transmission of Central Bank Communication to the General Public via the Print Media (E5, C8)
Central banks need to influence wage and price-setters' expectations to fulfil their objectives. Despite its importance, communication to the general public is far less studied than communication to financial markets. This paper posits that a key channel through which the general public receives central bank communication is through the print media. We examine which features of central bank text are associated with increased newspaper reporting of central bank communication. We write down a model of news production and consumption in which news generation is endogenous. We use our model to show that standard econometric techniques will likely (i) provide biased estimates and (ii) fail to deal with the high-dimensionality of the estimation problem. We use computational linguistics to measure the the extent of news coverage a central bank communication receives, and the textual features that might cause a communication to be more (or less) likely to be considered newsworthy. We apply our model to the case of the Bank of England, and utilise machine learning techniques designed for high-dimensional equations to estimate the relationship between news coverage and central bank communication. We find that the interaction between the state of the economy and the way in which the Bank of England writes its communication is important for determining news coverage. Content and the state of the economy on their own do not seem to have an effect on news coverage. We provide concrete suggestions for ways in which central bank communication can increase its news coverage by improving readability in line with our results.
Market Power in Product and Labor Markets and Average Stock Returns (G1, E1)
In this paper, we analyze the asset pricing implications of market power in product and labor markets. To investigate the effect of these changes in labor and product market power on firms' asset prices, we first construct a real business cycle (RBC) model where firms have oligopoly and oligopsony power in product and labor markets, respectively, and show that the presence of market power in either market is associated with a lower equity premium. We then provide empirical support for the above results using univariate and multivariate portfolio analysis along with firm-level panel regressions. Our results suggest that investors demand a premium for holding stocks that are in low labor market concentration industries. In addition, we show a negative and significant relationship between labor market concentration and future returns even after controlling for other systematic factors such as market risk, size, profitability and momentum.
Market-Based Private Safe Asset Creation (G2)
I model a shadow banking sector that originates risky loans that are used as collateral for creating private safe assets. Asset managers offer investment vehicles financed by different liabilities and maximize safe debt capacity by committing to contingent collateral trades. Due to a pecuniary externality that arises from this commitment, individually optimal balance sheet choices generate an inefficient market structure. As a result, the sector has too many bank-like intermediaries but under-produces safe assets. Using collateralized loan obligations as a laboratory, I show evidence from the post-crises era, including the COVID-19 period that is consistent with the theory’s predictions. The model provides a welfare-based analysis for the US Risk Retention regulation in this market.
Marriage and Development: Cross-Country Evidence (O4, J1)
Marriage is an important economic subject because it matters not only for labor supply decisions but also for fertility and child-raising decisions. In this paper, we draw microdata from 92 countries to document that marriage rates decrease with development: one log point increase in GDP per capita is associated with a decrease of 5 percentage points in marriage rates for adults above age 22 for both genders. Over the life cycle from age 15 to 89, marriage rates for males are always higher in low- than in middle-, and high-income countries. In low-income countries, the male marriage rate is 73% on average across all 5-year age bins, which is 7 and 13 percentage points higher than that in middle- and high-income countries, respectively. For females, the average marriage rate in low-income countries is 8 and 17 percentage points higher than in middle- and high-income countries before age 50, although it becomes lower than that in high-income countries during age 50 and 75. Furthermore, we show that variations in desired fertility and demographic compositions including distributions of age, education, and urban status can account for at most one third of the decreasing marriage rates with GDP per capita. Finally, we propose a framework to quantitatively decompose the contributions from reduced gender wage gap and changing social norms with development.
Marriage Market Signaling and Women’s Occupation Choice (J1, J3)
It has been long puzzling the economists why women have been underrepresented in some occupations while overrepresented in others, typically low-paying jobs. Despite the gender equality movements and the decline in occupation segregation by sex over the last 50 years, this segregation is still far from disappearing. On explaining this robust pattern, previous literature basically falls into three broad categories: preference, discrimination and social norms, and human capital. In my paper, however, I tried to provide explanation of women’s occupation sorting from the aspect of marriage market signaling. By establishing a model with women born with heterogenous family-orientation and men born with heterogenous preference for wife, I propose that on a marriage market with imperfect information, if occupation choice could signal women’s family-orientation, women then have the incentive to sort into “childcaring” occupations, which typically are not well-paid. I further use data from American Time-use Survey, O*NET, and National Longitudinal Survey of Youth to empirically prove this.
Mass Shootings and Infant Health in the United States (I1, I3)
We study the causal effect of mass shooting incidents during pregnancy on infant health outcomes. Our identification strategy exploits the spatial and temporal variation of mass shooting incidents across counties of the United States. We find that increased severity of mass shooting incidents leads to lower average birth weight, increased incidence of low birth weight (less than 2500 grams), and higher infant mortality rates. Further, we use the exogenous variation in the media coverage of mass shooting incidents due to competing international newsworthy events to show that the effects are exacerbated by the coverage of mass shooting events. Our findings suggest that the adverse effects on health might be due to psychological stress from exposure to news coverage of shooting incidents.
Measures of Model Risk for Continuous-time Finance Models (C5, G1)
Measuring model risk is required by regulators on financial and insurance markets. We separate model risk into parameter estimation risk and model specification risk, and we propose expected shortfall type model risk measures applied to L´evy jump models and affine jump-diffusion models. We investigate the impact of parameter estimation risk and model specification risk on the models’ ability to capture the joint dynamics of stock and option prices. We estimate the parameters using Markov chain Monte Carlo techniques, under the risk-neutral probability measure and the real-world probability measure jointly. We find strong evidence supporting modeling of price jumps.
Measuring Interdependence of Inflation Uncertainty (E3, C1)
The unprecedented fiscal and monetary policy responses during the COVID-19 crisis has increased uncertainty about inflation. During crises periods, the strength of the transmission of inflation uncertainty shocks from one country to another tends to intensify. This paper examines empirical methodologies to measure the strength of the interdependence of inflation uncertainty between the UK and the euro area. We first estimate inflation uncertainty by ex post forecast errors from a bivariate VAR GARCH model. The interdependence of uncertainty is estimated using a probability model. The results imply that the spillover of uncertainty is stronger for uncertainty about distant future than near future. The evidence from quantile regressions shows that such empirical method could suffer from bias if endogeneity is not properly addressed. To identify structural parameters in an endogeneity representation of interdependence, we exploit heteroskedasticity in the data across different regimes determined by the ratio of variances. The results no longer exhibit stronger interdependence at longer horizons. Estimated by different sample periods, the strength of the propagation of inflation uncertainty intensifies during the Global Financial Crisis while the interdependence significantly weakens during the post-crisis period.
Measuring Systemic Financial Stress and Its Risk for Growth (G0, E3)
This paper proposes a general statistical framework for systemic financial stress indexes rooted in standard definitions of systemic risk. We consider systemic stress as an ex post measure of systemic risk. Our statistical framework defines systemic stress as a state of the financial system in which a representative set of individual stress measures is considered extremely high and strongly co-dependent. The composite indicator results from a matrix association index that combines two matrices quantifying the extremeness and the co-dependence hypotheses. We demonstrate that several indicators from the financial stress and systemic risk literatures can be represented as special cases of our general framework. The paper also introduces a daily enhanced variant of the ECB’s composite indicator of systemic stress (CISS) for the US and the euro area. The CISS aggregates index components using their time-varying cross-correlations as co-dependence measures, thereby putting more weight on situations in which stress becomes widespread and thus systemic. From a statistical point of view, the various design steps are geared towards delivering a composite indicator which does not suffer from look-ahead bias, is sufficiently robust to outliers and largely unaffected by differing distributional properties of the underlying raw data. We develop a bootstrap algorithm to test critical levels of the CISS. The final part runs a quantile vector autoregression model for the CISS, the PMI and annual real GDP growth. The results confirm the CISS as a significant driver of economic activity, but mainly in the lower tails of the growth distributions in line with the recent growth-at-risk literature. Conditional forecast exercises and variance decompositions suggest a dominant role of financial stress in explaining the severe recession during the financial crisis in 2008/9. This is different from the Covid-19 crisis in which financial stress shocks only play a minor role compared to aggregate output shocks.
Measuring the Natural Rate of Real Interest for Chinese Economy (E4, E5)
In this paper, we deliver several measures of the natural rate of real interest for Chinese economy and assess their reliabilities as the indicators of Chinese monetary policy conducts. First, we use the simple univariate methods including the historical average level and filtering techniques to quantify the natural real rate. Second, a VAR/VECM model based on Taylor rule is built up to calculate the natural real interest rate. Third, also as a robustness check, we quantify the natural real rate using a New Keynesian system and then evaluate all the estimates. The empirical results suggest that the natural real rate varies over time with a countercyclical pattern, and the estimate in terms of the Taylor rule is most efficient and successful for indicating the stance of monetary policy in China. Our study provides new insights to investigate the monetary policy operation and its effects on the business cycle in China.
Medicaid Expansion's Effect on Self-Employment Reporting (H2, J4)
This paper tests whether the expansion of Medicaid following the Affordable Care Act impacted the supply of labor toward work which does not offer employer supplied health insurance. I find a reduction in engagement in self-employment in states which expanded Medicaid, with roughly 350,000 to 700,000 fewer individuals reporting earnings from nonemployer establishments among states which expanded Medicaid. This represents approximately an 18.9% - 37.7% reduction in comparison to the expected number of self-employed. Using data on the deployment of Uber, I test whether this effect is driven by an employment lock effect, or if instead it is evidence of tax evasion through a reduction in declared self-employment income. I find evidence that the reduction in declared self-employment is biased by tax evasion. This result supports previous work on reported earnings from self-employment being manipulated in reference to means-tested programs (Andreoni, Erard and Feinstein, 1998; Saez, 2010; Chetty et al., 2012;Chetty, Friedman and Saez, 2013).
Medical Loss Ratio Regulation and Insurer Pricing (I1, L5)
The Affordable Care Act Medical Loss Ratio (MLR) regulation limits each insurers' profit by setting a minimum requirement on the ratio of medical spending to premium revenue. This regulation may undermine the incentives for insurers to bargain for lower prices when negotiating with health care providers. I build a bargaining model of how MLR constraint affects price negotiation between insurers and providers. This model illustrates the insurer trade-off between lower premiums and higher service prices and reveals how bargaining for lower prices is reduced. Predictions from the model are tested in a structural model of MLR regulation on negotiated prices and insurers' costs using data from the individual Health Insurance Exchange Marketplace. Welfare calculations using estimated demand, cost, and bargaining parameters suggest that, with the presence of insurer-provider price negotiation, the MLR regulation led to higher health service prices and consumer welfare loss. The counterfactual analysis suggests that either a health service price regulation or a well-designed public insurance option could make health services and health insurance plans more affordable and further improve consumer welfare.
Migration Workers and Their Left-Behind Children’s Mental Health (I1, J1)
Following massive flows of internal migration of rural population in China, a large quantity of literature has emerged to explore the income gains and the subsequent welfare improvement. However, only a few studies examine the problems related to left-behind children’s mental health (or, noncognitive ability development). In this paper, we develop a theoretical framework a la Cunha, Heckman and Schennach (2010) and analyze the effect of parents’ migration on children’s mental health, using the China Family Panel Survey (CFPS) 2012-2018 panel dataset. We use factor analysis to measure the noncognitive ability of children of age between 10 and 15 in three dimensions: depression, self-esteem, and self-control. Our OLS regression analysis shows that, on average, left-behind children are more depressed (by an increase of 0.0472 standard deviations) and less self-esteemed (by a decline of 0.023 standard deviations), while their self-control ability is not significantly different from others. The negative impact on mental health is particularly more profound to young children (10-12 years old). All our results are robust to using IV and DID models. We explain this impact with two possible transmission channels: the company channel and the communication channel. For the first channel, we find that the rise of income in a low- and middle-income family due to parents’ off-farm work compensates for the lack of parents’ company in children’s noncognitive skill development. This leads to an insignificant impact on the left-behind children from low- or middle-income families, while this impact is significant for the left-behind children from high-income families. For the second channel, we find that the parents’ migration results in deterioration in parent-child relationships due to lack of communication, which contributes to a worsening of children’s mental health.
Minority-Owned Startups: Discrimination in Access to Capital and Industry Selection (E2, J7)
We use multiple micro-data sources (Survey of Small Business Finances and Kauffman Firm Survey) to document three key facts regarding the impact of racial inequalities in accessing startup financing on firm outcomes. First, Black-owned firms generally have lower capital-labor ratios than White-owned firms, and they also typically start and stay smaller over time. Second, Black-owned firms have lower average revenue products of capital (ARPK) than White-owned firms. Finally, Black-owned firms have lower average revenue products of labor (ARPL) than White-owned firms. We argue that our findings present a puzzle for the extant literature on racial inequalities in startup financing. Our first fact is consistent with a long literature documenting discrimination against Black entrepreneurs in startup financing. However, viewed through the lens of a standard investment model, our second fact is mutually inconsistent with our first fact. In standard investment models where returns to capital is globally decreasing, financially constrained firms employ lower capital-labor ratios since capital is scarce, but enjoy higher internal returns to capital, that is, higher ARPK. To resolve this puzzle, we propose an alternative model of entrepreneurial choice and production with two key ingredients: (1) an extensive margin of selection where financially constrained entrepreneurs potentially selective negatively into sectors with lower implicit costs of capital, and (2) a production function with locally increasing returns to scale. We use our model to match our three facts above, and then further study the macroeconomic impacts of discrimination in startup financing.
Modelling Yield Curve in Less Liquid Markets (G1, C4)
Less liquid markets for government bonds (LLMs) are characterized by many well recognized challenges which reduce the reliability of the classic Nelson-Siegel-Svensson NSS parsimonious approach. We document key stylized facts about government bond markets concerning liquidity, diversity of maturities available, bid-ask spread in price quotes, as well as price distortion in the very short end of the curve due to switch auctions. Based on these facts, we augment the NSS approach with model- and data-driven endogenous system of weights which permits reliable estimation of yield curves in LLMs. We apply our approach to the data for one of the largest European emerging markets: Poland. Through a battery of sensitivity analyses we show that there exists a class of weights that systematically gives better results than the classic NSS approach. The best fit weights have at least the same weight for the short end of the curve as a sum for all other tenors of bonds. It proves that inferring from the liquidity in particular maturities raises the information content and quality of yield curve estimation, which links our results to the pure expectation hypothesis (PEH).
Monetary Instruments and Inflation in Nigeria: A Revisit of FAVAR (E5, C5)
This paper revisits the potency of the Factor Augment Vector Autoregression (FAVAR) model to trace the impact of notable monetary policy instruments on inflation in Nigeria. Having identified noisy variables with higher uniqueness using the factor analysis approach, we retained the variables with higher communality instead and eliminate others. Our results show no evidence of "price and liquidity puzzles" in our approach thus, the estimated factors improve the inflation forecast for Nigeria. In addition, our results show that there exists a high impact of monetary policy rate on Inflation, and the persistent increase in the unobserved information of short-term interest rates and monetary aggregates serves as an effective monetary channel or transmission mechanism of the Nigerian monetary system. Moreover, we find that inflation is mostly driven by food consumer price indices and a good FAVAR estimation should not only involve increasing the selected factors but also eliminate variables whose variance contributes least to the variances shared by the common factor. Consequently, for Monetary Policy Rate (MPR), Treasury Bill Rate (TBR), and Cash Reserve Ratio (CRR) to be an effective policy instrument, they need to be anchored on efficient operating and monetary targets.
Monetary Policy Interactions: The Policy Rate, Asset Purchases, and Optimal Policy with an Interest Rate Peg (E5, E4)
Is central bank balance sheet expansion inflationary? How does optimal policy affect inflation dynamics? We derive the welfare loss function in a New Keynesian model with a debt-capacity constraint. The term premium enters the welfare loss function. With optimal discretionary interest rate policy, balance sheet expansion drives the economy toward the policy rate's effective lower bound (ELB). At the ELB, balance sheet expansion is inflationary with optimal policy. With endogenous balance sheet policy, the central bank can simultaneously target inflation, the output gap, and the term premium. Endogenous balance sheet policy rules exist that support a permanent policy rate peg. Furthermore, we compute Ramsey optimal balance sheet policy under a permanent policy rate peg. We compare this balance sheet policy to balance sheet policy when the policy rate responds flexibly in combination with the balance sheet to support divine coincidence.
Monetary Policy, Excess Reserves and Credit Supply: Old-Style versus New-Style Central Banking (E5, E4)
The operational framework in the U.S. changed in 2008 as a consequence of the unconventional policies implemented to mitigate the Great Financial Crisis. This change in the operational framework at the Fed can be summarized as a switch from a “corridor” to a “floor” system.
In this paper I examine the effects of this switch on the monetary transmission mechanism. Concretely, I analyze if the bank-lending channel changed due to the alterations in the operational system.
To this end, I develop a regime-switching two-agent New-Keynesian (TANK) model with an interbank market to compare both systems. Monetary policy is implemented via Open-Market Operations and thus, the model is able to capture both dimensions of conventional policy, i.e. the interbank target (or Federal Funds) rate and the central bank's balance sheet size.
I find that under a corridor system real activity declines after a monetary contraction in line with Bernanke and Blinder (1988). However, under a floor system, monetary tightening stimulates credit supply, and hence aggregate demand. The reason for this counterintuitive result is that, without a binding reserve requirement constraint, a fall in reserves reduces a friction introduced by the presence of liquidity management costs. A smaller friction enables banks to obtain more deposits, and thus to increase credit supply.
Additional simulations support the Fed’s decision to keep a floor system, since a run on the non-bank money market participants has no consequences for aggregate demand under the new operational framework. Finally, I assess the Fed’s 2019 plan to increase the level of excess reserves and conclude with a word of caution: under the new system, monetary expansion could affect real economic activity in a contractionary manner.
Monetary-Fiscal Policy Interactions in an Era of Unconventional Policies (E6, E5)
This article assesses whether the introduction of unconventional monetary policy has helped the Bank of Japan respond actively to the inflation rate. Then we address what roles monetary and fiscal policy stances play on the behavior of inflation in Japan. A Markov switching dynamic stochastic general equilibrium model with changes in monetary/fiscal policy interaction is estimated. The special feature is to integrate the shadow interest rate in our model, which can be interpreted as an aggregate that captures the overall effect of unconventional monetary policies as well as conventional monetary policy. Focusing in particular on the prolonged deflation observed in Japan, we find that Japan's policy regime was characterized by the combination of passive monetary and passive fiscal policy. We quantitatively clarify that this combination played a substantial role in propagating negative demand shocks, leading to prolonged deflation. But, after the introduction of unconventional monetary policy, monetary policy then has switched from a passive regime to an active regime. We show that the policy regime change has helped to push up inflation.
More than Just Babies: Paid Family Leave and Caregiver Outcomes (J1, J4)
The large and rapidly growing paid family leave literature primarily focuses on health and labor outcomes for new parents and children, overlooking the 20 percent of claims for caregiving. As the Baby Boomer generation ages, an increasing share of the labor force will be older workers, and many of these older workers (particularly women) will also have caregiving responsibilities for their older parents and in-laws. This project will utilize Health and Retirement Survey (HRS) data and American Time Use Survey (ATUS) data to estimate the impact of state paid family leave programs (primarily California Paid Family Leave implemented in 2004) on employment rates, retirement age, and time use among caregivers. We expect that the availability of paid leave will allow caregivers to take a short leave to care for a family member and return to the labor force rather than exiting entirely to provide the necessary care. As a result, we expect employment rate, retirement age, and time spent on caregiving in states with paid family leave programs to increase relative to states without. This research will help elucidate the burdens aging women face balancing work and caregiving to inform future policy to support this generation.
Move On Up: Does Electrification Reduce Internal Migration? (H4, J6)
This study uses the large scale roll-out of electric transmission infrastructure in Nigeria during the years 2010 to 2015 to quantify the effect of electrification on agricultural productivity and internal migration. I combine household panel data on agricultural production and migration of household members with remote-sensing data on observed cropland area at grid cell level. The estimation strategy relies on spatio-temporal variation in distance to the electric transmission grid. To address endogenous grid location, I create a hypothetical grid based on least construction costs and condition on distance to the closest electric substation. Results show an increase in agricultural production, wages and land inputs following the construction of new transmission lines, while agricultural labor demand does not increase. In addition, transmission lines increase the migration propensity of household members, in particular children aged 13-18. Analysis of origin-destination data using a gravity model reveales that electrification accelerates rural to urban migration. Results are robust to the inclusion of numerous economic and political control variables, the inclusion of changes in road infrastructure and placebo tests using data on planned future grid.
Moved to Poverty? A Legacy of Apartheid in South Africa (I0, N0)
This paper examines the consequences of the homeland system set up by the apartheid government in South Africa. This system forced Black people to live in homelands and as consequence several million individuals were moved to such areas during the 1960s and 1970s, resulting in one of history’s largest segregation policy experiments. We examine how and why relocation to the homelands affected human capital attainment. We exploit the staggered timing of homeland establishment in a cross-cohort identification strategy that compares migrants and homeland natives. Our basic finding is that moving to the homelands during childhood significantly reduces educational attainment. The magnitude of this effect is particularly pronounced for early childhood exposure, consistent with the importance of events and circumstances during critical periods of child development. The policy also had important consequences on individual success in the labor market: exposed cohorts are significantly less likely to work and have lower income in adulthood. Our examination of possible mechanisms suggests an important role for place effects. Moving to the homelands at earlier ages implies greater childhood exposure to poorer neighborhoods and it disproportionally reduces human capital attainment. The findings of this paper illustrate how discriminatory policies against specific ethnic groups can have long-lasting consequences and increase ethnic inequality.
Mutual Risk Sharing and Fintech: The Case of Xiang Hu Bao (O3, G5)
Xiang Hu Bao (XHB), meaning "mutual protection" in Chinese, is a novel online platform operated by Alibaba's Ant Financial to facilitate mutual risk sharing of critical illness exposures. XHB reached nearly 100 million members in less than one year since its launch and so far has offered its members critical illness protections at significantly lower cost than traditional critical illness insurance. There are three major distinctions between XHB and traditional insurance products. First, XHB leverages the tech giant's platform and digital technology to lower enrollment and claim processing costs. Second, different from insurance applying sophisticated actuarial pricing models, XHB collects no premiums ex ante from members, but instead equally allocates indemnities and administrative costs among participants after each claims period. Third, XHB limits coverage amount, often below critical illness insurance products, particularly for older participants. We show this restriction potentially leads to separating equilibrium, a la Rothschild-Stiglitz, where low-risk individuals enroll in XHB while high-risk individuals purchase critical illness insurance. Proprietary data from XHB shows that the incidence rate of the covered illness among XHB members is well below that of comparable critical illness insurance. Our findings further suggest the role of advantageous selection in explaining the cost advantages of the Fintech-based mutual protection programs.
New Economics of Regulation: Financial Stability as a Social Dilemma (H4, G1)
This article is an exploratory essay on financial regulation and stability in the light of the 2007-2008 turmoil. It seeks to bring forth the major characteristics of the regulatory environment that let the crisis dynamics develop in spite of numerous warning signs that became apparent from the years 2005. In this vein, I draw upon the New Economics of Regulation (Laffont, 1994).I then suggest a non-cooperative communication game between an extra-market (a public) regulator and the market players (the regulatees) and study the rules that could lead to a relevant incentives mechanism. These rules should perform in their communication and coordination role by allowing financial institutions to undertake activities that are consistent with systemic stability. Compared with the literature developed on this issue, the article offers an alternative perspective to financial stability. It assumes that systemic stability is a public good that decentralized market mechanisms cannot provide. I borrow from the analysis of Ostrom (1998) on the commons and collective action and argue that financial regulation takes on systemic importance since a smooth functioning of markets requires continuous and sustainable provision of financial stability. The implementation of the model rests on a comparison between two alternative communication processes: a binding mediation-based revelation model and a low-constraints-based cheap talk process. The optimal framework is a function of a given environment. Each alternative reveals to be relevant only under a set of specific conditions. The relevance of a regulatory framework then rests on the consistency of an institutional environment with regard to the institutional characteristics of the economy. A possible implementation of the results can be considered in all areas of public choice that partially involve market mechanisms and that have to face coordination problems in order to respond to societal shocks such as the current Covid pandemic.
New Evidence on Monetary Transmission: Interest Rate versus Inflation Target Shocks (E5, E4)
We present empirical evidence on monetary transmission from estimated New Keynesian and empirical VAR models, that allow for a standard nominal interest rate shock and an inflation target shock. In response to the highly persistent inflation target shock we largely find evidence of a Neo-Fisher effect: the nominal interest rate co-moves positively with inflation and output. In an estimated model version where agents do not have full but only imperfect information about the nature of monetary shocks, Neo-Fisherian effects arise only with a lagged effect and not in the immediate short-run, because, in such case, inflation expectations do not adjust immediately to the target shock. We then use these insights to inform our VAR, adopting a methodology that allows to account for the uncertainty about identifying assumptions with respect to the target shock, i.e., that economic agents might need time to learn whether a monetary shock is temporary or persistent. We again find that nominal interest rates and inflation increase significantly and co-move positively in the short aftermath of the target shock, and output increases immediately.
New Products (O3, L2)
We introduce a new measure of innovation based on important product launches by public firms in the US. Our measure is based on stock-market reactions to media articles – classified by a convolutional neural network approach as referring to new product introductions – and has two distinct advantages. First, it covers the entire spectrum of industries and is not limited to products sold by retail firms. Second, we rely on collective wisdom about product value expressed through financial markets. This lends a forward-looking aspect to our measure, and helps avoid issues associated with valuing new types of output in a changing economy. Using our measure, we derive a few stylized facts. We show that product innovations are highly persistent, both at the firm- and at the industry-level. Firms that launch more new products are larger, and they typically operate in industries that are more competitive. New product introductions correlate with productivity measures at the aggregate level. However, most of these new products are launched in industries that are not among the largest employers; moreover, employment falls further following product launches.
Newborns during the Crisis: Evidence from the 1980s’ Farm Crisis (I1, G5)
This paper overcomes an identification challenge in the literature of wealth-health relationship by proposing a novel research design. By studying an interest rate shock that triggered a severe farm credit crisis in the 1980s, I analyze the health effect of wealth losses due to the crisis on infant birth outcomes. To address the endogeneity concern about wealth losses, I construct an instrument using the interaction of land specialization and starting time of the shock. The variation of land specialization stems from the difference in geo-climatic condition suitable for agricultural production, which is exogeneous to local economic condition. I find that great wealth losses were detrimental to the well-being of newborns. A ten percent increase in net wealth raised the low-birth weight rate by 0.1 percentage points and reduced the actual birth weight by 3 grams. I also present evidence that household expenditure reductions in food and health care might be the main contributor to worse infant health outcomes. Households living in areas having more wealth losses also reduced more food consumption and prenatal care visits.
News and Networks: Using Text Analytics to Assess Bank Networks During COVID-19 Crisis (G2, C0)
We study the 'interconnectedness' of stress-tested banks by exploiting how they are mentioned together in the context of financial news. We start by constructing weekly co-occurrence network matrices following Ronnqvist and Sarlin (2015) text-to-network approach. Using the COVID-19 pandemic as an external shock, we examine how bank networks behave during high stress periods. We find that banks become more interconnected during peaks of COVID-19 induced stress. We put forth a new measure of systemic risk that utilizes text-based eigenvector centrality. This measure provides a more stable ranking system than the traditional SRISK measure during both high and low stress periods.
Nonparametric Analysis of Financial Development and Consumption (G2, E2)
In this paper we analyze the effect of financial development on consumption using nonparametric regression methods for panel data.
In high-income countries, financial development has a positive impact on consumption. However, the responsiveness of consumption to financial development decreases as the level of financial development rises.
In low-income countries, intermediate levels of financial development appear to be associated with lower consumption, while variations in the level of financial development in the tails appear not to affect consumption.
The response of consumption to the remaining regressors has a magnitude and sign along the lines of what previous literature on the subject has found. However, the results suggest that these responses are nonlinear, depending not only on the level of the regressor, but also on the level of financial development.
Not that way! An Exploration of the Social Free-rider Problem as Cause of the Boomerang Effect from Social Norm Information (C9, D8)
Social norms are often used to nudge people towards prosocial behavior, yet occasionally results are unexpected and end in a boomerang effect that can potentially cause unintended harm. While there have been several explanations for why boomerang effects occur, the standard literature does not always provide successful explanations. As these nudges are often used in campaigns of societal importance, it is valuable to understand under what circumstances the boomerang effect will occur. One cause of a boomerang effect that we explore is a social free-rider problem whereby the social norm information acts as a mechanism of updating prior probabilities on the actions of others. In this study, we formally propose a model of the social free-rider problem in the context of the boomerang effect from social norm information. Furthermore, we test our model predictions in a laboratory experiment in the context of a social dilemma environment with two different framing environments that alter the priors of the social norms; in this way we outline the circumstances under which this problem may occur.
On Optimal Currency Areas and Common Cycles: Are the Acceding Countries Ready to Join the Euro? (F3, E3)
The former EU president Jean-Claude Junker has proposed that all countries of the European Union should also adopt the euro as their currency and recent research has shown that countries currently pursuing this goal indeed fulfill the classical Optimal Currency Area (OCA) criterion of positively correlated shocks with the European Monetary Union (EMU). We illustrate, however, that not only the correlation of shocks but also a common impulse response pattern over time is needed for a currency area to be optimal. We test this additional OCA criterion using the concept of a common serial correlation test. The test clearly rejects the notion that the potentially acceding countries share a common cyclical response pattern with the EMU aggregate – except for Sweden. Instead, the business cycles in most of the other countries exhibit only a very weak form of codependence.
On the Macroeconomic Effects of Shadow Banking Development (E3, O4)
We build and estimate a dynamic stochastic general equilibrium model with risky innovation and shadow credits to study the macroeconomic implications of shadow banking (SB), particularly on productivity. Our analysis is motivated by negative relationships between SB development and innovation outcome or total factor productivity (TFP) growth. In our model, information asymmetry associated with technology utilisation leads to an agency problem in which shadow intermediation reduces banks’ incentives to screen project quality. An SB boom crowd-out traditional financial services, decreases innovation quality and technology efficiency, and thereby reduces TFP. In the light of model mechanisms, we analyse cross-country differences and deliver important implications of SB. SB development mainly driven by financial factors (e.g., the US case) leads to significant loss on TFP while that relatively prompted by real-sided factors (e.g., China and the EA cases), could be less harmful.
On the Origin of Cognition: How Childhood Conditions Shape Cognitive Function at Older Age (J1, I1)
The paper examines the long-term impact of early childhood conditions on the cognitive function at older age. Using a unique data from a longitudinal survey of Chinese elderly aged 65 and above, we find the old people who experienced poorer nutrition and medical conditions during their childhood and whose fathers were uneducated tend to have lower cognitive ability. In addition, poor nutrition in childhood and the uneducated father or mother predict significant faster deterioration in cognition at older age. We also find significant gender differences in the association. Girls suffer more in cognitive development in late life due to worse childhood conditions than boys. Further mechanism analyses suggest educational attainment accounts for about 20 to 30 percent of the associations between early life circumstances and both the level and the speed of decline of cognitive function at older age.
Opacity, Signaling, and Bail-ins (G2, E4)
Should banks be transparent in the presence of bail-ins? Banks experiencing losses may bail-in creditors to allocate resources efficiently. However, if banks privately learn the losses, bail-ins may signal the asset quality. When bail-ins do not signal the quality, banks immediately bail-in creditors and sell assets at a pooled price, insuring creditors against asset risks. When bail-ins signal the quality, banks attempt to delay bail-ins to sell assets at a higher price. Such incentives can cause excessive short-term repayments. To avoid costly signaling, banks choose to be transparent or opaque so that they will not know asset qualities privately.
Optimal Sample Allocation in Experimental Studies Probing Mediation Effects (C9, C6)
Experimental studies investigating main treatment effects have been widely used in economics, psychology, education, and other social sciences to evaluate whether a treatment works or not. Meanwhile, mediation analysis can probe the mechanisms through which an intervention impacts an outcome. Experimental studies employing mediation analyses to investigate the mechanisms through which a treatment operates on an outcome often lay the cornerstone for the critical development and iterative improvement of a specific treatment.
A key consideration in designing experimental studies probing main and mediation effects is to determine an efficient sampling plan such that the design will use the minimum resource to achieve an adequate statistical power (e.g., 80%). However, the literature has rarely provided guidance on how to design efficient and well-powered experiments aim to assess both main and mediation effects. In this paper, we developed a flexible optimal design framework for experimental studies investing both main and mediation effects by leveraging and modeling the cost of sampling in a power analysis framework.
In our derivation, we considered multiple inferential tests of mediation (e.g., Sobel test, joint significance test, Monte Carlo confidence interval test) and the inclusion of covariates. The results suggest that the proposed framework can identify efficient sampling plans and is robust to misspecifications of concomitant design parameters (e.g., cost structures of sampling). To facilitate end-user calculations, we have implemented these formulas in the R package XXX.
Outmigration and Tax Rates: Will the Wealthy Leave? (H7, J6)
Are wealthy people likely to move where tax rates are lower? I show that differences in tax rates between two counties influence the migration of the poor and wealthy. I exploit variation in outmigration and real estate tax rates at the county-level from the American Community Survey (ACS) and Zillow's Home Value Index. I find that the wealthy migrate less than the poor and typically move to counties with higher income tax rates and real estate tax rates.
Overborrowing & Shadow Banking (G2, G1)
In this paper I show that the existence of unregulated financial institutions (“shadow banks”) generates overborrowing (relative to the socially optimal level) in competitive equilibrium. This is because borrowers fail to internalize that an additional unit of borrowing, when the financial sector is well functioning, has adverse effects on the credit conditions they will face if a crisis arises. The social planner therefore has a motive to intervene to reduce borrowing in good times so as to limit the severity of a crisis. The discretionary planner’s optimal allocation can be decentralized either using borrower based instruments or via regulation of financial intermediaries. The optimal regulation of commercial banks is pro-cyclical.
Parental Inputs and Child Outcomes (I0, J0)
This paper exploits the introduction of a Swedish “daddy-month” reform in a
difference-in-discontinuities design to study how fathers’ parental leave affects children’s
human capital. Results show that the reform improved average compulsory
school-leaving GPAs of sons of college educated fathers by 0.06 standard deviations,
while GPAs of sons of non-college educated fathers declined by 0.05 standard
deviations. We estimate that the reform increased intergenerational persistence of
human capital by ten percent for boys with no corresponding effects for girls. We
provide suggestive evidence that these results are linked to asymmetric impacts on
family stability and role model effects.
Parents in a Pandemic Labor Market (J1, D0)
Gender gaps in labor market outcomes during the pandemic are largely due to differences across parents: Employment and labor force participation fell much less for fathers as compared to women and non-parent men at the onset of the pandemic; the recovery has been more pronounced for men and women without children, and; the labor force participation rate of mothers has resumed declining following the start of the school year. The latter is partially offset in states with limited school re-openings. Evidence suggests flexibility in setting work schedules offsets some of the adverse impact of the pandemic on mothers’ employment, while the ability to work from home does not. We additionally find that employment and labor force participation fell more for black and Hispanic mothers, and that their recovery took longer than that of white mothers. Preliminary results using a matched sample show that the distinction between part-time and full-time employment prior to the pandemic have an effect on rates of exiting the labor force, with part-time workers leaving at greater rates than full-time workers.
Passive Investing and Price Efficiency (G1, G2)
This paper studies how falling fees for delegated investments affect price efficiency in a theoretical framework, in which the investors’ allocations, management fees, and asset prices are all determined in a general equilibrium. Importantly, investors optimally decide whether to participate in the financial market or simply hold the safe asset, and active managers trade strategically, adjusting the traded quantities according to market liquidity. Perhaps surprisingly, and in contrast to the broad theoretical literature, prices of the index fund become more efficient as passive fees decrease and more investors choose the uninformed index fund. Prices can become more efficient even when the inflow to passive funds comes at the expense of the outflow from active funds, as has been the case more recently since 2007. Combined with the observed downward trend in fees, the finding is consistent with recent empirical evidence that prices, especially those of S&P 500, have become more informative over time.
Passive Investors Should Manage Their Trades Actively (G1, G2)
I find 98% of the U.S. equity index ETFs rebalance their portfolio abruptly, i.e., buy and sell the underlying stocks within one day. As a result, the underlying stock returns become a reversed U-shape when they rebalance to buy, leading to 26 bps execution costs on average. Thus, as uninformed, passive investors, ETFs pay even higher execution costs than often informed, active asset managers. ETFs that track publicly available indices (e.g., S&P, FTSE, MSCI index series) pay 33 bps of execution costs. Yet, those ETFs that track private indexes (which are not available for public subscriptions) pay negligible execution costs. Therefore, pre-announced rebalancing trades (“sunshine trades”) are associated with higher transaction costs, indicating that the ETFs might be front-ran by other market participants. The sub-optimal rebalance strategy costs passive investors 1.2 billion dollars per year.
Patent Boxes and the Success Rate of Applications (H2, O3)
Patent boxes significantly reduce the corporate tax rate applied to income earned from a patent. This incentivizes firms to increase the likelihood of a patent application being granted by creating more novel research and using more successful legal representation when filing the application. Conversely, it supports submitting applications for marginally novel innovations that otherwise would not have been submitted, lowering the probability of success. We use data from applications to the European Patent Office from 1978 to 2019 and find that the introduction of a patent box increases the average success rate of applications from large, corporate innovators by 6.9 percent- age points. This impact only materializes two years after a patent box takes effect, suggesting that improved research effort is the dominant response by firms. Therefore patent boxes may help to increase innovation novelty and improve the overall quality of research.
Pattern Making and Pattern Breaking: Measuring Novelty in Brazilian Economics (O3, Z1)
How do new ideas emerge in academic contexts and what forces determine which ideas get selected and which are forgotten? We analyze more than 1,600 papers presented at the ANPEC Brazilian Economics Meetings from 2013 to 2019 using topic modeling and relative entropy measures. In contrast to simply counting citations or reference combinations, these methods explore the information in the actual texts to detect the rise of new patterns and whether these patterns persist once they have been established. We find that novelty is highly correlated with transience so that most new ideas are quickly forgotten. However, of the ideas that persist those that are more novel have a higher impact. We show that our text-based measure of impact is correlated with subsequent citations. Our results provide a metric to compare the nature of research at the level of Brazilian Economics departments as well as for individual researchers. Finally, we analyze how the selection procedures for the ANPEC meetings affect the incentives for economists to pursue more novel or conventional research.
Paying Interest on Reserves to Promote Liquidity (E5, G2)
Liquidity regulation is the conventional policy deployed to correct insufficient liquid-asset holdings by banks. In a standard banking model in which under laissez faire banks hold excessively illiquid asset porfolios, I find that payment of interest on reserves, by strengthening banks’ incentive to hold liquid assets, equivalently corrects this distortion. In fact, it is the optimal policy intervention, since it does not act as a tax on banks and thus, unlike liquidity regulation, avoids financial disintermediation. In an extension, I show that payment of interest on reserves has a fiscal cost and therefore distortionary taxation implies a lower optimal level of interest on reserves.
People’s Context, Corruption as Experience and Culture, and Tax Morale (H2, D9)
Corruption breaks the fiscal contract between government and taxpayers, so it hurts the future motivation to pay taxes. This research identifies the indirect effect of corruption on individual tax morale. I answer three main research questions in this paper: (1) what happens to an individual’s motivation to pay taxes when he/she experiences a corrupt transaction, (2) what is the spillover effect of corruption on individual tax morale, and (3) how persistent is the effect of corruption on tax morale? By using hierarchical multi-level estimates and exploiting several research designs, this research provides important policy information that corruption significantly hurts tax morale of not only those individuals who experience corruption in person but also a large spillover effect on other people who did not experience corruption. Corruption and tax morale have also a cultural component that is often difficult to separate from the institutional effects. After exploiting the origin country corruption of a large immigrant population in the destination country that allows separating the institutional effect of origin country, this paper finds that the cultural norms of corruption persist on the tax morale of first-generation immigrants although the effect attenuates for the second-generation immigrants. Policies that aim to improve tax compliance with minimal enforcement should address corruption and understand the cultural influences. It is unlikely that elimination of corruption will allow people to adjust their tax morale quickly, instead, the adjustment process can be prolonged to at least one generation. As the corruption experience drives down the individual tax morale, several factors such as fairness in public and welfare goods in their community, ethnic diversity, feeling of guilts, sense of responsibility, crowd-in of public policies, and others play key mechanistic roles.
Perception and Reality. Evaluation of Elite Football Women’s and Men’s Performance (C9, Z2)
There is ample evidence that humans misconceive reality. In some cases, misconceptions can be harmless and do not have to be corrected. In other cases, rigid misconceptions influence the social and economic situation of a sizable share of the population. Research has continuously shown, however, that science can reveal misconceptions. A process that improves the situation of people who suffered from the misconception. Using a field experiment, we analyze how respondents evaluate the performance of men and women. The results show that respondents prefer men when they can see the gender. Respondents are undecided, however, when they cannot see the gender.
Political Action Committee Activity and the Elasticity of Demand (H0, E6)
Considerable attention is paid to the amount of money in politics, including the role of Political Action Committees (PACs). It is generally assumed the money raised is spent to pursue political ends. This makes sense on the face. However, that does not seem to sufficiently explain the variations in activity among political committees. This work examines the amount raised and spent by PACs from different economic sectors. The author will then incorporate the price elasticity of demand as the independent variable to explain PACs' activity better.
The research incorporates the aggregation of PAC dollars across five election cycles, according to public records, and will draw from existing studies of the price elasticity of demand for specific goods. Not surprisingly, many PACs are distributed by sector and good.
This work is influenced by political science research which suggests that corporate donations have minimal influence on policy outcomes. (Hansen, Rocca, & Ortiz) Also, Acemoglu et al. look at the effect of the price elasticity of demand on other political events and argue a “key parameter determining the incentives for war is the elasticity of demand.”
This paper is part of a broader research agenda that looks at the influence the collective demand for goods has on politics.
Political Repression, Media Propaganda, and Nation Building (N4, D7)
In the conquest of China in the mid-17th century, the Manchu-led Qing government oppressed the Han Chinese, the native population of China. Two and a half centuries later, when modern newspaper technology became available, revolutionary propagandists took advantage of a retelling of the political repression and resistance to fan the flames of discontent. Applying machine learning to analyze 0.3 million newspaper article titles, I examine the interaction between the anti-Manchu propaganda and the historical repression and resistance. I find that prefectures with repression and resistance responded more to the anti-Manchu propaganda and produced more revolutionaries. After the revolution, revolutionaries strove to build a modern nation-state by organizing the Kuomintang party, army, and government. The results indicate that propaganda utilizing repression and resistance shaped the political identity and played a pivotal role in the political transformation of modern China.
Polygyny, Inequality, and Social Unrest (J1, D7)
Is polygyny related to social unrest? We propose three theoretical mechanisms related to different dimensions of grievance-inducing and greed-related inequality, which may occur in polygynous societies:
(i) Economic, reproductive and social inequality resulting in relative deprivation among non-elite men (‘vertical inequality’ where polygyny implies a monopolization of women by the elite).
(ii) Inequality within elites when it comes to the distribution of resources and inheritance, both related to the relative position of dependent family members in a clan (‘horizontal inequality’ where the polygynous elite itself may fall victim of instability and internal violence).
(iii) Gender inequality in general, since highly patriarchal structures and inequality between sexes make up the core of polygynous family structures. In order to earn the bride price, women are ‘sold’ into marriages by their families, thereby reinforcing the dominant role of males and foregoing the benefits of more gender-diverse societal and political structures, such as women’s more peaceful strategies of conflict resolution.
Using data for 41 African countries from 1990-2014, we provide evidence for these mechanisms and their relationship to four different types of social unrest. We find that medium levels of polygyny, gender inequality and vertical economic inequality are positively associated with both the intensity and incidence of all four types of social unrest under consideration. For horizontal inequality, the feasibility of mobilization seems to be particularly important. Countries with higher average incomes and high levels of inequality within polygynous families have a higher probability of unrest compared to similar countries without horizontal inequality.
The robustness of our results is tested through alternative and complementary hypotheses such as strategic behavior of elites, population growth that relaxes tight marriage markets, and an excess of male deaths that could possibly be balanced by polygynous marriages (implying possible reverse causation), all of which we are able to reject.
Practice the Purpose Preach: Experimental Evidence on the Effect of Corporate Purpose on Workers’ Willingness to Go the Extra Mile (M0, M5)
Academics and business leaders increasingly call for a (re)definition of corporate purpose beyond profit-maximization to create value by contributing to the welfare of society and planet. In this context, the present paper employs a two-phase natural field experiment to investigate whether, when and how the communication of a corporate purpose influences workers’ willingness to complete unrequired extra work. The main findings show that receiving information about an employer’s corporate purpose causes workers to complete more extra work. Workers whose personal preferences match with the organization’s purpose are most responsive. For those workers the communication of a corporate purpose in combination with an authentic purpose practice further increases their additional work performance, whereas a non-authentic purpose practice backfires. Furthermore, we find evidence that the underlaying mechanism is primarily driven through an increase in workers’ meaning of work. In a broader context, the findings that workers are willing to go the extra mile working for an organization with purpose provide some empirical indications in support of the theoretically proposed business case of purpose.
Preschool and Child Health: Evidence from China’s Subsidized Child Care Program (I2, H5)
Early childhood education programs have been found to be effective for promoting children’s social and cognitive development. However, less research has investigated the health impacts associated with these programs. Using a quasi-experiment of the first universal child care program in China from 2010, this paper aims to identify whether preschool attendance together with nutrition and health services at school produces any short-term effects on health-related outcomes of preschoolers. I exploit the variation in the number of subsidized preschools across provinces and implement difference-in-differences and triple-difference strategies. Results confirm the effectiveness of this program by showing a strong and positive impact on preschool attendance. This study then documents the benefits to alleviating undernutrition. Specifically, a preschooler from high-intensity provinces is 6 percentage points less likely to be underweight than their counterparts from low-intensity provinces. Estimates show a larger effect in rural areas, suggesting a result of narrowing the rural-urban gap in education access and undernutrition preventions. However, I find no impact on overweight probability. There is also evidence showing that the program has encouraged caregivers of preschoolers to refer kids to a doctor when kids get sick, instead of finding medicines by themselves.
Present-Biased Households and Monetary Policy (E7, E5)
We build a behavioral New Keynesian model with present-biased households’ and show that the natural interest rate and the present bias are positively related. The higher (weaker) the bias, the higher (lower) the natural interest rate in the economy. Our present bias endogenization demonstrates that higher present bias is likely to occur under lower macro volatilities, higher cognitive costs, higher relative risk aversion, and lower Frisch elasticity of labor. Our results elucidate key behavioral mechanisms driving the Zero Lower Bound (ZLB) during crises. High volatilities characterizing crises tend to increase the agents’ awareness and weaken their present bias, which in turn lowers the natural interest rate. This mechanism increases the likelihood of ZLB occurrence during crises. We study optimal monetary policy and Forward Guidance (FG) to assess the influence of the present bias on the optimal central bank behavior. To the extent that the present bias dampens the role of expectations, our model provides a solution for the FG puzzle.
Price Pressure and the Turn-of-the-month Effect: Evidence from Retirement Accounts (G1)
One popular explanation behind the turn-of-the-month (ToM) effect – the fact that stock returns are higher on days surrounding the turn of calendar months – is that people typically get their salaries at this time, and their investment in equities creates price pressure around those days. We test this hypothesis using a hand-collected comprehensive sample of mutual funds contained in 401K retirement accounts in the US, which constitute a substantial chunk of equity investments of the salaried. We find no evidence to support this hypothesis. While the ToM effect is still present in the data, stocks held by retirement funds which get large inflows of capital at ends-of-month do not exhibit a stronger ToM effect. On the other hand, we find evidence that stocks heavily held by retirement account funds have significantly higher returns in the middle-of-the-month. Evidence from Target date funds corroborates both these results.
Pricing Protest: The Response of Financial Markets to Social Unrest (E7, G1)
Using a new daily index of social unrest, we provide systematic evidence on the negative impact of social unrest on stock market performance. An average social unrest episode in a typical country causes a 1.4 percentage point drop in cumulative abnormal returns over the two-week event window. This drop is more pronounced for events that last longer and for events that happen in emerging markets. Stronger institutions, particularly better governance and more democratic systems, mitigate the adverse impact of social unrest on stock market returns.
Product Recalls, Market Size, and Innovation in the Pharmaceutical Industry (I1, O3)
The idea that research investments respond to market rewards is well established in the literature on markets for innovation (Schmookler, 1966; Acemoglu Linn, 2004; Bryan Williams, 2021). Empirical evidence tells us that a change in market size, such as the one measured by demographical shifts, is associated with an increase in the number of new drugs available (Acemoglu Linn, 2004; Dubois et al., 2015). However, the debate about potential reverse causality is still open (Cerda et al., 2007). In this paper, we analyze market size’s effect on innovation as measured by active clinical trials. The idea is to exploit product recalls, an innovative instrument, tested to be sharp, strong, and unexpected. The work analyses the relationship between US market size and innovation at the ATC-3 level through an original dataset and the two-step IV methodology proposed by Wooldridge et al. (2019). The results reveal a robust and significantly positive response of the number of active trials to market size. The first stage results display a negative and significant (Huber-White robust standard errors) impact of recalls and their lag on the log of sales at the market level. Such findings are enforced by analyzing abnormal values, which display the strong negative impact of recalls on sales. As measured by the impact of log sales on innovation, the second stage's outcomes reveal a robust and significantly positive response of active trials to market size. According to our estimates, an increase of 10% in market size leads to an increase of almost 6.3 % of active clinical trials. The effect of control variables (i.e., the share of generics, scientific production) aligns with theory.
Profit Shifting and Destination-Based Taxes (H2, F1)
We study the choice between source-based corporate taxes and destination-based corporate
taxes in a two-country model, allowing multinational firms to use transfer pricing to allocate profits across tax jurisdictions. We show that source-based taxation is a Nash equilibrium of a tax revenue maximizing jurisdiction if both domestic and foreign firms generate large revenues. Otherwise, the equilibrium involves destination-based taxes. We also show that a country has an incentive to unilaterally adopt destination-based taxes if it is much larger than the Foreign country.
Progressing towards Efficiency: The Role for Labor Tax Progression in Privatizing Social Security (E2, H3)
We show that labor tax progression can effectively substitute for the insurance implicit in redistributive social security, challenging the existing view in the literature that linking pensions to individual incomes wages reduces distortions associated with social security, but severely reduces welfare due to removing insurance. Our study shows that privatizing social security can deliver aggregate welfare gains if alternative channels of providing insurance are implemented.
We develop a stylized theoretical model. The agents participate in fully redistributive social security and pay a progressive labor income tax. In this setup, both the pension benefits and labor taxation provide insurance against income uncertainty. Then, we introduce a pension benefit proportional to individual contributions of each agent. In this setup, income shocks from the working period carry over to retirement, and the only source of redistribution is labor tax progressivity. We replace redistributive social security, where benefits partially pool between idiosyncratic income shocks with an arrangement where the individual pension benefit depends solely on individual contributions. We complement this change with an increase in redistribution through labor tax progressivity in during the working period. We show that there exist changes in labor income taxes accompanying such social security reform, which are fiscally neutral and at the same time raise welfare.
We take the intuitions derived from this stylized setup to a calibrated general equilibrium setup, replicating the features of the US economy. We characterize the conditions under which disincentives from redistributive social security may be reduced without loss of welfare. We compensate insurance loss of insurance from privatizing social security through greater progression in the instantaneous labor taxation. We find that for plausible calibrations of Frisch elasticity, privatizing social security coupled with labor tax progression may deliver aggregate welfare gains.
Promoting Student Diversity in the European Higher Education Area: Lessons Learned from a Good Practice in Germany (I2, H5)
The European Higher Education Area (EHEA) originates in the Bologna Declaration signed by education ministers from 29 European countries. Meanwhile, the supra-national network encompasses 49 member countries. Consistent with the United Nations Sustainable Development Goals, the EHEA now gives strategic priority to the social dimension of higher education. Specifically, policies of social inclusion and campus diversity are of great interest. Understanding which mechanisms lead to wider participation and success in higher education is essential for the optimal design and effectiveness of such policies.
In this paper, we analyze the relevance of non-traditional pathways to university and their importance for students with a migration background. In particular, we examine a good practice from Germany that prepares for fully fledged access to university. Our results show that historically underrepresented groups benefit from the availability of the good practice. Using a representative survey of first year university students in Germany, we find sizeable and significant effects. Specifically, second-generation immigrants are ceteris paribus 14 percentage points more likely to benefit from the good practice. Likewise, students from less educated family backgrounds are almost 8 percentage points more likely to enter university using entry channel enabled by the good practice. Students from academic households, however, are generally drawn to the traditional route to university. These findings add novel insights regarding the understanding of diverse communities of learners at universities and allow to draw policy implications for purposeful action to put forward one of the EHEA’s major goals.
We conclude that keeping young adults in compulsory education longer plus supporting them to upgrade school tracks can contribute to alleviating remaining disadvantages in the educational achievements. The good practice examined in this paper might, therefore, be considered for a rollout experiment across other EHEA countries.
Public Financing and Academic Achievement (I2)
I causally identify the effect of public financing on academic achievement and find improvement in public financing conditions results in a significant increase in both English Literature and Math test scores. However, the treatment effect distributes unevenly within county. I detect meaningful improvement among White students, but not among Black or Hispanic students. Consequently, the academic achievement racial gap significantly widens.
Public Infrastructure Procurement in Post-Conflict Power-Sharing Arrangements: Evidence from Lebanon’s Council for Development and Reconstruction (H1, L3)
Post-conflict power-sharing arrangements not only allocate political power but also economic re-sources among powerful elites. Previous research on such economic arrangements has largely focused on the access of powerful elites to natural resources, such as to oil fields or mines. This article investigates the mechanisms of rent allocation of another major source of revenues: Public procurement of large infrastructure projects. We analyze a unique dataset of all infrastructure procurement contracts awarded between 2008 and 2018 by Lebanon’s Council of Development and Reconstruction (CDR), the major state institution to finance and implement large infrastructure projects after the country’s civil war (1975-1990). We qualify the extent to which politically connected firms capture higher project values by differentiating their “quality” of connections. We find that connected firms capture higher project values, however, only those firms that are directly connected to the board of CDR and their political protégés, rather than firms connected to the wider set of powerful political elites. We argue that the arrangements to share economic resources based on such formal state institutions are upheld by norms of power-sharing behavior. The case of CDR moreover illustrates that efforts for capacity building via international donors or technical interventions are unlikely to yield the desired results in the absence of reforms to enhance accountability of elites.
Public Sector Balance Sheet Strength and the Macro Economy (H6, H5)
This paper introduces concepts of public sector balance sheet (PSBS) strength, taking into account different aspects of what governments own in addition to what they owe. It develops measures of PSBS strength and investigates their macroeconomic implications. Empirical estimations show that in their pricing of sovereign bonds, financial markets account for government assets and net worth in addition to their liabilities. Furthermore, economies with stronger public sector balance sheets experience shallower recessions and recover faster in the aftermath of economic downturns. This faster return to growth can be explained by the greater space for countercyclical fiscal policy in countries with stronger balance sheets.
Public Support and Enterprise Performance: Evidence from Firm-Level Data in Ethiopia (M2, J0)
The purpose of this study is to evaluate the impact of public contract to small and medium-sized enterprises (SMEs) on their performance. We use a rich firm-level dataset from the World Bank Enterprise Surveys and apply quasi-experimental approaches (endogenous treatment effect and propensity scores weighted difference-in-difference) to test the effect of public contract awards on SMEs sales and employment growth. Our results indicate that public contracts enhance SMEs sales and employment. The results highlight the positive role of public contracts as a promotion strategy to improve firm performance and address the unemployment. Further, our results have implications for government supports to businesses in the era of COVID-19 crisis that affected businesses with demand shock. We conclude by providing policymakers advice to diagnose the challenges that firms face in low-income countries and craft a more targeted support policy to address the unemployment and firms’ growth issues in developing countries like Ethiopia.
Pulp Friction: Long-Term Contracts as Quantity Insurance (L2, D8)
In decentralized markets for homogeneous goods, firms often trade using long-term contracts even though spot trade is more flexible to changing market conditions. We argue that long-term contracts are valuable in decentralized markets with frictions because they provide informal quantity insurance from the risk of being matched with low value trading partners -- buyers and sellers become endogenously risk averse with respect to quantity. Using proprietary invoice and production data from a large seller, we estimate a model of long-term contracts and spot trade in the pulp and paper industry to size the value of long-term contracts and the costs of inflexibility. We find that eliminating long-term contracts would decrease median surplus by 70%, while absent any trading frictions it would increase it by 139%.
Quantile Connectedness of the U.S. Interbank Liquidity Risk Network: a Bayesian Nuclear Norm Estimation Approach (G2, C1)
Understanding the connectedness of financial networks is central to the study of shocks transmission and systemic risk.
Many studies investigate the connectedness of various financial networks, focusing on financial assets connectedness among financial institutions, at the conditional mean.
In contrast, we study the quantile-dependent U.S. interbank liquidity risk network based on a non-publicly available supervisory dataset.
Both methodological and empirical contributions are new to the literature.
We first propose a new Bayesian nuclear norm estimation method that automatically estimates the quantile connectedness of the network.
We employ a common factor structure to deal with unobserved heterogeneity that may exhibit endogeneity within the network.
Together with a quantile vector autoregressive model with common factors, we then apply the proposed Bayesian method to the supervisory dataset to study the quantile-dependent liquidity risk network among large U.S. bank holding companies between April 2017 and December 2020.
We find that the liquidity risk network varies across quantiles and has changed substantially during the COVID-19 pandemic relative to the pre-pandemic period.
The estimated quantile liquidity network connectedness measures could be useful for bank supervision and financial stability monitoring by providing leading indicators of the system-wide liquidity risk connectedness not only at the median but also at the tails, as well as identifying systemically important banks and vulnerable banks in the liquidity risk transmission of the U.S. banking system.
Quantitative Forward Guidance through Interest Rate Projections (E5, G1)
We assess the effectiveness of quantitative forward guidance through interest rate projections along four key dimensions: (i) predictability, (ii) credibility, (iii) consistency and (iv) non-redundancy. Based on data for the Reserve Bank of New Zealand, the Norges Bank, the Swedish Riksbank and the Federal Reserve we find that the interest rate projections released by these four central banks are predictable and credible. Market expectations of the future path of interest rates anticipate changes in the central bank projection path and adjust to path surprises. The adjustment is, however, not one to one and decreases with the projection horizon. Moreover, high uncertainty around the projection path reduces the impact of path surprises. We also find the interest rate projections to be consistent with the macro projections that are released by the four central banks in parallel as these projections are empirically linked by a stabilising Taylor rule. Finally, interest rate projections are not redundant as they impact market expectations also when controlling for the effects of macro projections.
Rainy Day Liquidity (G1, G2)
Insurance firms are a key player in the corporate bond market. In this study, we consider the role of life insurers as "rainy day" liquidity providers who improve liquidity in stressful conditions due to persistent activity of buy-and-hold investing for long-term horizon. To this end, we present evidence that insurers' corporate bond purchases improve bond liquidity during the financial crisis and among downgraded bonds facing selling pressure. Life insurers are net purchasers of downgraded bonds within investment grades. Downgraded bonds that life insurers purchase in larger amounts are significantly better priced.
Raising the Inflation Target: What Are the Effective Gains in Policy Room? (E4, E3)
The return of the zero lower bound on interest rates poses a serious challenge for central banks because it leaves them without their main instrument. This situation is prompting policymakers to consider raising their inflation target to regain policy room. We show that the gains generated by this strategy are not one-to-one: Because a higher inflation target leads to a steeper Phillips curve, to get, for instance, 2 percentage points of extra room, policymakers need to raise their inflation target from 2\% to 5\%. Moreover, taking this mechanism into consideration raises the optimal inflation target by 1 pp.
Re-estimating Potential GDP: New Evidence on Output Hysteresis (E3, E4)
We propose a simple structural method to estimate potential (or flexible-price) output. Our
baseline method is derived from a New Keynesian model with wage rigidities, yet it is consistent with a wide range of assumptions on technology, preferences, policy rules and expectation
formation. Our approach has two main advantages over existing methods. First, it provides an
unbiased estimate of potential GDP that is not subject to the Lucas Critique. Second, unlike
other model-based approaches, our method does not require prior knowledge of all deep parameters and it does not resort to Bayesian estimation or calibration of the underlying model. Hence,
our estimates are consistent with a large set of possible parametrizations. We compute potential
GDP for the US and find that the output gap implied by our method is highly correlated with
the one computed by the Congressional Budget Office (CBO), a widely used estimate. However,
there is a stark difference between the two series during and after the Great Recession. We also
use our potential GDP estimates to contribute to the debate on the effects of demand shocks
on aggregate supply. We show evidence supporting hysteresis hypotheses claiming that demand
shocks can affect potential GDP
Reading between the Lines − Using Text Analysis to Estimate the Loss Function of the ECB (E5, E3)
We apply textual analysis to extract the tone (sentiment) from the introductory statements to the ECB’s press conferences regarding economic outlook. By combining this information with Eurosystem/ECB staff macroeconomic projections, we are able to directly estimate the Governing Council’s loss function. Our analysis suggests that the de facto inflation aim of the ECB may have been considerably below 2%. We also find evidence that the loss function has been asymmetric, which would mean that the ECB has been more averse to inflation above 2% than below 2%.
Reading Between the Lines: Quantitative Text Analysis of Banking Crises (G2, C4)
This paper develops five indicators based on a large corpus of economic news articles to forecast financial crises. The methodological approaches feature the identification of key topics within a large volume of texts, as well as the measurement of similarity between texts. A Banking Crisis Lexicon Index and Sentiment Index are developed through analysing a vast amount of economic articles to detect the evolution of banking sector discourse. Findings from Granger causality highlight leading indicators status and receiver operating characteristics suggest robust forecasting performance strength of the Banking Crisis Lexicon Index, globally and for developed economies up to two years preceding a crisis. While the aggregated Sentiment Index constitutes a coincidental indicator, for developed economies it is a short-term leading indicator. A combined lexicon and sentiment index exhibit solid forecasting performance. Statistical models Wordscores and Wordfish are introduced to study banking crises and underscore crisis classification strength. A hand-coding approach is used to verify the veracity of the indices and provides credence to the vital contribution of published deliberations in understanding and detecting banking sector frailties. In reading between the lines, this paper contributes to the literature on quantitative text analyses in constructing text-based latent banking crisis indicators.
Reaganomics 101 – Reaganomics Is More than the Laffer Curve (H6, E6)
What is commonly called Reaganomics is often inadequately covered (if at all) in Principles of Macroeconomics textbooks. Working with the Ronald Reagan Presidential Library, the California Council on Economic Education seeks to remedy that by developing three classroom lessons about President Reagan’s economic policies. What is commonly called Reaganomics includes deregulation, tax cuts, and a balance Federal budget. Attend this session to receive interactive in-class activities illustrating inflation, unemployment and supply side economics and get web-based research activities that evaluate the success of Reaganomics.
Recovery from Mobility Limitations among Older Americans: The First-Interview Effect (I1, J1)
Measures of mobility limitations (even self-reported ones) have been considered as objective measures of health status and few concerns have been raised about their potential biases. Using data from the Health and Retirement Study, we observe that the incidence of recoveries from mobility limitations is very high between the first and second time that individuals are interviewed and declines sharply to a stable level in subsequent waves. Based on this observation, we question the initial accuracy of measures of mobility limitations, which could be later ameliorated through individuals’ ability to learn about their health capacities over time. We provide empirical evidence to show that this unusually high incidence of recovery from mobility limitations could be in part the result of respondents’ improvement on health knowledge of questions on mobility limitations or improved experience with the activities instead of caused by real health improvement. Our results are not only relevant to any empirical researchers using mobility limitation indicators in a panel data setting, but also to researchers analyzing the effectiveness of policy interventions on health outcomes using self-reported health measures, since the effects of these policies are likely to be confounded with improvements in health knowledge, and the results be biased towards surprisingly large short-run effects and much smaller medium and long-run effects.
Redistribution of Return Inequality (H3, E6)
Wealthier households obtain higher returns on their investments than poorer ones. How should the tax system account for this return inequality? I study capital taxation in an economy in which return rates endogenously correlate with wealth. The leading example is a financial market, where the rich acquire more financial information than the poor. Contrary to conventional wisdom, rather than calling for more redistribution, the presence of this scale dependence provides a rationale for lower marginal tax rates. The leading example is a financial market, where the rich acquire more financial information than the poor. Contrary to conventional wisdom, rather than calling for more redistribution, the presence of this scale dependence provides a rationale for lower marginal tax rates. The endogeneity of returns generates an inequality multiplier effect between wealth and its returns. Therefore, standard elasticity measures that determine the responsiveness of capital to taxes must be revised upwards. At an aggregate level, a rise in redistribution induces a compression effect on the distribution of pre-tax returns. In the financial market, I identify general equilibrium trickle-up externalities that provide a force for more redistribution relative to the partial equilibrium. Finally, I estimate partial and general equilibrium responses and demonstrate the quantitative importance of scale dependence for tax policy.
Referral Behavior When Matching Is Important (C7, I1)
Why do physicians in a medical specialty on occasion refer patients to another physician in the same specialty? Altruistically doing so may make sense: physicians each have different capabilities and for a severely ill patient being referred to the right physician may determine life or death. While altruism, capacity constraints and reciprocity all play their part, in this paper we develop a dynamic model of strategic referrals. Consider a market with information asymmetry between heterogeneous, well informed physicians and poorly informed patients. Each patient is aware that he faces a problem, but can determine neither its nature nor the appropriate physician. Consequently, he chooses to consult the physician with the best reputation for successful outcomes. A physician is paid first for diagnosing the patient’s problem. After diagnosis, she then has an incentive to generate further revenue from treating the patient. She, however, may be reluctant to continue with cases that, given her skills, are difficult and may result in a poor outcome tarnishing her reputation. Referral to a more apt rival may then be advantageous, despite the revenue loss, because it results with higher probability in a cure, and is thus beneficial for the referring doctor’s reputation. In our model of strategic referrals, we find that physicians have an incentive to refer, but generally not at efficient levels. Moreover, they lack an incentive to invest in their own skills. This is because, in equilibrium, not absolute but relative reputations determine profits. We show that the matching between physicians and patients improves when physicians sign revenue pooling contracts as is common in law, accounting, and business consulting generally. These contracts may also alleviate physicians’ reluctance to invest in their craft. Finally, we argue that this intuition remains true when we consider capacity constraints and disease severity.
Regional Risk and Aggregate Fluctuations (E3, E6)
This paper highlights time-varying regional risk and federal fiscal transfer policy as two competing forces driving regional risk sharing over the business cycle and quantifies their impacts on aggregate fluctuations. I uncover that regional risk is strongly countercyclical: the conditional standard deviation of idiosyncratic shocks to U.S. state-level output growth is estimated to be 40% larger in NBER recessions than in normal times. However, the extent to which regions insure against these risks has not varied much over the business cycle. I argue that federal fiscal transfers that operate in a "state-contingent" manner during recessions can reconcile these results. To understand what these patterns of regional risk and fiscal integration imply for the aggregate, I build a New Keynesian model of heterogeneous regions with an incomplete asset market. The calibrated model suggests that during an economic downturn, increased regional risk worsens risk sharing, due to more regions being constrained at the borrowing limit; meanwhile, it amplifies the impact of aggregate productivity shocks, due to stronger precautionary saving motives leading to lower aggregate demand. However, state-contingent federal government transfers stabilize both the regional and aggregate economy, by providing insurance to the regions that need it the most - those with low productivity and high marginal propensity to consume. Taken together, this paper demonstrates the power of regional redistributive policies as automatic stabilizers in a fiscal union.
Reinforcement Learning for Household Finance: Designing Optimal Policy for Mortgage Market (G5, C6)
We apply reinforcement learning (RL) techniques from machine learning (ML) to improve efficiency of residential mortgage market in a dynamic setting which consists of three agents: the borrower (household), lender (financial institution that loans the money) and servicer (intermediary that sends mortgage statements). Our objective is to understand intertemporal choice of the borrower on the one hand and pricing, investment and portfolio allocation, transaction costs, tax consequences of the issuer/lender on the other. To attain this objective, we optimize the hedging of contingent claims (cashflow from different plausible paths that a loan take) by the servicer. By implementing the optimal policy, the servicer can preempt moral hazard from the borrower's behavior and increase the mortgage market efficiency. Our approach differs from the conventional (adhoc) approach which depends on the qualitative legal and industry expert judgement. RL is a model-agnostic methodology of solving dynamic optimization problems, adding action space on top of state space used in classical approaches.
Relational Contracts in Frictional Markets with Rematching (C7, L1)
How do market opportunities affect the value and dynamics of relationships? This paper studies a market in which matched principals and agents interact via relational contracts, and unmatched principals and agents can randomly and anonymously rematch with each other. Market characteristics like search friction and market thickness determine the probability of rematching and forming new relationships. I show that a large rematching probability leads to non-stationary and less productive relationships, while a small rematching probability leads to stationary and productive relationships. Welfare is non-monotonic in the rematching probability, as a smaller rematching probability facilitates incentive provision but makes forming relationships harder.
Relationship between Students’ and Families Growth Mindset with Students’ Academic Performance in Math: The Case of the Dominican Republic (I2, I3)
Recent academic literature on education, psychology, and education of economics has widely documented a strong significant relationship between growth mindset - the belief that intelligence is not fixed and can be developed - and students’ academic achievement, particularly when these two variables are controlled by students’ socioeconomic background. Current findings in the Latin America region suggest that student growth mindset can temper the effects of poverty on academic achievement, but our best knowledge no study has considered the interplay between students’ and their families’ growth mindset on academic achievement. We use a novel database of 9th graders students and their families in the Dominican Republic (N=110,000). Preliminary results show that families’ growth mindset strongly predicts students’ growth mindset, and also their math academic achievement. Multilevel regression analysis depicts differences in math scores ranging from 0.1 to 0.4 standard deviations. These findings suggest that students’ mindset are highly influenced by families’ mindset and may be a mechanism to elicit explicit academic behaviors, especially in under-privileged students’ and families
Relaxing Credit Constraints: Credit Risk and Financial Intermediation Quality (G2)
This paper studies the effect of credit supply shock on firm’s credit risk and bank lending quality. We exploit a quasi-experimental setting generated by a regulatory change that extended the range of firms eligible for priority sector lending (PSL) in India. Comparing profiles of firms around the cutoff of PSL eligibility, we find a substantial local effect on firm’s liabilities and credit risk. The credit supply shock is targeting firms with better economic outlook (higher sales and profitability growth, higher credit rating) and commonly considered financially constrained (private and younger) suggesting that the mandated credit expansion does not compromise the intermediation quality of banking sector.
Reluctant Savers and Mortgage Subsidies (E7, E6)
The Mortgage Interest Deduction (MID) ranks among the largest tax expenditures in the US tax code. Despite the many proposals aiming to revise the MID, the ultimate effects of modifying it are still ambiguous and controversial. One argument against this policy points to its regressivity, as it mostly benefits wealthy homeowners. On the other hand, the MID could alleviate self-control problems by promoting savings in an illiquid asset. This paper focuses on this interaction between progressivity and self-control, by studying the effects of eliminating the status-quo MID within a heterogeneous agents model with incomplete markets and Gul-Pesendorfer preferences. Ignoring self-control issues can lead to underestimating the average welfare gains associated with the MID repeal. Crucially, MID tax savings constitute a liquid form of income which exacerbates self-control costs of the losers from the policy reform.
Resource Cost of Income Sheltering Behaviours and Elasticity of Taxable Income (H2, H0)
How much are high-income taxpayers willing to forgo to shelter one dollar of their income from taxes? Exploring a policy change in Australian individual income tax schedule and using administrative data, we provide the first estimate of marginal resource cost of income sheltering behaviours and Elasticity of Taxable Income (ETI) for high-income individuals.
The policy change increased the top tax threshold to AUD$180,000 from AUD$150,000 in 2008-2009 tax year. The marginal taxes below and above the threshold are respectively 40 and 45 percent. This policy change induces large incentives for income sheltering behaviours to decrease the tax liability. Individuals face incentives to locate – “bunch”– right below the threshold where the marginal tax rate is lower. After the policy change, some individuals bunch at the new threshold while some others continue bunching at the former threshold, suggesting that they face costs for sheltering their income from taxes. We use the amount of bunching at the new and former thresholds to estimate our bunching model.
Progressive income tax systems, in which the marginal tax rate increases as taxable income rises, are used as a policy instrument to redistribute income from high- to low- income individuals. The ETI for high-income individuals is relatively high, suggesting that their taxable income is very sensitive to tax rates. However, it has been suggested that the higher sensitivity might be due to a low marginal resource cost of income sheltering behaviours,1 the amount they are willing to forgo to shelter one dollar of income from taxes, rather than a more overly sensitive labour supply. Whether the income sensitivity is driven by income sheltering behaviours rather than labour supply has important implications for the efficiency cost
Retirement, Housing Mobility, Downsizing and Neighbourhood Quality-A Causal Investigation (J1, R2)
This paper provides the first causal evidence on the impact of retirement on housing choices. Our empirical strategy exploits the discontinuity in the eligibility ages for state pension as an instrument for the endogenous retirement decision and controls for time-invariant individual characteristics. The results show that retirement leads to a statistically significant and sizable increase in the probability of making a residential move or the likelihood of becoming outright homeowners. We also find that individuals downsize both physically and financially and tend to move to better neighbourhoods or closer to the coast upon retirement. We additionally discover that some housing adjustments take place up to 6 years before retirement. Moreover, our results reveal significant heterogeneity in the retirement impact by gender, marital status, education, housing tenue, income and wealth. Within coupled households, housing mobility choices are primarily influenced by the wife’s retirement while housing downsizing decisions are only affected by the husband’s retirement. The results suggest that failing to address the endogeneity of retirement often under-states the retirement impact on such housing arrangements. From a policy perspective, our findings suggest that policies to increase retirement ages would also postpone the retirement attributable housing adjustments among older people. The evidence of the substantial housing mobility and downsizing around retirement may be useful input to housing or fiscal policies as it demonstrates that the unleashing of accumulated housing equity occurs well before some major life course events identified in the current literature.
Retiring from Unemployment: The Role of Personal Finances (I3, J6)
Hetschko, Knabe, and Schöb (2014) demonstrate that unemployed individuals’ life satisfaction (LS) increases after they retire, and use social identity to explain the positive effects of retirement. The explanation is widely accepted in academia.
This paper, however, suggests that LS increases are due to financial situation changes. As in Hetschko, Knabe, and Schöb (2014), this paper uses a difference-in-differences (DiD) estimation to compare the retirement transition of employed and unemployed individuals based on the German Socio-Economic Panel (SOEP). The innovation of this paper is to combine the DiD with a matching approach based on entropy balancing. Moreover, in the matching process, two procedures are utilized for control variable selection: manual selection and automated selection based on the state-of-the-art machining learning methods Lasso and LightGBM.
Using the above empirical strategy, this paper first shows that for the employed, the retirement transition substantially decreases income. By contrast, the unemployed experience an increase in income after retirement. The transition from unemployment is associated with a greater LS increase. This paper then conducts two heterogeneity analyses, which demonstrate that the LS increase is concentrated on those whose financial status is better.
This paper makes three contributions. First, to the authors’ knowledge, this paper is the first to suggest that the LS increase during retirement from unemployment is due to financial improvement. Second, novel empirical methods, including entropy balancing matching and Lasso/LightGBM feature selection, are utilized. Third, this paper also contributes to the literature on unemployment and happiness.
Explaining why LS increases during retirement from unemployment is important for policy recommendations. Nonmaterial-based explanations suggest that finding a job is the most important consideration in fighting against unemployment, regardless of job quality or wage. Material-based explanations, by contrast, emphasize a decent living standard. This suggestion poses more challenges to policy makers.
Return Migration and the Labor Outcomes of Adolescents: Evidence from the USA-Mexico Corridor (J1, O1)
The US-Mexico corridor represents one of the most massive migrant movements in human history. Since 2007, the trend has reversed with more Mexican migrants returning home than those migrating to the U.S. In this study, I examine the causal effect of return migration on the school-work tradeoff and selection into employment types of children aged 12-19 years in Mexican households. I use the Mexican census of 2010 and various other sources to construct a unique dataset. I employ the control function approach and use U.S. state-level immigration enforcement acting as push factors as an instrument to address the endogeneity of return migration. My results suggest an increase in the probability of school attendance, a decrease in labor market participation, and a decrease in the probability of working and going to school simultaneously for children of households with return migrants, relative to non-migrant households. Moreover, I find a decrease in the probability of employment in wage/salaried work, and an increase in self-employment among children in return migrant households. I speculate that these improvements are driven by the migrants' experience, accumulation of human and financial capital in the United States, as well as better labor market opportunities when they return. My paper suggests return migration from a developed to a developing country as a mechanism through which migrant flows may benefit origin developing countries worldwide. Policies aimed at assisting the reintegration of return migrants in local markets may substantially improve the quality of education and can act as a channel to reduce child labor.
Returns to Scale in Property Value Assessment: Evidence from New York State (H7, H2)
This paper takes advantage of cooperative agreements among small tax-assessing jurisdictions to explore economies of scale in property valuation. New York State incentivizes neighboring cities and towns to unify their assessment function while maintaining separate taxing authority. With administrative data for years 2003-2014, we estimate cost function models with an instrumental variable estimator to examine whether collaboration reduces assessment cost. We use border intersection and prior cooperation in service provision at the jurisdiction level to address potential bias in selecting coordination partners. Our results show consistent evidence of economies of size; expanding assessing unit size may save operational and contractual cost. We also find that potential collaboration among assessing units may incur additional contractual expense in the short run.
Revenue Elasticities and Macroeconomic Dynamics (E3, D2)
Recent literature on market power in macroeconomics has noted the limitations of using revenue elasticities to proxy output elasticities when estimating markups. Although revenue elasticities do not unlock markups, we can -- to some extent -- circumvent markup estimation and make use of revenue elasticities. We investigate the possibilities and limitations of using revenue elasticity, instead of markups, as a determinant of business cycle behaviour. Using U.S. firm-level data, we measure revenue elasticities and estimate impulse responses using local projections. Consistent with theory, higher revenue elasticity firms generate greater business cycle amplification.
Reward, Punishment and Children’s Cooperation Preference (C7, J1)
Based on a large-scare field experiment of public goods games from more than 1600 students aged 7-17, this paper compares and discusses the effects of different incentive mechanisms, including endogenous reward, exogenous reward, endogenous punishment and exogenous punishment, on children’s cooperation preference. First, we found that all the mechanisms could significantly promote children’s cooperation level, but the effects are different. Exogenous mechanisms are more effective than endogenous mechanisms, and punishment is more effective than reward. However, at the same time, some students still showed a tendency to maintain the same supply level of public goods even there existed an institution. Then, we asked children about their preference for rewards/punishments and found that reward is more popular. More cooperative individuals are more likely to be supporters of incentive policies. Thirdly, we introduced uncertainty into reward/punishment and found that mechanisms with 50% probability of enactment could also improve children’s cooperation significantly, but the effects are inferior to certain ones. Furthermore, we found a significant positive correlation between students’ supply level and their belief in the contribution of other members. Finally, we found heterogeneous effects of rewards and punishments on different group of subjects. Girls contributed more facing rewards rather than punishments, while boys showed the opposite side. Compared with previous studies, this paper systematically analyzed the role of reward and punishment in the improvement of children’s cooperation. Our findings provide experimental evidence for policymakers to adopt cost-saving and individualized incentive measures to guide the development of children’s non-cognitive skills.
Right-to-Work Laws, Unionization, and Hourly Wages (J5, J3)
Right-to-work (RTW) laws prohibit labor unions from requiring that all workers covered by a union-negotiated contract contribute financially to the union. This allows workers to opt out of paying dues or fees while retaining union contract benefits like wage raises — a classic free-rider problem. If too many workers free-ride, union funding decreases, which may weaken contract bargaining and reduce union organizing in nonunion workplaces. Although RTW laws have historically existed in states with historically low union concentration, five states in the Rust Belt — a region with a historically high prevalence of unionism and manufacturing — enacted RTW laws since 2010. Drawing on 14 years of data on private sector workers from the Current Population Surveys, I use Sun and Abraham’s (2020) interaction-weighted difference-in-differences event study estimator to find that the recent wave of RTW legislation led to sustained reductions in union membership ranging between 1.0 and 2.7 percentage points per year, with similar effects on union contract coverage. Meanwhile, RTW reduced wages by between 0.6% and 2.7% over its first three effective years. However, I find only a short-term spike in free-riding (defined by union contract-covered workers not paying membership dues) and no wage reduction among contract-covered workers. This suggests that wage effects are driven by reduced union expansion into new workplaces, potentially associated with union leaders’ fear of future free-riding rather than free-riding itself. Thus, workers in RTW states who would otherwise join unions or work under union contracts do not recieve a union wage premium, dragging their wages down compared to their non-RTW state counterparts. Further, I show how RTW’s effect varies by demographics and socioeconomic status, finding that workers of color, workers in traditionally unionized industries, and male workers experience higher magnitude reductions in union membership and wages than others.
Risk-Aversion and the Bifurcated Interest Responses of Corporate Investment: Theory and Evidence (E4, E7)
This paper analyzes how the bifurcated interest response of risk-averse investment challenges a negative relationship between investment and interest rates. Risk-averse firms weigh the certainty equivalent of a risky investment return against the safe return of a riskless project. If investment return (θ) and the interest rate (r) are sufficiently low, while firms often prefer holding safe assets over undertaking risky investment projects, there is an increasing θ-r boundary along which investment can be maximized when its return bifurcates its interest response. Specifically, my model shows that investment increases with the interest rate if the return on investment exceeds its boundary value but decreases with the interest rate otherwise. Furthermore, investment tends to respond more to an increase than a decrease in the interest rate. Essentially, the improved investment return is the driving force for risk-averse investment whereas the peripheral role of a higher interest rate is mainly to lower the safe return and associated risk premium. I empirically demonstrate that the theoretical predictions of my model - i.e., bifurcated interest responses of investment and an asymmetric response pattern with respect to increases versus decreases in the interest rate - are empirically borne out using U.S. nonfinancial business data.
Robustly Optimal Monetary Policy in a Behavioral Environment (E3, E5)
There is a wide consensus that uncertainty affects the conduct of monetary policy. There is less consensus on the impact of uncertainty on monetary policy decisions. The first ever contribution to this question, (Brainard, 1967), has established what is called Brainard’s attenuation principle; i.e. the presence of uncertainty implies an attenuated policy response compared to settings where uncertainty is not taken into account. A recent literature contested this result showing, in particular setups, that uncertainty
leads to aggressive policy actions (e.g. Giannoni, 2002)). This paper provides a reexamination of Brainard’s attenuation principle in a specific setting, a behavioral New Keynesian model entailing a behavioral (psychological) parameter; i.e. cognitive discounting. We assess the robust optimal policy with uncertainty regarding this behavioral parameter, but we also revisit the robust policy with uncertainty regarding the slope of the Phillips curve. Introducing uncertainty in the behavioral parameter seems natural, as we lack solid empirical evidence on its numerical values. Further, as this parameter pertains to psychological underpinnings of individual decision-making, it is hard to be pin it down to a specific value. More interestingly, this parameter falls in the characterization of Barlevy (2011), in parameters affecting the persistence, but not in the conventional way.
We show that in face of uncertainty regarding the cognitive discounting parameter, the robust optimal policy implies an interest rate reaction lower than what is suggested by the otherwise optimal policy without robustness. We confirm Brainard (1967)’s attenuation principle in this case, which is against Barlevy (2011)’s rationalization theory of aggressive responses to uncertainty. For price rigidity uncertainty, our results are in line with the previous literature. Interestingly, when the central bank faces a joint uncertainty, we show that the robust policy is in line with the attenuation principle.
Role of Academic in Improving the Value-Based Education in India (I2, A2)
Quality education is a very powerful instrument for combating poverty and inequality in India. Therefore, ensuring universal access to quality education is central to the economic and social development of the country (The World Bank, 2011). The current Indian educational system lacks access to technology and qualified teachers. The pedagogical strategies used in higher education do not support character development, as is evidenced by an increase in violence and social unrest. India can learn from its ancient history, when its educational system was first designed to develop the whole person, by adopting this approach for its current system. India should consider incorporating value-based education in order to promote a sense of civic responsibility and social values in students.
Satellite University Campus and Local Government Finance: Evidence from China's Higher Education Expansion (H7, H3)
In the last decade, a growing body of research shows that higher education institutions (HEIs) bring about local economic growth (Wang & Tang, 2020; Hausman, 2017; Liu,2015; Cantoni & Yuchtman, 2014 ). Less is known, however, about when and why local governments would be interested in attracting HEIs to their localities. Existing evidence from developing country context is even more scarce. China, the fastest-growing large-size emerging economy in the past forty years, has set up 452 new university campuses nationwide to accommodate the unprecedented expansion of higher education since 1999. This large natural experiment offers a unique opportunity to examine the decision-making process of local government in this regard and its consequent impact on local government finance.
Drawing on the public choice theory by James Buchanan and the fiscal federalism literature by Musgrave and Oates, we hypothesize that local governments that are under severe fiscal pressure would be more likely to engage HEIs to set up satellite campuses and the influx of HEIs would boost the local fiscal revenues through appreciation of land price first and other sources of revenues later on.
We manually collect the detailed information of each and every new campus between 1999 and 2007 and link it to the local government finance yearbook to construct a unique panel dataset at the county-level. Survival analysis, difference-in-differences, and event study model are performed drawing on the temporal and geographical variation in the setup of new campuses. The findings are consistent with our hypotheses. Heterogeneity analysis shows that effects are much larger in less developed western regions.
These findings shed new lights on the micro mechanisms of universities’ impact on economic growth and the role of higher education in the process of China’s urban construction and economic development during the late 90s and early 2000s.
Saving Neonatal Lives for a Quarter (I1, C4)
Over 400,000 children die annually from a bacterial infectionn the blood despite several RCTs suggesting that this can be eradicated by chlorhexidine umbilical cord care (CHX) -- for only US$0.23 per dose. Unresolved treatment effect heterogeneity across RCTs and lack of evidence outside an experimental setting contribute to the slow international scale-up of CHX. Exploiting the national roll-out of CHX in Nepal, we combine a two-way fixed effects approach and recently developed machine learning methods to estimate the effect of CHX in real-life conditions, investigate sources of heterogeneity, and derive optimal targeting policies. We confirm the effectiveness of CHX at scale and show evidence of significant treatment effect heterogeneity consistent with the mixed conclusions found across different RCTs. Policy rules based on a few easily observable characteristics can ensure that subpopulations with large benefits receive CHX.
School Entry and Leaving Laws and Compensating Wage Variations over Life Cycle (I2, J3)
This study investigates the effects of school starting age and compulsory schooling duration on labor market outcomes from a life-cycle perspective. For identification, I use geographical and temporal variation from independent changes in cutoff rules for enrollment and compulsory schooling requirements that occurred in Germany during the 1950s and 1960s. Using administrative data with nearly career-long labor market biographies, I study the impacts of the two common policies for the same cohorts and within the same institutional setting. This allows me to shed light on their relative importance and potential interaction effects. I find that school starting age has little impact on labor market outcomes from a life cycle perspective. If anything, a later school start generates small earnings losses despite increased educational attainment. In contrast, extended compulsory schooling yields significant benefits in terms of higher lifetime earnings and increased labor supply. However, returns to compulsory schooling differ substantially by school starting age. Specifically, younger school starters benefit relatively more from extended schooling compared to older entrants, which suggests that later policies can mitigate initial disparities due to school entry rules. This yields important policy implications.
School Track Decisions and Teacher Recommendations: Evidence from German State Reforms (I2, J2)
I study the effects of admission requirements, a common regulation to determine program
assignment, in the context of school track decisions. Depending on the federal state in Germany, either teachers or parents have the discretion to decide which secondary school track
a child may pursue after primary school. Applying a differences-in-differences approach,
I exploit variation in the implementation and abolition of binding teacher recommendations across states and over time to investigate its effects on students’ academic outcomes.
Using data from Germany-wide large-scale skill assessments, I show that binding teacher
recommendations significantly improve student achievement in fourth grade, prior to track
assignment. Effects persist into ninth grade, after consequential track assignment. Further
analyses suggest that effects are driven by increased time investments in students’ skill
Selection into Informative Consumer Credit Markets (G5, D1)
Technological innovation facilitated households’ access to information about their cost of credit. We exploit a quasi-natural experiment in an online consumer credit market to identify which households take advantage of informative markets. We find that when a platform switched from personalized loan prices to prices by credit grade – less experienced individuals immediately and disproportionately exit the market, especially among riskier borrowers. We conclude that less experienced borrowers sort into markets offering personalized information. Additional analysis confirms that their behavior is consistent with learning from personalized prices. Our results highlight the important, yet overlooked, informative role of the growing Fintech sector.
Self-Fulfilling Liquidity Traps in a World Economy (F3, F4)
This paper studies a New Keynesian model of a two-country world with a zero lower bound (ZLB) constraint for nominal interest rates. A floating exchange rate regime is assumed. The presence of the ZLB generates multiple equilibria. The two countries can experience recurrent liquidity traps induced by the self-fulfilling expectation that future output and inflation will be low. These “expectations-driven” liquidity traps can be synchronized or unsynchronized across countries. In an expectations-driven liquidity trap, the domestic and international transmission of persistent shocks to productivity and government purchases differs markedly from shock transmission in a “fundamentals-driven” liquidity trap.
The paper is available here:
Author web page: http://www.robertkollmann.com/
Sentiment in Bank Examination Reports: Supervisory Information and Bank Outcomes (G2, C8)
We conduct textual analysis on over 3,000 small to medium-sized commercial bank examination reports from 2006 to 2016. These confidential examination reports provide textual context to each component of the supervisory CAMELS ratings; capital adequacy, asset quality, management, earnings, liquidity, and sensitivity to market risk. Each component is given a categorical rating, and each bank is then designated an overall composite CAMELS rating along the same scale, which are used to determine the safety and soundness of banks. We find that controlling for these ratings and a variety of other factors, the sentiment supervisors express in describing the asset quality and earnings in their respective sections provides significant additional information in predicting future indications of asset quality and various measures of profitability, respectively. The sentiment in describing management appears to help predict supervisory actions. This suggests that bank supervisors play a meaningful role in the surveillance of the banking system, especially at critical junctures when risks are transitioning.
Separating Retail and Investment Banking: Evidence from the UK (G2, E5)
The idea of separating retail and investment banking remains controversial. Exploiting the introduction of UK ring-fencing requirements, we document novel implications for credit supply, competition, and risk-taking via a funding structure channel. By preventing retail deposits from funding capital market activities, this separation leads universal banks to rebalance towards mortgage lending and away from supplying corporate credit lines and underwriting services. By redirecting the benefits of deposit funding towards the retail market, this rebalancing reduces the cost of household credit, without eroding lending standards. However the rebalancing also increases mortgage market concentration and risk-taking by smaller banks via competition effects.
Shaping the Future: Policy Shocks and the GDP Growth Distribution (E6, C5)
When central banks lower interest rates or governments raise spending during downturns, they aim to improve the economic outlook and avert the worst possible outcomes. Standard econometric approaches focus on the most likely scenario. We instead study the influence of policy interventions on the entire distribution of future GDP growth, with attention to the lower tail.
Building a new quantile regression-vector autoregression hybrid model (QR-VAR), we investigate the dynamics of the distribution of GDP growth following policy shocks. The VAR structure characterizes economic dynamics over time and allows for a structural meaning of shocks based on traditional sign or zero restrictions. The variable of interest, in our application GDP growth, is modelled through quantile regressions. Here we employ a distribution-free method, allowing for heteroskedasticity, skewness and multi-modality. We project the quantiles one step ahead to simulate a conditional density distribution of GDP growth from which we can re-sample to project the whole distribution of GDP growth multiple periods ahead. We perform simulations under alternative data generating processes to assess the performance of our QR-VAR.
After estimating our QR-VAR on a panel of six countries from 1964Q1 to 2019Q4, we investigate how monetary and fiscal policy shocks shape the future GDP growth distribution. We obtain three key insights: First, monetary policy easing shifts the distribution of future GDP growth up, but without significantly changing its shape. In contrast, fiscal stimulus creates positive skewness by increasing upside risks more than it reduces downside risks. Third, the fiscal effect is stronger when monetary policy is constrained by the zero-lower bound and when government spending originates from the central government.
Our results suggest that fiscal stimulus can shape the distribution of future GDP growth and is well suited to hasten the recovery phase of the COVID-19 pandemic.
Short-Time Work and Precautionary Savings (E2, E5)
In the Covid-19 crisis, most OECD countries use short-time work schemes (subsidized working time reductions) with the aim to preserve employment relationships. This paper studies whether and to what extent short-time work can save jobs through stabilizing aggregate demand in recessions. We build a New Keynesian model with incomplete asset markets and labor market frictions, featuring an endogenous firing as well as a short-time work decision. In recessions, short-time work reduces the unemployment risk of workers, which mitigates their precautionary savings motive. As a result, aggregate demand falls by less. Using a quantitative model analysis, we show that this channel can increase the stabilization potential of short-time work over the business cycle up to 55%, even more when monetary policy is constrained by the zero lower bound. Our results also suggest that an increase of the STW replacement rate may be more effective compared to an increase of the unemployment benefit replacement rate.
Size, Heterogeneity and Distributional Effects of Self-Employment Income Tax Evasion (H2, H3)
We measure tax evasion in Italy by estimating a food expenditure equation that disentangles households with prevalent income from self-employment, which is self-declared, from those with mostly third-party reported income. By using a novel dataset that links the 2013 Italian Household Budget Survey with individual tax records over a period of 7 years, we reduce measurement error by a great extent. We also depart from the usual constant share of underreporting, showing that underreporting heterogeneity among self-employed is significant, and is larger for singles and for college-educated households. We show that self-employed workers in Italy exhibit a similar attitude to tax evasion as those in other developed countries. Therefore, we point to the structure of the economy for an explanation of why aggregate tax evasion in Italy is larger than in other developed countries.
The estimated heterogeneity of underreporting behavior of households combined with the use of a tax-benefit microsimulation model have allowed us to shed light on the distributional effects of income tax evasion, showing that almost 73% of the missing revenue is attributable to taxpayers at the top of the income distribution.
Skill Loss during Unemployment and the Scarring Effects of the COVID-19 Pandemic (E2, I1)
We integrate the SIR epidemiology model into a search and matching framework
with skill loss during unemployment. As infections spread, fewer jobs are created, skills
deteriorate and TFP declines. The equilibrium is not efficient due to infection and skill
composition externalities. Job creation increases infections due to increased interactions
among workers. However, lower job creation decreases TFP due to skill loss. A three-month lockdown causes a 0.56% decline in TFP, i.e. nearly 50% of productivity losses
in past recessions. We study the efficient allocation given the trade-off between both
externalities and show that quantitatively the skill composition externality is sizeable.
Social Comparison, Employee Attrition, and Productivity: Evidence from a Workplace Field Experiment (M5, J6)
Social Comparison among coworkers is prevalent in organizations and often has consequences for employee turnover, productivity, and well-being? Can firms mitigate the costs of performance comparison by redesigning the information environment? In a 7-month randomized control at a leading multi-national spa chain with more than 7,000 workers in China, we sent spa workers twice-weekly messages to inform them of the performance trajectories of their high-performing senior coworkers. This information treatment reduced the overall attrition rate of new workers by 12%, and the effect is most pronounced for high-performing new workers. The lower attrition rate is mostly driven by the improvement in new workers' stress levels and mental health conditions, which in turn results from the lowering of their beliefs about high-performing senior co-workers' past performance. Overall, this study demonstrates that showing junior workers the "resumes" of senior workers mitigates the social comparison costs within firms.
Social Preferences on Networks (D9, C9)
We develop a model of social preferences for network games and study
its predictions in a local public goods game with multiple equilibria.
The key feature is that players’ social preferences are heterogeneous.
This gives room for disagreement between players about the “right”
payoff ordering. When preferences are compatible, however, players
coordinate on a refined equilibrium set. How easily the requirements
for preference compatibility are met crucially depends on a property
of the network structure: neighborhood nestedness. This means that
equilibrium selection succeeds in small, connected structures but also
in centralized networks. All predictions are confirmed in an experiment.
Sovereign Debt Crises and Firm Dynamics (F3, E4)
This paper quantifies the role of entry and exit in propagating the effect of the sovereign default risk on the real economy. Using enterprise-level data from Portugal, we show that the higher sovereign default risk during the European debt crisis (2010-2012) is associated with a substantial decline in firm entry and a rise in exit. Moreover, cohorts of firms exposed to high sovereign default risk consist of fewer firms and employ persistently and significantly fewer workers over their life cycle. We find that the cumulative drop in employment across exposed cohorts is significant and has a long-lasting negative effect on the dynamics of the economic aggregates. Motivated by the empirical facts, we develop a heterogeneous firm dynamics model with endogenous entry and exit, sovereign default risk, and financial frictions. The calibrated model generates a close match to firms' life-cycle dynamics and salient features of the sovereign-bank-firm relationships in Portugal. We find that the sovereign-liquidity channel is one of the main drivers of the observed dynamics in the extensive margin, which, in turn, has a long-lasting negative effect on aggregate employment. We argue that policies that support small and young business operations could reduce the long-run adverse effects of the high sovereign default risk.
Sovereign Debt Repatriation during Crises (F3, F4)
We use a comprehensive dataset on the sovereign debt investor base to document threenovel empirical facts: (i) sovereign debt is repatriated - that is, shifted from external to domestic investors prior to sovereign defaults; (ii) not all crises are equal: evidence for repatriation during banking and currency crises is more limited; and (iii) the nature of debt restructuring matters: external investors do not leave during preemptive defaults. The dataset we use is uniquely suited to analyzing investor base dynamics during rare crises due to its large cross-section and time series, covering 180 countries from 1989 to 2020.
Specification Testing for Prediction (C1, C5)
This paper proposes a prediction-based specification test for parametric regression models. The soundness of regression analysis relies on the model specification's correctness. On the other hand, given a set of regression models, we are also interested in finding a model with a better prediction ability, even if there is no correctly specified model in the set. This paper considers a situation where two stratified datasets are given: one has a covariate and the corresponding outcome, but the other only has a covariate without the corresponding outcome. This paper aims to conduct statistical inference for the unobserved outcome of the latter dataset based on a model estimated from the paired former dataset. We introduce a weak model specification concept, which suffices for prediction tasks. Let us consider the risk of predicting the unobserved outcome. When the risk minimizer is the same between the regressions of two stratified datasets, the proposed specification test does not reject the alternative hypothesis, even if the model is potentially misspecified in the conventional meaning. Our proposed test is not consistent but suffices for prediction task and counterfactual inference of the unobserved outcome. We can apply the proposed method to a wide range of real-world applications, such as the conditional average treatment effect estimation. Through experiments, we investigate the effectiveness of the proposed method.
Spousal Insurance, Precautionary Labor Supply, and the Business Cycle (E2, D1)
I document that married women are less likely to leave the labor force and are more attached to employment in recessions. Using a two-person household incomplete assets markets model with labor market frictions, I show that married women exhibit precautionary labor supply in response to the higher threat of job loss experienced by their husband in recessions. Quantitative analysis shows that married women's precautionary labor supply behavior is an important mechanism of intra-household risk sharing and accounts for 30% of married women's low cyclicality of employment. Furthermore, I show that spousal insurance reduces consumption volatility in married households by 30% over the business cycle.
Staffing Composition and Patient Selection in Nursing Homes: Evidence from Regulation on Nurse Mandatory Overtime (I1, J0)
Due to their uncertainty in future patient volume and health status, nursing homes typically use regular scheduled staff hours combined with overtime hours or third party agency nurse hours during demand peak. This paper studies how nurse staff hours composition affects patient selective admission against Medicaid patients in nursing homes. Leveraging on the state level regulations prohibiting mandatory overtime for nurses as an exogenous shock to nursing homes' ability to use overtime hours, I find that nursing homes reduced total staff hours by 4.5% and the share of Medicaid patients in nursing homes increased by 2% in states that passed regulation. Estimated impact is mainly driven by non-hospital-based nursing homes. The result suggests that targeting overtime hours as an labor input used more in treating Medicare patients can be an effective solution to temper the selection against disadvantaged patients.
State Health Insurance Benefit Mandates and Health Care Affordability (I1, H0)
Every US state requires private health insurers to cover certain conditions, treatments, and providers. These benefit mandates were rare as recently as the 1960’s, but the average state now has more than forty. These mandates are intended to promote the affordability of necessary health care. Our study aims to determine the extent to which benefit mandates succeed at this goal. Using fixed effects and difference-in-difference research designs with data from the restricted Medical Expenditure Panel Survey- Household Component (MEPS-HC), we provide the first empirical estimates of how health insurance benefit mandates affect out-of-pocket costs and total spending on health care. Further, we develop a framework of health care affordability that puts the empirical results into context and allows for an evaluation of how mandates affect the overall affordability of health care.
Stock Market Reactions to Legislated Tax Changes: Evidence from the United States, Germany, and the United Kingdom (E6, G1)
We study the effect of tax policy on stock market returns in the United States, Germany, and the United Kingdom using GARCH models and a unique daily dataset of legislative tax changes during the period 1978 to 2018. We find that days of discretionary tax legislation during all stages of the process often matter for returns, both in terms of statistical significance as well as economic relevance. Further disaggregating the tax shocks shows that news about personal income tax cuts affects stock market returns positively, where-as business tax legislation is rarely influential. We find evidence of stock market spillovers, mainly from US tax changes to European stock markets. In several cases, we measure significant effects of changes in tax legislation on the days the changes are implemented. The US House Committee Report appears to be the most influential legislative stage. During the financial crisis, stock markets were more responsive to tax legislation. Finally, S&P500 returns tend to react at earlier legislative stages than do DAX re-turns, whereas FT30 returns barely react on days of domestic legislative action.
Strategic Trading Around Predictable Liquidation Events: Evidence from Index Reconstitutions (G1, G2)
This study tests the competing theoretical predictions of strategic trading around anticipated liquidation events. Anticipated liquidation events occur in equity markets when stocks are added or deleted from market indexes. This is due to index fund managers rebalancing their portfolios to reflect changes to index constituents. Accordingly, this empirical study considers a sample of stocks added to and deleted from the Australian S&P/ASX 200 market index from 2006 to 2014. We choose this setting as the Australian market for equities represents a well-developed and highly concentrated market that is dominated by a single market index.
The empirical analysis yields two key findings. First, we do not find any evidence of “predatory” strategic trading in the form of significant changes to permanent or temporary price impact costs during liquidation events. This is in contrast to the predatory trading model of Brunnermeier and Pedersen (2005) and consistent with the Bessembinder, Carrion, Tuttle, and Venkataraman (2016) strategic trading model. Second, we find that market resiliency is a key determinant of market quality during index reconstitutions. As predicted by the Bessembinder et al. (2016) model, there is a positive relationship between market resiliency and liquidity provision during predictable liquidation events. We find that a one standard deviation increase to market resiliency prior to index reconstitutions leads to a 5.08 basis point decrease to the relative bid-ask spread on index reconstitution days. This effect is greater at price levels further from the mid-quote. That is, the equivalent figure for the depth-weighted relative bid-ask spread measured across 10 price levels is 37.03 basis points.
Given the strong evidence in favor of the Bessembinder et al. (2016) strategic trading model, this study asserts the need for protocols that promote market resiliency and enhance competition among strategic market participants.
Strategies and Expansion of Intermittent Renewables: Disentangling Pass-Through Costs in Electricity Markets (Q4, L1)
We investigate technology as a source of product differentiation and its impact on strategic behavior and wealth distribution in the German day-ahead market. We compare the performance of our model to a benchmark, using elasticity-adjusted markups and without bid data. We represent uncertainty on the demand side as an intermittency of renewables or a flexible demand response. The results show that both model estimates converge at off-peak hours, being robust to rampingcost and renewable forecast assumptions. Producers pass on fuel and CO2 costs differently with implications for reinforced regulations by the European Emissions Trading Scheme. Consumers are better off under a carbon price floor up toe25/tCO2, but producers are worse off, particularly during the morning peak.
Strengthening Farmer Competences to Increase Climate Resilience of Cocoa Farms: A Randomized Control Trial in Honduras (D9, Q1)
Cocoa cultivation is the income source for more than 350,000 farmers in Latin America; however, this income hardly surpasses the poverty level. Consequently, this lack of income increases farmers' vulnerability to climate change. Cocoa farm management relates to the individual's ability (men or women) to make the right decisions in terms of distribution of shade, fertilization, and pest and disease control. Many agricultural programs focused on providing technical assistance targeting these three factors individually, without having a significant impact on lifting men and women out of poverty; because increasing cacao yields do not automatically translate into higher net income, the farmer must have the competencies to decide what and how to adapt to specific changes. This research project's general objective is to provide scientific evidence on how empowering farmers' competencies reduce their vulnerability to climate change. We build on the concept of `boost' from behavioral economics to strengthen farmers' mental tools to employ new procedures and redesign the tools if necessary. We implement a randomized controlled trial in Honduras, providing heuristic-based training through farmer schools. This research project measure farmers' competencies, adoption, and degree of adoption from different integrated agricultural models which include the attributes of shade, fertilization, pest, disease management, and risk literacy boost. The results of this project proposal will contribute to the literature by providing empirical and scientific evidence and policy recommendations on the design of interventions to lift cocoa farmers out of poverty in Latin America.
Structural Changes in the Job Ladder and the Flattening of the Phillips Curve (E3, C3)
Routine jobs are destroyed every time an economy enters in a recession (Jaimovich and Siu (2020)). This paper shows that such a structural change in the composition of the job ladder –triggered and matured during a crisis– has important implications for inflation dynamics and the Phillips Curve (PC). Using data from the European Monetary Union, we show that countries experiencing a bigger shift in the occupational structure during a downturn exhibit slow inflation recovery and a flatter PC afterward. In particular, we estimate that the shift in the composition of the job ladder during the Great Recession and the following Sovereign Debt Crisis jointly explain 34% of the flattening of the PC between 2002 and 2018. We reconcile this evidence with a New Keynesian DSGE Model accounting for heterogeneity in the labor market. Different wage flexibility is underneath the flattening of the PC generated by both an endogenous and exogenous shift towards non-routine employment during recessions.
Stylized Facts of Consumer Prices From Multi-Channel Retailers in Mexico (E3, E5)
Evidence on price rigidity is used to inform macroeconomic modeling. Using web scraping techniques, this paper characterize the frequency, size and dispersion of price changes as observed on the website of eight large retailers in Mexico, and compare them with price statistics stemming from brick and mortar stores of the same retailers. Strikingly, the evidence suggests that prices observed in brick and mortar stores (offline) change more frequently than those observed on websites (online). However, given a price change, online prices tend to exhibit larger price changes than their offline counterparts. Regarding the distribution of price adjustments, this study shows that online price changes are centered at focal points (multiples of 5%) and report a smaller fraction of tiny price changes relative to the distributions drawn from offline price changes. When focusing on the 2020 data, affected by the Covid-19 pandemic, the results suggest that, for the product categories and retailers in the study, the frequency of price changes increased roughly by the same magnitude in both online and offline sales channels, while the average size of price adjustments do not change relative to previous years.
Sudden Stop with Local Currency Debt (F3, G0)
Over the past two decades, emerging market economies have improved their external liability structures by increasing the share of debt denominated in local currencies, while foreign currency debt is considered a major source of financial instability. This paper embeds the debt denomination choice in a sudden stop model and explore its implications for the optimal capital control policy. As its payoff depends on the real exchange rate, the local currency debt provides better risk-sharing for emerging market economies but introduces additional distortions. Compared to the competitive equilibrium, a discretionary planner has incentives to deflate the debt burden denominated in local currencies, which increases its issuance cost ex ante. In contrast, a social planner with commitment would promise a higher future payment to obtain a more favorable local currency bond price. Quantitatively, the optimal policy under commitment encourages more borrowing in local currencies, mitigates the severity of crises, and improves welfare relative to the laissez-faire.
Switching-Track after the Great Recession (E1, O4)
We propose a theoretical framework to reconcile episodes of V-shaped and L-shaped recovery, encompassing the behaviour of the U.S. economy before and after the Great Recession. In a DSGE model with endogenous growth, negative demand shocks destroy productive capacity, pushing the economy towards a lower trend. A Taylor rule policy designed to reduce the output gap may counterbalance the shocks, preventing the destruction of economic capacity and inducing a V-shaped recovery. However, when shocks are deep and persistent enough, like during the Great Recession, they call for a downward revision of potential output measures, the so-called switching-track, weakening the recovering role of monetary policy and inducing a L-shaped recovery. When calibrated to the U.S. economy, the model generates V-shaped recoveries from small shocks, whilst large and persistent demand shocks lead to a switching-track and affect the supply side permanently.
Synthetic Control Methods for Comparative Case Studies: An Empirical Model for Bangladesh’s Remittance Incentive Program (E6, G2)
This paper explores the impact Bangladesh’s 2% remittance incentive program (the first of its kind) had on formal channel remittances. Globally, remittances are the most significant mechanism of foreign financial inflows with an estimated $550 billion (USD) being sent annually. Globally, remittances dwarf international development aid by three fold and are currently outperforming foreign direct investment as the primary external form of financing. And while the potential positive impact on economic development conditions is vigorously debated by the literature, there is clear empirical support for the impact remittances have on elevating individuals out of poverty. With an approximately 25% poverty rate in Bangladesh, the opportunity to amplify remittances that specifically target Bangladesh’s poor is a compelling opportunity. The 2% incentive program is effectively a 53% reduction in the average 3.75% transaction fee that most Bangladeshi foreign nationals pay. This could translate to anywhere between a 13% and 85% increase in overall remittances - with one outcome nearly doubling overall remittances. Through the use of the synthetic control method, this paper has designed a comparative case study to explore the impact Bangladesh’s remittance incentive program would have compared to a synthetic counterfactual control unit. This paper established a strong synthetic control model composed of Kenya, India, and Cambodia that was able to tightly balance covariates and withstood both placebo and robustness testing during the pretreatment periods. This experimental design is well suited for future analysts to quickly incorporate future remittance and covariate data into the model to study the impact the first remittance incentive policy had on overall remittances.
Tackling the Volatility Paradox: Spillover Persistence and Systemic Risk (G0, E4)
I propose Spillover Persistence as a measure for financial fragility. The volatility paradox predicts that fragility builds up when volatility is low, which challenges existing measures. Spillover Persistence tackles this challenge by exploring a novel dimension of systemic risk: loss dynamics. I document that Spillover Persistence decreases when fragility builds up, during the run-up phase of crises and asset price bubbles, and increases when systemic risk materializes. Variation in financial constraints connects Spillover Persistence to fragility. The results are consistent with the volatility paradox in recent macro-finance models and highlight the usefulness of Spillover Persistence to disentangle fragility from amplification effects.
Tail Risk and Expectations (E7, E0)
We study changes in expectations by incorporating tail risk in a Bayesian learning framework with information frictions. Using a signal extraction problem, we find that economic agents behave differently in the face of tail risk. First, under tail risk, uncertainty shocks lead to a decrease in expectations, which implies more pessimistic forecasts. In comparison, in the absence of tail risks, uncertainty shocks do not influence expectations. Second, we show that individuals overreact under tail risk, that is, individuals are excessively optimistic and pessimistic as compared to a Bayesian learning framework without tail risk. Third, the magnitude of overreaction under tail risk depends on the level of uncertainty in the economy. To validate these theoretical findings, we use quantile regressions to estimate the GDP growth distribution and its measure of tail risk. By comparing the state of the economy in tail risk and non-tail risk episodes, we find that uncertainty shocks lead to a larger fall in expectations when the economy exhibits tail risk. In addition, using an event study approach, we show that there is more overreaction behavior under tail risk episodes. Also, consistent with the theoretical findings, the magnitude of overreaction is larger in periods of higher uncertainty. Our findings seek to shed light on factors driving overreaction in expectations, and to highlight the importance of uncertainty shocks in propagating macroeconomic stability.
Tailored Stories (D9, D8)
Is it possible to persuade others only by providing interpretations of future events? I study the general problem of manipulating a boundedly rational agent by controlling her interpretation of signals she is about to receive. Persuasion arises because the agent updates her beliefs and takes an action based on the narrative she finds most plausible given her prior beliefs. Leveraging this, not only is the persuader able to strategically manipulate the agent to maximize his expected utility, but he can also induce the agent to hold non Bayes-plausible beliefs across signal realizations. Allowing for multiple stories to be communicated before the signal realizes, I propose a disciplined relaxation of the Bayes-plausibility constraint. This paper seeks to provide insights on the mechanism behind incoherent perceptions of the observed facts and the impact of persuasion via storytelling. I illustrate the model with applications in politics, finance, nudging policies, and self-control task.
Talents from Abroad. Foreign Managers and Productivity in the United Kingdom (F2, J6)
In this paper, we test the contribution of foreign management to firms’ competitiveness. We use a novel dataset on the careers of 165,084 managers employed by 13,106 companies in the United Kingdom in 2009-2017. We find that domestic manufacturing firms become on average 4.9% more productive after hiring foreign managers while increasing volumes of activity and capital intensity. In particular, we find that previous industry-specific experience by foreign managers is the primary driver of productivity gains in domestic firms. Foreign-owned firms do not show significant gains from recruiting foreign talents. Our identification strategy combines matching techniques, difference-in-difference, and pre-recruitment trends to challenge reverse causality. Results are robust across different specifications and to sample composition effects. Eventually, our findings pinpoint how limits to the global mobility of managerial talents risk hampering domestic markets’ competitiveness.
Target Inflation and Forward Guidance (E5, E6)
This paper studies the impact of higher inflation targets on the effectiveness and credibility
of forward guidance announcements. Higher inflation rates induce a more forward looking
behavior of firms. This implies that agents "see through" announcements on future interest
rates and the positive current period demand effects they generate reducing the overall effectiveness of forward guidance. This result is robust to expectations processes that are not
subject to the forward guidance puzzle, e.g. inattention, imperfect credibility or learning.
Optimal policy in an environment with positive target inflation holds interest rates lower
for even longer but this becomes less credible the higher the target.
Taxation of Top Incomes and Tax Avoidance (E2, H2)
In the US, the debate on how to tax the rich has highlighted two salient empirical facts. First, there is a high concentration of entrepreneurs with small and medium-sized businesses at the top of the US income distribution. Second, the estimated response of reported income to marginal tax rates is larger for the top 1 % income earners compared to the rest of the population. This difference may be attributed to tax avoidance.
Motivated by these empirical facts, this paper incorporates tax avoidance by entrepreneurs in a quantitative macroeconomic model with heterogeneous agents and occupational choice to study the aggregate and distributional effects of income taxation at the top. Specifically, entrepreneurs can reduce their tax burden in two ways. On the extensive margin, entrepreneurs can choose the legal form of their business organization (sole proprietorships, S- corporations, and C- corporations). On the intensive margin, they can shift their income between different tax bases.
We calibrate the model to the US economy. Importantly, our targets include the elasticities of taxable income with respect to marginal tax rates across income groups and occupations. We find that poor entrepreneurs choose to run their businesses as sole proprietorships to take advantage of the low administrative costs. Wealthier entrepreneurs switch to S-corporations to avoid the social security taxes by disguising their labor income as business income. Because of the double-taxation, only a few entrepreneurs choose to run C-corporations. This is in line with the empirical reported in Smith et al. (2019).
We show that increasing the top marginal income tax rate incentivizes rich entrepreneurs to operate as C-corporations to take advantage of the capped dividend tax rate. While recent literature suggests high taxes on the rich, we show that the presence of tax avoidance significantly reduces the optimal marginal tax rate at the top.
Teacher Flexibility and Scale-up of Education Interventions: Evidence from a Remedial Program in India (I2, O1)
The grade-level learning gap is a persistent problem of Indian education. We use a randomized controlled trial (RCT) to evaluate the impact of remedial teaching on improving the learning of secondary school students in India. We also examine whether granting teachers more flexibility impacts learning improvement. We find that remedial teaching improves student learning significantly. In addition, we find evidence that granting teachers more flexibility to implement the remedial program does not affect the learning improvement of students. These findings suggest that remedial education can be adopted as a meaningful strategy to tackle the problem of the learning gap while the administrative cost can be minimized by granting teachers more flexibility on how to implement the program.
Terrorism and Patriotic Investors: Evidence from the 1920 Wall Street Bombing (G4, N2)
This paper argues that patriotism does not necessarily account for stock market resilience in the face of terror. Studying the 1920 Wall Street bombing, I find that a rallying-around-the-flag effect does not explain the upbeat market reaction at the time. On the contrary, firms with patriotic terms (such as "American'' or "U.S.'') in their name underperformed. Using hand-collected data for 79 NYSE firms, I estimate that abnormal returns for patriotic firms were around one percentage point lower on the day after the bombing. The absence of a clear international actor behind the attack may explain this limited role of patriotism.
Testing Superneutrality When Money Growth Is Endogenous (E0, C3)
This paper develops a model that tests long-run superneutrality of money while controlling for endogenous money growth. We find that an exogenous permanent increase in inflation has a positive long-run effect on output. This finding rejects superneutrality based on Friedman's famous dictum that a permanent movement in "inflation is always and everywhere a monetary phenomenon." We show that our result is robust to a potentially important source of misspecification by extending an existing econometric technique to a novel setting. Using similar methods we show that a simpler vector autoregression model that was popular in previous work is biased against finding evidence for superneutrality because that model fails to account for endogenous money growth. Overall the paper concludes Mundell-Tobin effects seem more prevalent than they used to.
The Aggregate Effect of Granular Shocks: The Role of Bank Specialization (G2, E3)
Are idiosyncratic shocks to one industry transmitted to other industries through common financial intermediaries? In this paper, we find that when banks are exposed to shocks due to their specialization in a certain sector, then they are more likely to increase lending to the treated sector due to higher compensation. In addition, we find that banks cut their lending disproportionally in unrelated and non-affected sectors due to the bank's funding capacity. At the aggregate-industry level, the effects are attenuated during good times as firms raise financing from other banks or from other markets or use internal funds. We do not find these shocks to have any effect on unrelated industries in good times or any associated real effects. However, during bad times, or when financing conditions are tight, these shocks translate into real effects for unrelated industries and thus highlight that idiosyncratic shocks to industries can be passed on to other unrelated industries through common lenders when times are bad and thus amplifying the initial shock. Critically, bank linkage