Consumption

Paper Session

Friday, Jan. 6, 2017 7:30 PM – 9:30 PM

Hyatt Regency Chicago, Field
Hosted By: Econometric Society
  • Chair: Marco Di Maggio, Harvard University and NBER

Stock Market Returns and Consumption

Marco Di Maggio
,
Harvard University and NBER
Kaveh Majlesi
,
Lund University
Amir Kermani
,
Massachusetts Institute of Technology

Abstract

Stock Market Returns and Consumption
Marco Di Maggio Columbia Business School Amir Kermani UC Berkeley Kaveh Majlesi Lund University This paper employs unique data containing household-level information on stock holdings and consumption for the entire Swedish population. We find that the marginal propensity to consume (MPC) from stock market returns is about 2% and is constant across the entire wealth distribution, except for investors in the top 10%, who exhibit an MPC close to 0.5%. In contrast, households’ consumption is very responsive to dividend payouts across all parts of the wealth distribution. Specifically, on average households consume more than 50% of dividend payments. Combining these findings with information on the distribution of stocks holdings across households, we show that the total MPC out of stock returns is about 3% ; most of it driven by dividend payments. These results have three broad set of implications: 1.Monetary Policy: Stock returns increases generated by the recent unconventional monetary policies have very limited effects on consumption inasmuch as they do not affect firms’ dividend payments. 2.Inequality. A 5 percent increase in the stock ownership share by the wealthiest 10 percent (and a corresponding reduction for the bottom 60%) reduces the total consumption response to changes in stock returns by about 20%. 3.Consumption Literature. Classical models predict that investor consumption patterns should not be affected by the division of stock returns into dividends and capital gains, while our findings support the mental accounting theory, which suggests that investors have a higher MPC from stock returns in the form of dividends.

Reported Preference Versus Revealed Preference

Jonathan Parker
,
Massachusetts Institute of Technology
Nicholas Souleles
,
University of Pennsylvania

Abstract

The majority of economic research develops, tests, and estimates theoretical models using a revealed preference approach in which causal inference is made from identifying assumptions and observed behavior and outcomes. Increasingly researchers are also using an alternative methodology in which inference is conducted using hypothetical or reported choice behavior, as is more common in other areas of social science. In this paper, we evaluate the consistency of inference based on subjective reported behavior with inference based on observed behavior in a natural field experiment in the domain of consumption smoothing. Specifically, we contrast the extent to which households report that they spent their Economic Stimulus Payments in 2008 with the extent to which the spending behavior of these same households reveals that they spent their payments. This example is of interest both because there has been extensive study of the spending responses to expected fiscal payments using both methodologies and because the answer is a canonical example, pertinent for both the modeling of human economic behavior and the design of optimal policy. We measure the reported-preference spending by surveying household with questions developed by Shapiro and Slemrod (1995, 2003a, 2003b, 2009) to measure spending responses to various tax policies. We measure the revealed-preference spending behavior following previous research using randomization in the timing of the disbursement of the payments and survey information on household spending from both the Consumer Expenditure Survey and the Nielsen Consumer Panel. Despite some potential differences in the meaning of “spending,” the match between the behavior captured by the subjective questions and the behavior identified by the variation in timing are close in both datasets, with one significant exception. We conclude with suggestions for improving the consistency of the subjective analysis.

The Consumption Effect of Positive and Negative Income Shocks

Luigi Pistaferri
,
Stanford University
Dimitris Christelis
,
University of Naples Federico II
Dimitris Georgarakos
,
Deutsche Bundesbank
Tullio Jappelli
,
University of Naples Federico II
Maarten Van Rooij
,
Bank of the Netherlands

Abstract

We use the responses from a representative sample of Dutch households to a survey that asked how much of an unexpected, transitory and positive income change they would consume, and by how much an unexpected, transitory and negative income change would reduce their consumption. The questionnaire distinguished between relatively small income changes (a one-month increase or drop in income), and relatively larger ones (equal to 3 months of income). The results are broadly in line with models of intertemporal choice with precautionary saving, borrowing constraints, and finite horizons.
JEL Classifications
  • E2 - Consumption, Saving, Production, Investment, Labor Markets, and Informal Economy