« Back to Results
Marriott Marquis, Grand Ballroom 10
American Economic Association & Committee on Economic Statistics
Economic Measurement Challenges in the Digital Economy
Saturday, Jan. 4, 2020 10:15 AM - 12:15 PM (PDT)
- Chair: John Haltiwanger, University of Maryland
The Welfare Effects of Social Media
AbstractThe rise of social media has provoked both optimism about potential societal benefits and concern about harms such as addiction, depression, and political polarization. We present a randomized evaluation of the welfare effects of Facebook, focusing on US users in the run-up to the 2018 midterm election. We measured the willingness-to-accept of 2,743 Facebook users to deactivate their Facebook accounts for four weeks, then randomly assigned a subset to actually do so in a way that we verified. Using a suite of outcomes from both surveys and direct measurement, we show that Facebook deactivation (i) reduced online activity, including other social media, while increasing offline activities such as watching TV alone and socializing with family and friends; (ii) reduced both factual news knowledge and political polarization; (iii) increased subjective well-being; and (iv) caused a large persistent reduction in Facebook use after the experiment. Deactivation reduced post-experiment valuations of Facebook, but valuations still imply that Facebook generates substantial consumer surplus.
Innovation-a: What Do IP-intensive Stock Price Indexes Tell Us about Innovation?
AbstractStock prices are a leading indicator of economic activity in the United States, e.g., they are a component of The Conference Board’s U.S. Leading Economic Index. This paper re-examines stock prices and business productivity in light of the growing importance of intangible investment in overall investment in recent decades (Corrado and Hulten, 2010; Haskel and Westlake, 2017). The new information brought to bear in this paper is the M·CAM database (Martin 2004, 2013; Luse and Martin 2014). This database includes traditional full-text patent and other IP data (such as state-granted rights) and includes both explicit citation information together with implicit conceptual association calculated using M·CAM’s proprietary linguistic genomic algorithms that provide estimates of the uniqueness, quality, and value chain associations of patents across companies. The M·CAM database is used to estimate a stock price index of companies determined to have the strongest ties between their holdings of intangible assets and the company’s future profitability. We find that (a) the intangibles-driven stock price index is 10 percent higher than the S&P 500 since its real-time inception in July 2015 and greatly outperforms the S&P 500 over its backcasted history, which extends to July 2007; (b) IP and other innovation assets are essential business assets of over 70 percent of the Standard & Poors 500 and the Russell 1000; and (c) that M·CAM’s sector-level IP data are “value added” indicators of the sector’s technological capability vis a vis “raw” indicators such as WIPO or USPTO patent counts.
Estimating Family Income from Administrative Banking Data: A Machine Learning Approach
AbstractJPMorgan Chase Institute aims to publish insights that are representative of the US population. To do this, we require a method to reweight research based on key characteristics, with income foremost among them. Given that we do not have full coverage of income information across our portfolio of customers, we have developed a proof-of-concept method for estimating income. With that method now in in use by Institute researchers, we seek to refine and expand the original estimate. JPMC Institute Income Estimate (JPMC IIE) version 1.0 uses gradient boosting machines (GBM) to estimate gross family income based on a truth set drawn from credit card and mortgage application data. The estimation relies on administrative banking data – such as checking account inflows – in combination with ZIP code-level characteristics available through public datasets, as well as Census data at the tract level. Deposit account inflows alone are insufficient to approximate gross family income. The combination of administrative banking data with other data sources and a machine learning approach yielded a significantly more accurate prediction of income.
David M. Byrne,
Federal Reserve Board
Massachusetts Institute of Technology
Massachusetts Institute of Technology
Washington Center for Equitable Growth
- O3 - Innovation; Research and Development; Technological Change; Intellectual Property Rights
- C8 - Data Collection and Data Estimation Methodology; Computer Programs