Members of the NBER's Public Economics Program met April 4-5 in Cambridge. Program Director Raj Chetty of Harvard University, Research Associate John N. Friedman of Brown University, and Faculty Research Fellow Eric Zwick of the University of Chicago organized the meeting. These researchers' papers were presented and discussed:
François Gerard, Columbia University and NBER, and Joana Naritomi, London School of Economics
Job Displacement Insurance and (the Lack of) Consumption-Smoothing
The most common forms of government-mandated job displacement insurance are Severance Pay (SP, lump-sum payments at layoff) and Unemployment Insurance (UI, periodic payments contingent on nonemployment). While there is a vast literature on UI, SP programs have received much less attention, even though they are prevalent across countries and more common than UI schemes in developing countries. In particular, little is known about the insurance value they provide to displaced workers, which relies critically on workers' ability to dissave their lump-sum amount optimally after layoff. Gerard and Naritomi contribute to filling this gap. They follow a standard approach in the UI literature that evaluates workers' need for insurance and the insurance value of various policies by studying workers' ability to smooth consumption after layoff. The researchers take advantage of a rare combination of high-frequency expenditure data and matched employee-employer data for more than 400,000 workers over five years in the state of São Paulo, Brazil. This is a rich empirical setting in which displaced workers are eligible for both UI and SP, with variation in benefits across workers that the researchers can exploit. The researchers find that workers increase consumption at layoff by 35% despite experiencing a long-term consumption loss of 17% when they stop receiving any benefits. They refer to "consumption" because these patterns are robust across expenditure categories, and are not driven by durable goods. The long-term loss is comparable to estimates from other papers that study UI programs, despite the higher labor market informality in the setting. In contrast, the increase in consumption at layoff is unique to this research and the presence of SP. It is larger for workers eligible for higher SP amounts, it is robust across worker groups and realized non-formal-employment durations, but it is not found for workers ineligible for any benefits who experience a sharp and persistent drop in consumption at layoff. Using administrative data on UI payments, the researchers also find that workers spend 20% more in the week they receive their monthly UI paycheck. In addition, they fail to smooth consumption in anticipation of the (expected) drop in income at UI exhaustion, which is associated with a 10% drop in consumption. These results highlight the importance of the difference between SP and UI in their "disbursement" policy -- beyond their different "contingency" policy -- when consumption is highly sensitive to the timing of benefit payment, and shed new light on the need for job displacement insurance in a context of high informality.
Jacob Bastian, University of Chicago, and Maggie R. Jones, U.S. Census Bureau
Do EITC Expansions Pay for Themselves? Effects on Tax Revenue and Public Assistance Spending
Bastian and Jones study how behavioral responses to the Earned Income Tax Credit (EITC) affect the program's budgetary cost. The EITC encourages labor supply and increases income, thereby reducing public assistance payments to households and increasing taxes paid by households. These sources of revenue reduce the EITC's net cost. The researchers use administrative Internal Revenue Service tax data linked to Current Population Survey data on enrollment in public assistance programs to estimate the EITC's net cost. The evidence from three decades of EITC policy expansions implies that the EITC decreases public assistance received by mothers and increases payroll and sales taxes paid. The estimates suggest that the EITC has a self-financing rate of 83 percent, so that the EITC's true cost is only 17 percent of the budgetary cost. Although the EITC is one of the largest and most important public assistance programs in the U.S., the researchers show that the EITC is actually one of the least expensive anti-poverty programs in the U.S., costing taxpayers less than the school lunch and breakfast programs.
Tatiana Homonoff, New York University, and Jason Somerville, Cornell University
Program Recertification Costs: Evidence from SNAP
Homonoff and Somerville document low rates of recertification for the Supplemental Nutritional Assistance Program, which they attribute partly to procedural issues associated with the recertification process. Current recipients, who must complete a recertification interview by the end of their recertification month, are 9 percentage points less likely to recertify when assigned an interview date at the end rather than the beginning of the month. The majority of these case "churn" back on the
program shortly after suggesting that these discontinuations were likely not due to ineligibility. The effects are larger for long-term recipients and cases with children, characteristics associated with higher need.
Victor Stango, University of California, Davis, and Jonathan Zinman, Dartmouth College and NBER
We are all Behavioral, More or Less: Measuring and Using Consumer-Level Behavioral Sufficient Statistics (NBER Working Paper No. 25540)
Can a behavioral sufficient statistic empirically capture cross-consumer variation in behavioral tendencies and help identify whether behavioral biases, taken together, are linked to material consumer welfare losses? Stango and Zinman's answer is yes. They construct simple consumer-level behavioral sufficient statistics -- "B-counts" -- by eliciting seventeen potential sources of behavioral biases per person, in a nationally representative panel, in two separate rounds nearly three years apart. B-counts aggregate information on behavioral biases within-person. Nearly all consumers exhibit multiple biases, in patterns assumed by behavioral sufficient statistic models (a la Chetty), and with substantial variation across people. B-counts are stable within-consumer over time, and that stability helps to address measurement error when using B-counts to model the relationship between biases, decision utility, and experienced utility. Conditional on classical inputs -- risk aversion and patience, life-cycle factors and other demographics, cognitive and non-cognitive skills, and financial resources -- B-counts strongly negatively correlate with both objective and subjective aspects of experienced utility. The results hold in much lower-dimensional models employing "Sparsity B-counts" based on bias subsets (a la Gabaix) and/or fewer covariates, illuminating lower-cost ways to use behavioral sufficient statistics to help capture the combined influence of multiple behavioral biases for a wide range of research questions and applications.
Rebecca Diamond and Petra Persson, Stanford University and NBER; Michael J. Dickstein, New York University and NBER; and Timothy McQuade, Stanford University
Take-Up, Drop-Out, and Spending in ACA Marketplaces (NBER Working Paper No. 24668)
Diamond, Dickstein, McQuade, and Persson study the dynamics of participation and health care consumption in the Affordable Care Act's health insurance marketplaces. Leveraging credit and bank account micro-data, they find half of all new enrollees drop coverage before the year's end. Many of these dropouts re-time health spending to months of insurance coverage, creating large moral hazard effects. This behavior generates a new type of adverse selection: plans selected by dropouts will face high costs, relative to premiums paid. Insurers shift these costs to non-drop-out enrollees, whose inertia generates low price sensitivity. Penalties targeting drop- out consumers could lower premium levels and improve market stability.
Shifrah Aron-Dine, Stanford University; Aditya Aladangady, David Cashin, Wendy Dunn, Laura Feiveson, Paul Lengermann, and Claudia R. Sahm, Federal Reserve Board; and Katherine Richard, University of Michigan
High-frequency Spending Responses to the Earned Income Tax Credit
Aladangady, Aron-Dine, Cashin, Dunn, Feiveson, Lengermann, Richard, and Sahm estimate spending out of federal tax refunds to recipients of the Earned Income Tax Credit (EITC) in the weeks around refund issuance. To do so, they use a new dataset of anonymized daily, state-level spending and plausibly exogenous variation in the timing of refund issuance to this group. The researchers find EITC recipients spend 15 cents out of each refund dollar at retail stores and restaurants within two weeks of issuance, with two thirds of the spending occurring in the week of receipt. The researchers also document an effect on non-durable consumption, as spending at grocery stores and restaurants rises with receipt of the EITC refund. Given that these refunds are large, predictable payments, the results point to excess sensitivity among low-income working families in the U.S. and suggest that alternatives to the lump sum EITC refund payments might better support consumption throughout the year.
Paul Hufe, Ifo Institute for Economic Research; Ravi Kanbur, Cornell University; and Andreas Peichl, University of Munich
Measuring Unfair Inequality: Reconciling Equality of Opportunity and Freedom from Poverty
Rising income inequalities are widely debated in public and academic discourse. Hufe, Kanbur, and Peichl contribute to this debate by proposing a new family of measures of unfair inequality. To do so, they acknowledge that inequality is not bad per se, but that its underlying sources need to be accounted for. Thereby, this research is the first to reconcile two prominent fairness principles, namely equality of opportunity and freedom from poverty, into a joint measure of unfair inequality. Two empirical applications provide important new and interesting insights on the development of inequality both over time (in the U.S.) and across countries (in Europe).
Bruce D. Meyer, University of Chicago and NBER, and Derek Wu and Victoria D. Mooers, University of Chicago
The Use and Misuse of Income Data
and the Rarity of Extreme Poverty in the United States
Recent research suggests that rates of extreme poverty, defined primarily as a person living on less than $2 per day, are high and rising in the United States. Meyer, Wu, and Mooers re-examine the rate of extreme poverty by linking 2011 data from the Survey of Income and Program Participation and Current Population Survey, the sources of recent extreme poverty estimates, to administrative tax and program data. Of the 3.6 million non-homeless households with survey-reported cash income below $2 per person per day, the researchers find that more than 90% are not in extreme poverty once they account for in-kind transfers, errors in earnings reports, errors in transfer reports, and the ownership of substantial assets. More than half of all mis-classified households have incomes from the administrative data above the poverty line, and several of the largest mis-classified groups appear to be at least middle class based on measures of material well-being. In contrast, the households removed from extreme poverty by in-kind transfers appear to be among the most materially deprived Americans. Nearly 80% of all mis-classified households are initially categorized as extreme poor due to errors or omissions in reports of cash income. Of the households remaining in extreme poverty, 90% consist of a single individual. An implication of the low recent level of extreme poverty is that it cannot have risen substantially over time or due to welfare reform.
Alisa Tazhitdinova, University of California, Santa Barbara
Increasing Hours Worked: Moonlighting Responses to a Large Tax Reform
Holding multiple jobs -- or moonlighting -- is increasingly popular in OECD countries, with 5 to 10% of workers holding two or more jobs. Yet little is known about the determinants of moonlighting and its responsiveness to financial incentives: research has been held back by the lack of identifying variation, as most policies treat primary and secondary employments equally. Tazhitdinova circumvents these limitations by studying a unique reform in Germany that allowed workers to hold small secondary jobs tax-free, thus decreasing the tax rate on secondary earnings by between 19.5 to 66pp. Using a difference-in-differences framework, Tazhitdinova documents three findings. First, they find that moonlighting participation elasticities are several times larger than participation elasticities in primary employment. Second, Tazhitdinova shows that individuals do not substitute primary earnings with secondary jobs, despite the large potential savings. Third, the number of low-income jobs increased rapidly after the reform, and did not result in decreased labor supply among low-income individuals. Finally, Tazhitdinova explores mechanisms behind the varying rates of response, and find that hour constraints and job access are key determinants of moonlighting.
Maria Polyakova, Stanford University and NBER, and Stephen P. Ryan, Washington University in St. Louis and NBER
Subsidy Targeting with Market Power
Public welfare programs have a long history of linking their benefits to observable characteristics of potential recipients, such as age, income, health, or employment status. Polyakova and Ryan argue that this common mechanism, tagging, whose goal is to improve efficiency by targeting in-kind transfers to the households with the highest expected utility gain, may lead to substantial market distortions in an environment where the benefit is provided to recipients by imperfectly competitive firms rather than the government. The efficiency losses may be further exacerbated when the level of transfers that subsidize the purchase of the good is anchored to the price information supplied by these firms. The researchers explore this possibility empirically, using data on the health insurance markets created under the Affordable Care Act. They build a model of supply and demand in this new market. Using a novel identification approach coupled with a highly flexible demand specification, they estimate model primitives that allow them to analyze the efficiency of the market. The researchers calculate the incidence of subsidies under the observed subsidy regime with tagging, as well as under counterfactual subsidization mechanisms. They find that a third of the subsidy surplus is captured by producers. Further, tagging subsidies generates a large efficiency-equity trade-off, reducing the overall consumption of the good, but strongly benefiting low income consumers at the expense of higher income consumers.
Marta Murray-Close and Misty L. Heggeness, U.S. Census Bureau
Manning Up and Womaning Down: How Husbands and Wives Report Their Earnings When She Earns More
Do gendered social norms influence survey reports of "objective" economic outcomes? Murray-Close and Heggeness compare the earnings reported for husbands and wives in the Current Population Survey with their "true" earnings from administrative income-tax records. Estimates from OLS regressions show that survey respondents react to violations of the norm that husbands earn more than their wives by inflating their reports of husbands' earnings and deflating their reports of wives' earnings. On average, the gap between a husband's survey and administrative earnings is 2.9 percentage points higher if his wife earns more than he does, and the gap between a wife's survey and administrative earnings in 1.5 percentage points lower if she earns more than her husband does. These findings suggest that gendered social norms can influence survey reports of seemingly objective outcomes and that their impact may be heterogeneous not just between genders but also within gender.
Daniel W. Sacks and Bradley Heim, Indiana University, and Ithai Lurie, Department of the Treasury
Does the Individual Mandate Affect Insurance Coverage? Evidence from the Population of Tax Returns
Sacks, Heim, and Lurie estimate the effect of the ACAs individual mandate penalty on insurance coverage using Regression Discontinuity and regression kink designs and tax return data for the population of single, childless tax filers. The researchers have four key results. First, the penalty paid per uninsured month is less than half the statutory amount. Second, nonetheless, the researchers find visually clear and statistically significant responses to the individual mandate. Coverage rises both in response to extensive margin exposure to the individual mandate and to marginal increases in the mandate penalty. Third, they find substantial heterogeneity in who responds: young people, men, and people without markers of serious health problems are all especially responsiveness. Fourth, the estimates imply fairly small quantitative responses to the individual mandate, especially in the Health Insurance Exchanges.
John N. Tsivanidis, Dartmouth College
The Aggregate and Distributional Effects of Urban Transit Infrastructure: Evidence from Bogotá's TransMilenio
How does public transit infrastructure shape the structure of cities, and how are the gains shared between low- and high-skilled workers? The standard approach in transportation economics measures the benefits in terms of the value of time saved, but new infrastructure can change individuals' decisions of where to live and drive land and labor market adjustment. Tsivanidis develops a quantitative general equilibrium model of a city where low- and high-skill workers with non-homothetic preferences sort over where to live, where to work, and whether or not to own a car. This theory provides a new reduced form framework to evaluate the effects of transit based on "commuter market access". Tsivanidis leverages detailed tract-level data spanning the construction of the world's largest Bus Rapid Transit system -- TransMilenio -- in Bogotá, Colombia, to show this explains the change in city structure in response to the new infrastructure. While the system caused increases in welfare and output larger than its cost, the high-skilled benefit slightly more.