Public Economics

April 9, 2016
Raj Chetty of Stanford University and Amy Finkelstein of MIT and Nathaniel Hendren of Harvard University and Neale Mahoney of University of Chicago, Organizers

Brian Baugh, Ohio State University

The "Amazon Tax": Empirical Evidence from Amazon and Main Street Retailers

Online retailers have maintained a price advantage over brick-and-mortar retailers since they were not required to collect sales tax. Recently, several states have required that the online retailer Amazon collect sales tax during checkout. Using transaction-level data, Baugh, Ben-David, and Park document that households living in these states reduce Amazon purchases by 8% after sales taxes were implemented, implying an elasticity of –1.1. The effect is more pronounced for large purchases, for which they estimate a reduction of 11% in purchases and an elasticity of –1.5. Studying competitors in the electronics field, the researchers detect substitution of the lost purchases towards competing retailers.

Stefan Pichler, ETH Zurich, and Nicolas R. Ziebarth, Cornell University

The Pros and Cons of Sick Pay Schemes: Testing for Contagious Presenteeism and Shirking Behavior

This paper proposes a test for the existence and degree of contagious presenteeism and negative externalities in sickness insurance schemes. First, Pichler and Ziebarth theoretically decompose moral hazard into shirking and contagious presenteeism behavior and derive testable conditions. Then, the researchers implement the test exploiting German sick pay reforms and administrative industry-level data on certified sick leave by diagnoses. The labor supply adjustment for contagious diseases is significantly smaller than for noncontagious diseases. Lastly, using Google Flu data and the staggered implementation of U.S. sick leave reforms, the authors show that flu rates decrease after employees gain access to paid sick leave.

Jake Mortenson, Georgetown University, Joint Committee on Taxation, and Heidi R. Schramm and Andy Whitten, Joint Committee on Taxation

The Effect of Required Minimum Distribution Rules on Withdrawals from Traditional Individual Retirement Accounts

Traditional Individual Retirement Accounts (IRAs) are a substantial source of retirement savings for current retirees. In 2013, individuals age 60 or older held $3.8 trillion in wealth in IRAs. Under current law, some fraction of these funds must be withdrawn each year beginning the year one turns 70.5 years of age, with the required fraction increasing in age. Mortenson, Schramm, and Whitten study the effects of these Required Minimum Distribution (RMD) rules on the decumulation behavior of retirees using a 16-year panel of administrative tax data. Their data consist of a 5% random sample of individuals age 60 and older from 1999 to 2014, with approximately 2.6 million individuals per year. This period encompasses a unique policy change that the researchers exploit for identification: a one-year suspension of the RMD rules in 2009. Though the RMD rules are modest — leaving one third of the original balance intact by age 90 even if investments generate zero returns — the empirical analysis shows they have large effects on individual behavior. Using a semiparametric technique developed by DiNardo et al. (1996), the authors estimate the counterfactual density of IRA distributions in 2009 that would have prevailed if the rules had not been suspended. They estimate that at least 41% of the individuals subject to the RMD rules would take an IRA distribution less than their required minimum if they were unconstrained. In addition, the authors document an extensive margin effect among individuals newly subject to the rules, and provide suggestive evidence of optimization frictions in retirees’ financial decisions.

Annette Alstadsæter, Norwegian University of Life Sciences; Wojciech Kopczuk, Columbia University and NBER; Martin Jacob, WHU - Otto Beisheim School of Management; and Kjetil Telle, Statistics Norway

Accounting for Business Income in Measuring Top Income Shares: Integrated Accrual Approach Using Individual and Firm Data from Norway

Business income is important in the upper tail of the personal income distribution, but the extent to which it is captured by measures of personal income varies substantially across tax regimes. Using linked individual and firm data from Norway, Alstadsæter, Kopczuk, Jacob, and Telle are able to attribute business income to personal owners as it accrues rather than when it is realized. This adjustment leads to an increase in top income shares, and the size of this effect varies dramatically depending on the tax regime in place. After a tax reform in 2005 that created strong incentives to retain earnings in the business, the increase is massive: accounting for earnings retained in the corporate sector leads to more than doubling of the share of income of top 0.1% in some years. As the result, traditional measures of top income shares become misleadingly low (even when accounting for capital gains). The researchers speculate on the implications of their findings for levels and trends in top income shares observed in other countries. In particular, they note that the major tax reforms of the 1980s in the United States correspond to a shift in the direction of business income being passed through to personal owners, and argue that top income shares constructed using income tax statistics before 1987 are likely to be significantly understated relative to those afterwards.

Alexander M. Gelber, University of California at Berkeley and NBER; Damon Jones, University of Chicago and NBER; Daniel W. Sacks, Indiana University; and Jae Song, Social Security Administration

Estimating Extensive Margin Responses on Kinked Budget Sets: Evidence from the Earnings Test

Gelber, Jones, Sacks, and Song develop a novel methodology for estimating the impact of incentives to remain employed on the employment rate, and they use it to estimate the impact of the Social Security Old Age and Survivors’ Insurance (OASI) Annual Earnings Test (AET) on the elderly employment rate. The AET reduces OASI claimants' current OASI benefits as a proportion of earnings, once a claimant earns in excess of an exempt amount, implying that the losses in current benefits due to the AET grow as earnings increase above the exempt amount. Using a Regression Kink Design and administrative data from the Social Security Administration, the researchers document clear visual and statistical evidence that the probability of working decreases discontinuously faster above the exempt amount than below it, paralleling the reduction in benefits. Under this novel estimation strategy, the authors' preferred estimate of the elasticity of the employment rate with respect to the average net-of-tax rate is 0.59, suggesting that the existence of the AET leads to a substantially lower employment rate among affected older workers.

Eduardo Dàvila, New York University

Using Elasticities to Derive Optimal Bankruptcy Exemptions

In this paper, Dàvila characterizes the optimal bankruptcy exemption for risk averse borrowers who use unsecured contracts but have the possibility of defaulting. It provides a novel general formula for the optimal exemption as a function of a few observable sufficient statistics. Knowledge of borrowers' leverage, interest rate schedule sensitivity to the exemption level, probability of bankruptcy, and change in consumption by bankrupt borrowers is sufficient to determine the optimal exemption. When calibrated to U.S. data, the optimal bankruptcy exemption implied by the model ($100,000) is larger than the average exemption in the U.S. ($70,000), but of the same order of magnitude.

Bradley Heim, Indiana University; Gillian Hunter, Department of the Treasury; Adam Isen, Department of the Treasury ; and Ithai Lurie and Shanthi Ramnath, Department of the Treasury

Income Responses to the Affordable Care Act: Evidence from the Premium Tax Credit

The effect of the Affordable Care Act (ACA) on economic behavior has been the subject of much public debate. In this paper, Heim, Hunter, Isen, Lurie, and Ramnath examine the extent to which taxpayers responded to the marginal income incentives implicit in the Premium Tax Credit subsidy schedule of the ACA for taxpayers who purchase health insurance through the government Marketplaces. Because these sizable subsidies fall to zero when modified adjusted gross income is 400% of the Federal Poverty Level (FPL), a notch is generated for some taxpayers. Using data from tax returns filed in 2013-2014, the researchers find clear evidence of bunching at 400% of FPL, albeit many taxpayers remain in strictly dominated regions on the right of the notch. The observed bunching suggests an income elasticity of 0.5, but the modest amount of overall bunching and larger implied elasticities for those who were eligible for larger subsidies, used paid tax-preparers, and received an advance on their subsidy are consistent with large adjustment costs. The authors further find some evidence that the responses they do observe are driven by avoidance behavior by means of both deductions and labor supply.

Sarena Goodman, Federal Reserve Board of Governors, and Adam Isen, Department of the Treasury

Un-Fortunate Sons: Effects of the Vietnam Draft Lottery on the Next Generation’s Labor Market

Goodman and Isen study how the Vietnam draft lottery affected the next generation’s labor market. Using the universe of federal tax returns, the researchers link fathers from draft cohorts to their sons' outcomes and demonstrate that sons of fathers randomly called by the draft 1) have lower earnings and labor force participation than their peers, and 2) are more likely to volunteer for military service. These findings highlight the strong role family plays in human capital development and occupational choice. More generally, the results provide sound evidence that policies can have intergenerational effects and that altering parents' circumstances can influence children's later-life outcomes.

Sharat Ganapati, Yale University; Joseph S. Shapiro, Yale University and NBER; and Reed Walker, University of California at Berkeley and NBER

Energy Prices, Pass-Through, and Incidence in U.S. Manufacturing

In this paper, Ganapati, Shapiro, and Walker use plant-level production data from a set of U.S. manufacturing industries to study how changes in energy input costs for production are shared between consumers and producers via changes in product prices (i.e., pass-through). The researchers show that in markets characterized by imperfect competition, cost pass-through and two other estimable statistics are sufficient to characterize the relative change in welfare between producers and consumers due to a change in input costs. They find that increases in energy prices lead to higher plant-level marginal costs and output prices but lower markups. Pass-through is therefore incomplete, with estimates centered around 0.75. The authors' confidence intervals reject both zero pass-through and complete pass-through. They find heterogenous incidence of changes in input prices across industries, ranging from consumers bearing a third of the burden in gasoline refining to consumers bearing large majorities of the burden in other industries.

Caroline M. Hoxby, Stanford University and NBER, and George Bulman, University of California at Santa Cruz

The Effects of the Tax Deduction for Postsecondary Tuition: Implications for Structuring Tax-Based Aid

The tax deduction for tuition potentially increases investments in education at minimal administrative cost. Hoxby and Bulman assess whether it actually does this using regression discontinuity on the income cutoffs that govern eligibility. Although many eligible households take the maximum, the researchers find no evidence that it affects attending college, attending full-time, attending four-year college, the resources experienced, the amount paid, or student loans. Their analysis suggests that the deduction's inefficacy may be due to salience, timing, and the method of receipt. The authors argue that the deduction might increase college-going if it were modified in simple ways that would not increase potential costs but would make it more likely to relax liquidity constraints and be perceived as a price change (which it is) as opposed to an income change. They find that households who would be just above a cut-off manage their incomes to fall slightly below it. Such income management generates bias due to reverse causality. The researchers choose optimal "doughnut-holes" that trade-off bias and statistical power.

Eric Bettinger, Stanford University and NBER; Oded Gurantz, Stanford University; Laura Kawano, Department of the Treasury; and Bruce Sacerdote, Dartmouth College and NBER

The Long Run Impacts of Merit Aid: Calculations from California's Cal Grant

Bettinger, Gurantz, Kawano, and Sacerdote examine the impacts of being awarded a Cal Grant, which is among the most generous of U.S. state based merit aid programs for high school graduates. The researchers instrument for Cal Grant receipt using high school GPA and family income cutoffs for eligibility which are time varying and ex ante unknown to applicants. Cal Grant receipt has only modest impacts on the choice of institution and degree completion, with reduced form estimates of four-year attendance, private school attendance, and bachelor's degree completion ranging from zero to four percentage points depending on the population studied. Receiving a Cal Grant significantly reduces total student loans while in school. Although merit-aid programs aim to decrease outmigration of college-educated individuals, Cal Grant receipt does not have statistically significant impacts on the likelihood of living in California at age 30. Measured earnings impacts are imprecise with IV estimates of award utilization ranging from -10 to 15 percentage point increases in earnings.

Sumit Agarwal, National University of Singapore; Gene Amromin, Federal Reserve Bank of Chicago; Souphala Chomsisengphet, OCC; Tomasz Piskorski, Columbia University and NBER; Amit Seru, University of Chicago and NBER; and Vincent Yao, George State

Mortgage Refinancing, Consumer Spending, and Competition: Evidence from the Home Affordable Refinancing Program (NBER Working Paper No. 21512)

Agarwal, Amromin, Chomsisengphet, Piskorski, Seru, and Yao examine the ability of the government to impact mortgage refinancing activity and spur consumption by focusing on the Home Affordable Refinancing Program (HARP). The policy allowed intermediaries to refinance insufficiently collateralized mortgages by extending government credit guarantee on such loans. The researchers use proprietary loan-level panel data from a large market participant with refinancing history and social security number matched consumer credit records of each borrower. A difference-in-difference empirical design based on eligibility requirements of the program reveals a substantial increase in refinancing activity by the program: more than three million eligible borrowers with primarily fixed-rate mortgages – the predominant contract type in the U.S. – refinanced their loans under HARP. Borrowers received a reduction of around 140 basis points in interest rate, on average, due to HARP refinancing, amounting to about $3,500 in annual savings per borrower. There was a significant increase in the durable spending by borrowers after refinancing, with larger increase among more indebted borrowers. Regions more exposed to the program saw a relative increase in non-durable and durable consumer spending, a decline in foreclosure rates, and faster recovery in house prices. A variety of identification strategies reveal that competitive frictions in the refinancing market may have partly hampered the program's impact. On average, these frictions reduced take-up rate among eligible borrowers by 10%-20% and cut interest rate savings by 16-33 basis points, with larger effects among the most indebted borrowers who were the key target of the program. These findings have implications for future policy interventions, pass-through of monetary policy through household balance sheets, and design of the mortgage market.

NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us