« Back to Results

Policy and Distributional Impacts

Paper Session

Friday, Jan. 5, 2018 10:15 AM - 12:15 PM

Pennsylvania Convention Center, 106-A
Hosted By: Society of Government Economists
  • Chair: Deirdre Nansen McCloskey, University of Illinois-Chicago

Prejudice in Discretionary Market Transactions: The Case of Markup Disparity in Indirect Auto Lending

Jonathan A. Lanning
U.S. Consumer Financial Protection Bureau


This paper uses a novel data set to explore the relationship between prejudice and
discrimination in the market for auto loans. Indirect auto loans (those a borrower
finances via a dealer rather than directly with a lending institution) are often subject
to discretionary markup, where a dealer increases the rate paid by the borrower in
order to receive additional compensation from the lender. Auto financing is often a
transaction of secondary concern (taking place only after a buyer determines the vehicle
to be purchased and negotiates the price), markups on auto loans are unobservable to
borrowers, and very few borrowers are even aware of the practice. All of this provides
uncommon potential for discriminatory tastes to manifest in market outcomes. Using
attitudinal measures from the General Social Survey I construct indices of prejudice
and examine their relationship with differences in discretionary markup. Using these
indices I am able to explicitly test the Becker model of discrimination by utilizing key
percentiles in the distribution of prejudice and disparities in market outcomes. I am
also able to credibly test for the impact of statistical discrimination by using proxies
for negotiation skill and financial sophistication. Finally, I am able to test predictions
of search models of discrimination separately, and in conjunction with the other models
of discrimination. Ultimately I find that the relationship between prejudice levels and
disparities in markup amounts are highly consistent with the predictions of a standard
Becker model of discrimination, implying that prejudice has an economically meaningful
(and statistically significant) impact on racial gap in markup. Additionally, I find that
neither search nor statistical discrimination appear to be a primary drivers of the racial
disparity observed in this market.

An Evaluation of Optimal Unemployment Insurance Using Two Natural Experiments

Po-Chun Huang
National Chengchi University
Tzu-Ting Yang
Academia Sinica


This paper identifies the liquidity and moral hazard effects of unemployment insurance (UI) using two policy changes in Taiwan: the introduction of a reemployment bonus program and a benefit extension for workers aged at least 45. The reemployment bonus counteracts the moral hazard effect of unemployment insurance without providing additional liquidity during unemployment. The benefits extension, however, increases workers' ability to smooth consumption when unemployed at the cost of distortion to search. Using the variation in the bonus offer brought about by its reach back provision and the age discontinuity in the eligibility for extended benefits, we separately identify the moral hazard effect and the liquidity effect of UI. We estimate that the liquidity effect accounts for 65% of the duration response to extended benefits for workers aged around 45. The moral hazard effect of extended benefits is estimated to decline with age, supporting a longer potential benefit duration for older workers.

The Influence of Pre-existing Conditions on the Risk of Using a Nursing Home

Samuel Tseng
U.S. Department of Labor


Preparing for the financial risk due to utilizing long-term services and supports (LTSS) has become an integral part of retirement planning. Long-term care insurance (LTCI) is an option to finance LTSS expenditures, and its role in retirement planning may grow as Americans are increasingly more responsible for their own retirement security. However, not all people are eligible to purchase LTCI coverage because to deter adverse selection, insurers deny coverage to individuals with pre-existing conditions. This paper estimates that 36% of individuals in their 50s have pre-existing conditions which may prevent them from obtaining LTCI coverage. The results show pre-existing conditions are associated with a 9-percentage-point increase in the lifetime likelihood of using a nursing home. This increase is a lasting effect. The findings can be applied to enhance consumers’ knowledge about the possibility of LTCI coverage rejection, resulting in improved retirement planning.

The Parental Gender Earnings Gap in the United States

Danielle H. Sandler
U.S. Census Bureau
YoonKyung Chung
Robert Graham Center
Barbara Downs
U.S. Census Bureau
Robert Sienkiewicz
U.S. Census Bureau


This paper examines the parental gender earnings gap, the within-couple differences in earnings over time, before and after the birth of a child. The presence and timing of children are important components of the gender wage gap, but there is selection in both decisions. We estimate the earnings gap between male and female spouses over time, which allows us to control for this timing choice as well as other shared external earnings shifters, such as the local labor market. We use Social Security Administration Detail Earnings Records (SSA-DER) data linked to the Survey of Income and Program Participation (SIPP) to examine a panel of earnings from 1978 to 2011 for the individuals in the SIPP sample. Our main results show that the spousal earnings gap doubles between two years before the birth of the first child and the year after that child is born. After the child's first year of life the gap continues to grow for the next five years, but at a much slower rate, then tapers off and even begins to fall once the child reaches school-age.
Robert Adams
Federal Reserve Board
Andrew Shephard
University of Pennsylvania
Ami Ko
Georgetown University
Wenhua Di
Federal Reserve Bank of Dallas
JEL Classifications
  • J0 - General