The Long-Term Impacts of Rent Control (Job-Market Paper)
Abstract: Rent control is a common policy tool enacted to limit the growth of rents and allow tenants to remain in their homes for longer. Prior empirical research has mainly focused on rent control’s impact on neighborhoods and housing markets while ignoring the potential long-term impacts of rent control for the people directly affected by the policy, particularly children. Using a nearest neighbor Mahalanobis distance matching strategy and publicly available outcome data at the census tract level, I report estimated treatment effects of rent control on average long-term outcomes for children. Consistent with the literature, I first provide evidence that rent control leads to increases in the average housing tenure duration. I also show evidence implying that rent control improves economic mobility for those who receive it while also creating negative spillover effects for those that do not but live in cities with rent control policies. In tracts with a high proportion of rental units, rent control is associated with a $1,300 increase in average tract-level income. Lastly, I report suggestive evidence that rent control has small long-term benefits for children at the bottom of the parent income distribution, but further study is required to validate these results.
Modernizing Person-Level Entity Resolution with Biometrically Linked Records with Michael Mueller-Smith.
Abstract: The increased utilization of administrative data in economics research has led economists to learn and deploy new empirical methods to prepare non-research data for analysis purposes. This paper focuses on record linkage, a common procedure to merge records from separate databases using common identifiers. We propose a supervised learning, record linkage algorithm that is trained using a large, novel dataset that includes biometric identifiers. We show that this large training data substantially improves model performance compared to the smaller training samples frequently reported in the literature. Next, we show that the model’s performance degrades only slightly when attempting to link datasets with different underlying characteristics than the training sample. Lastly, we run simulations to explore how precision and recall, two commonly measured concepts in the matching literature, are directly related to internal and external validity of estimates. We show that the choice of matching algorithm can affect a researcher’s ability to estimate unbiased and statistically precise parameters, reinforcing the importance of choosing the correct record linkage algorithm in applied work.
The Aftermath of Criminal Financial Sanctions with Elizabeth Liu and Michael Mueller-Smith.
The Demography of Rising Wealth Inequality with Fabian Pfeffer and Bob Schoeni. Revise and Resubmit at Demography.
Abstract: The growth of inequality in household wealth over recent decades is well documented. We determine the independent contribution of several demographic trends to rising U.S. wealth inequality over the last three decades. Using data from the Survey of Consumer Finances from 1989 through 2016 and novel decomposition techniques, we show that rapid growth in wealth inequality and increasing wealth concentration at the top coincided with important changes in the demographic composition of the country but that the two are not directly related. However, the shifts in the wealth distribution among demographic groups, in particular the move of households with less education and non-elderly households away from the middle of the distribution, explain much of the observed overall growth in inequality. Part, but not all, of these demographic contributions to rising wealthinequality operate through their contributions to rising income inequality.
The Determinants of Subprime Mortgage Performance Following a Loan Modification with Max Schmeiser (2016). The Journal of Real Estate Finance and Economics. 52(1), 1-27.
Abstract: We examine the evolution of mortgage modification terms obtained by distressed subprime borrowers during the recent housing crisis and the effect of the various types of modifications on the subsequent loan performance. Using the CoreLogic Loan Performance dataset that contains detailed loan level information on mortgages, modification terms, second liens, and home values, we estimate a discrete time proportional hazard model with competing risks to examine the determinants of post-modification mortgage outcomes. We find that principal reductions are particularly effective at improving loan outcomes, as high loan-to-value ratios are the single greatest contributor to re-default and foreclosure. However, any modification that reduces total payment and interest (P&I) reduces the likelihood of subsequent re-default and foreclosure. Modifications that increase the loan principal—primarily through capitalized interest and fees are more likely to fail, even while controlling for changes in P&I.