Household Behavior and Public Policy
Saturday, Jan. 5, 2019 10:15 AM - 12:15 PM
- Chair: Charlotte Otabor, District of Columbia Government
Does Affordable Housing Participation Reduce Default and Prepayment? The Case for the Montgomery County MPDU Program
AbstractThe Montgomery County Moderately Priced Dwelling Unit (MPDU) program has served as an affordable housing model to over 500 local jurisdictions in the United States. As the longest running inclusionary zoning policy in the country, the MPDU program has produced over 14,000 owner-occupied and rental units since enacted with over half of these properties sold to low and moderate-income households at below-market price. The units are under price controls with restrictions limiting buyers' potential capital gains from resale. Applicants of the MPDU program must attend a prepurchase certified classroom-based counseling program prior to participation. Using exhaustive data from the MPDU program and loan originations from Fannie Mae between 1995 and 2015, this paper examines the impact of being an MPDU borrower on default propensity and prepayment probability. Using a logit model, I find that MPDU borrowers are conditionally less likely to be ninety-day delinquent compared to non-MPDU borrowers when controlling for zip code of residence and year of purchase. Additionally, I find that the duration of the price control period does not have an effect on default or the likelihood of payoff. However, when I isolate those who prepay, I find that MPDU borrowers have slightly longer loan durations than non-MPDU owners.
Amenity Migration within a Millennial City Evidence from Washington DC Tax Data, 2005-2014
One difficulty in the quantitative study of migration stems from disentangling between economic, non-economic and housing market factors. By focusing on residential migration within a single city, one can minimize this caveat and isolate the impact of income and housing market factors on neighborhood choice. However, due to general equilibrium effects of the migration process on the relative attractiveness of each neighborhood, this solution requires the use of exhaustive data sources.
Using a gravity framework, we estimate patterns of residential migration flows between the 179 census tracts of a very polarized city that has become iconic of gentrification: Washington, DC. To that end, we analyze restricted-access individual income and real property tax rolls from 2005 to 2014, combined with data from the American Community Survey, National Bureau of Economic Research, and the National Neighborhood Indicators Partnership. We propose a measure of residential migration based on income flows across neighborhoods and show that it helps depict urban dynamics in a more complete pattern than classical individual-based flow measures.
Overall, we find that within city migrants, mostly low-to-moderate income households, move to more racially diverse neighborhoods, with cheaper housing, lesser amenities to the exception of EITC recipients, who are less sensitive to differences in neighborhood quality. These results are mostly stable over time, which means that the Great Recession did not alter households’ preferences regarding local amenities in Washington DC.
Payday Lending and the Unbanked Households
AbstractRecently, the media and legislators labelled payday lenders as “predatory lenders”. This is due, in part, to their triple digit, annualized interest rates and business model that traps borrowers in cycles of debt. Despite their costly loans, payday loans are frequently used by the most financially vulnerable populations, low-income and minority consumers with low educational attainment. The unbanked, those without a checking or savings account, are also more likely to be low income, minorities, and people with low educational attainment. Without ties to the conventional financial industry, they must receive financial services from alternative financial institutions, like payday lenders. This research estimates whether there is a relationship between payday loan borrowers and the financially vulnerable, unbanked population. Using the Current Population Survey (CPS) Unbanked/Underbanked Supplement for January 2009, June 2011, and June 2013, this research estimates a recursive bivariate nonlinear probability model. This research shows that the probability a household is unbanked is statistically correlated with the probability that a household uses a payday lender, but unbanked households are less likely to use a payday lender than banked households.
- R2 - Household Analysis
- H2 - Taxation, Subsidies, and Revenue