Variations in Perinatal Outcomes in Low-Income Census Tracts: An Analysis of Low-Income Housing Tax Credit Unit Receipt

Abstract

Introduction: The social determinants of health (SDOH) are the characteristics of a person?s physical and social environment that impact their health outcomes. Housing quality is a key aspect of the SDOH and is an important health determinant among low-income communities. The Low-income Housing Tax Credit (LIHTC) program, created in 1986, is the largest financial supporter of new and rehabbed low-income housing. LIHTC units are built to state-specific quality standards that have increasingly prioritized safeguarding occupant health. Despite the LIHTC program?s role in affordable housing, research on LIHTC?s impacts on resident health outcomes has been minimal. However, the unique role census tracts play in the administration of the LIHTC program provided a promising way to examine its health impacts.Objective: The paper focused on perinatal outcomes because health outcomes influenced by the environmental effects of LIHTC units would have been difficult to observe in older residents without extensive medical histories. This dissertation aims to answer whether, after controlling for tract socio-economic and demographic variables, does the number of LIHTC units constructed within a census tract relative to its population impact the low birth weight rates and preterm delivery rates of that census tract. Method: New York City census tract level birth data from 2000 to 2004 and 2010 to 2014 were matched with HUD LIHTC housing data. Tract characteristics from the same period were added to control for tract socioeconomic differences. A total of 1,774 tracts comprised the final sample, from which three additional analytical tract sub samples were created based on tract characteristics. I performed a cross sectional analysis on the full sample (n=1,774), sub sample two (n= 632) and sub sample three (n= 230). To address possible omitted variable bias, first differenced and fixed effects estimations were conducted on the full sample, sub sample two and sub sample four (n=402). Findings: Overall, my analysis indicated that the number of LIHTC units in a census tract relative to its population had no statistically significant effect on tract low birth weight or preterm delivery rates.