This project studies discrimination in online retail grocery stores. Do different consumers get charged a different price based on their perceived race? To answer this question, the author will implement a massive data collection exercise using web scraping. The data will combine firm-level data on products and prices with geography and aggregate socioeconomic indicators. This provides information on the underlying consumers in those areas. In the first phase of the project, the shopper’s race is based solely on geography. Future phases of the project will attempt to use browsing history to find racial differences in the shoppers. Web crawlers will be used to collect prices based on location to understand how online prices faced by consumers vary across socioeconomic and racial groups (imputed based on location). This research will identify whether online shopping allows retailers to price discriminate in ways that are harder to do in person.

This proposal will utilize a large and comprehensive dataset to evaluate whether historically Black colleges and universities, or HBCUs, can narrow or close racial gaps on numerous measures of economic well-being, not just typical measures such as income. The dataset links College Board SAT data with credit bureau data and National Student Clearinghouse data. It includes students who took the SATs between 2004 and 2010, tracking them from high school through college, and will allow the authors to look at financial outcomes at age 30. Using these data, the authors will compare the longer-term outcomes for Black students who attend these schools versus similar students who applied to but did not attend one. The authors will explore several outcomes, including those where racial disparities exist, such as college-related debt, other forms of debt, and whether the individual has a mortgage (a proxy for homeownership). The credit data also give a more complete picture of income than earnings since it covers all types of income. Existing research shows how important social supports and social capital are to economic mobility. This project will shed light on the distinctive social and psychological value-added features of historically Black colleges and universities.

This project explores how low-income people with immediate needs for cash make borrowing decisions in states where payday lending is heavily restricted versus states where it is not. It takes a qualitative approach to exploring the experiential processes that unfold across varying state policy contexts. As the author notes, there is a burgeoning line of scholarship on payday loans and states’ attempts to restrict them, but with mixed evidence on the effects on low-income borrowers. On one hand, these loans come with predatory lending rates that are often compounded for borrowers who are unable to pay back the loan in the original period and therefore roll it over, incurring more fees and often resulting in the borrower owing many times over what they originally received. On the other hand, credit is highly constrained for low-income individuals, with payday loans filling the gap. Yet there remains neither a consensus on the utility of such loans for low-income borrowers nor an understanding of how low-income individuals make decisions about borrowing. This gap limits policymakers from addressing the dual needs of credit access for low-income borrowers and the need to reduce the deleterious effects of payday lending, a gap this research will shed light on.

This project extends ongoing work by this research team on nursing home unionization and COVID-19 preparedness. The interdisciplinary team takes a mixed-methods approach to estimate the causal link between collective bargaining and workplace safety prior to and during the pandemic. The researchers have built a proprietary dataset of union status of all 15,000 nursing homes in the United States and merged it with publicly available Center for Medicare and Medicaid Services workplace-level data on COVID-19 outcomes in nursing homes. They will use this dataset to examine the impact of unionization on health and injury outcomes (and racial differences) before and during COVID-19 in the 2016–2021 period. Using an event-study difference-in-difference framework will allow them to capture the validity of parallel trends assumptions, and the proposed regression discontinuity design will help to establish the causal effect of unionization on worker COVID-19 infection outcomes. This is an important question that has significant resonance right now, as we are seeing a resurgence of union activity among major employers.

This project examines associations between municipality of residence during childhood and upward mobility. Notably, the project creates a new dataset by identifying municipalities across the United States and documenting and categorizing municipal policies for comparison. Research on municipalities is hampered because a single repository or dataset containing all municipalities and their characteristics and policies does not exist. In addition to the potential data contribution, from a policy solution standpoint, understanding municipal policy is critically important. It is neither practical nor reasonable to propose solutions for mobility that operate just at the neighborhood or commuting-zone level, outside of the context of local governance. City and county governments need to know what they can reasonably do within their jurisdictions in order to increase mobility. While the proposed study, like many others in this space, does not attempt to identify causality, the descriptive work has the potential to be telling since it could provide municipalities with evidence of how they are succeeding or failing at supporting upward mobility for their residents.

This project explores the question of how policy shocks propagate through the economy. The researchers will build a large dataset using Danish Bank and Danish government administrative data to build matrices of income, consumption, and production in different regions and sectors, as well as how income, consumption, and production in different regions and sectors are interconnected. This “disaggregated economic account” will be used to trace out how a shock that hits one part of the economy propagates to other parts and determine the aggregate impact of the shock. The resulting model will help define optimal policy responses to different kinds of shocks, to measure how certain shocks affect income inequality and growth, and to identify the most important channels that propagate shocks. Tracing how a shock in one sector filters through the economy is only possible with this type of administrative data linked in this way.