This page contains current research projects at various stages of development.
Working Papers
This paper examines potential bias in the Census Household Composition Key's (CHCK) probabilistic parent-child linkages. By linking CHCK data to the American Community Survey (ACS), we reveal disparities in parent-child linkages among specific demographic groups and find that characteristics of children that can and cannot be linked to the CHCK vary considerably from the larger population. In particular, we find that children from low-income, less educated households and of Hispanic origin are less likely to be linked to a mother or a father in the CHCK. We also highlight some data considerations when using the CHCK.
Find the working paper here.
Where Are Your Parents? Exploring Potential Bias in Administrative Records on Children
(with Kelsey Drotning and Katie R. Genadek)
Researchers use information about firms from household survey data in a variety of ways. For example, employer characteristics are often used as controls in wage regressions. This relies on a survey respondent to accurately represent the employer they work for. In this paper, we provide information on the degree of mismatch between survey reported job characteristics on the Survey of Income and Program Participation (SIPP) and corresponding information reported by employers in administrative data. We use a fuzzy matching technique that links respondent job data to that of employers in administrative records at the establishment level. The SIPP asks several questions about employer characteristics: whether the employer operated at more than one location, how many employees work there, and the type of industry the employer engages in. We then explore these differences in matching by demographic and socioeconomic characteristics of job holders and discuss possible explanations for the degree of mismatch between household and business data.
How well do you know your job? Differences in employer characteristics between survey and administrative data
(with Ruth Sarafin)
Jobs and Gender: Local Labor Market Outcomes and Gender-Specific Labor Demand
The labor market is characterized by a strong degree of sorting by gender into occupations and industries. Gender sorting implies that men and women are differentially exposed to changes in local labor demand. In this paper, I show that in the U.S. between 1980 and 2017, men have been more exposed to geographically concentrated changes in labor demand than women, and that men are exposed to these changes with higher variance and lower mean. I find that an aggregate labor demand analysis masks important heterogeneity by gender both in exposure and response to changes in gender-specific labor demand. I study differential responses to these shocks by gender, including migration and labor force participation. Migratory responses are greater for men, while labor supply responses are greater for women, and these effects are larger in rural areas. I provide a decomposition of the labor demand shocks to explore mechanisms, finding that industry sectors comprising most of the identifying variation of a shock vary by both gender and region of analysis.
The most recent version can be downloaded here.
Temporal and Spatial Effects of State and Local Taxes on Economic Growth
(with John E. Anderson)
This paper estimates the relationship between the economic growth of states and taxes, modeling both the effects of states own taxes and the spatial spillover effects of taxes in neighboring states on their economic growth. Our research goes beyond the usual temporal tax-growth analysis in the literature to incorporate spatial spillovers. Using data over the period 1999-2013, we analyze the effect of both differences and levels of state and local taxes on gross state product (GSP) growth. Our analysis includes consideration of each of the major state tax revenue sources: income (both personal and corporate), property, and sales taxes. While some previous studies have found strong inverse relationships between state taxes and economic growth, our results indicate that the temporal tax-growth relationship is sensitive to model specification and the time period of analysis. We extend the model to include spatial spillover effects using a spatial Durbin model in order to determine how neighboring states taxes may affect a states economic growth. Our results indicate that negative spillover effects are present in some cases, which we analyze for policy implications.
The most recent version can be downloaded here.
Internal Migration, Wages, and Employment
Young, college educated individuals have the highest rates of migration within the United States. At the same time, overall migration rates have declined, with some studies pointing to labor immobility even when faced with unemployment or poor labor market prospects. The growth of labor market attachment of women also creates an increase in migration ties, which may deter migration. Since location decisions are generally made at the beginning of couples' careers, this could potentially exacerbate the gender pay gap by creating a foundation where spouses willingly locate to an area with fewer opportunities in deference to a partner's labor market prospects. Using linked survey and administrative data, the proposed research will provide a detailed analysis of internal migration and earnings evolution within the United States. The research plans to address the following questions: i) what are the differential returns to migrants’ versus non-migrants’ earnings and future employment? and ii) how does the choice of migration affect earnings paths for individuals as well as spouses, especially after an unemployment spell?