Research

I use quantitative analytic methods and a variety of data sources to understand disruptive and surprising contemporary social phenomena. My research bridges the gap between formal demographic approaches to migration and mobility  and classic sociological topics, including inequality, segregation, residential mobility, urbanism and urbanization, and the influences of social, economic, geographic context.

I am particularly interested in understanding why people move, how they chose among potential destinations, and how that movement (or lack thereof) impacts individuals and places (or not).

American Mobility Decline

My main body of research considers the declining mobility of Americans, particularly since 1950. The reasons behind this decline, as well as its implications for individuals, families, and society remain unclear. My dissertation research clarifies these issues, examining the causes and consequences of mobility decline over the past four decades. Using traditional demographic decomposition analysis of data from the Current Population Survey, I show that nearly half of all mobility decline is attributable to changes in the composition of the American population. In particular, population aging, increasing racial and ethnic diversity, and the increasing prevalence of two-earner households contribute to the observed declines. I also identify substantial group differences in the rates at which mobility has declined. While most accounts stress the universality of the decline, I show that mobility has declined much more rapidly for non-Latino whites, the young, and renters than for other groups. These findings paint a more complicated picture of mobility in the US than has been previously recognized. This work was published in Demographic Research, and I presented this work at the 2016 meeting of the Population Association of America.

For some, declining mobility in the US is evidence of a neutral, if not beneficial, “mobility transition” or a cultural shift toward “rootedness”. My dissertation work suggests that other potentially more detrimental processes are at work. Analyzing data from the Panel Study of Income Dynamics, I show that increasing immobility is more likely the result of Americans being “stuck” in place. Contingent on the expressed expectation to move in the near future, householders are less likely to act on that expectation today than they were in 1970. These findings implicate rising housing costs, stagnant real wages, and other macroeconomic processes in the immobility of Americans. In turn, the inability of households to move when they intend to limits their ability to gain access to higher quality, more affordable, or more integrated neighborhood contexts. This work is currently under peer review, and related coauthored research was presented at the 2015 meeting of the Population Association of America.

My dissertation research also identifies mobility decline as yet another facet of racial and ethnic stratification in the US. Non-Latino whites are advantaged with respect to most residential mobility outcomes, while non-white householders find it difficult to escape impoverished neighborhoods, gain entry to racially-integrated contexts, or to relocate to good school districts. Racial gaps in the increasing tendency to remain “stuck” in place only exacerbate these inequalities. While non-Latino white mobility expectations have declined roughly in-step with observed declines in actual mobility, black mobility expectations have trended upwards despite declines in actual mobility. All Americans are increasingly likely to remain “stuck”, but black householders are much more likely to remain stuck than their white counterparts. This work was published in Sociology of Race and Ethnicity. A related, coauthored analysis examining the link between mobility expectations and mobility outcomes for white and black, as well as Latino and Asian householders is currently under review.

You can read more about this research here.

The Changing Relationship between Housing and Inequality

In work  coauthored with Rachel Garshick Kleit and published in 2015 in Housing Policy Debate, I consider the effects of subprime lending on county-level inequality. The foreclosure crisis and Great Recession have cemented in our minds the harmful effects of subprime and other exotic loan products, but the inherent volatility of these products was not always clear. In fact, as of the early 1990s subprime lending was promoted as an efficient equalizer of wealth, increasing access to mortgage credit and opening the possibility of homeownership to a broader swath of Americans. Using data from the Home Mortgage Disclosure Act from 1990 to 2010, we show that subprime lending did, in particularly unequal contexts with high housing costs, exert an equalizing effect on distributions of wealth and income. However, as lenders developed more predatory lending practices and more exotic, higher-rate, lower-down-payment mortgage products in the 2000s, subprime lending lost its equalizing effect. In addition to publication, this work has been presented at the annual meeting of the American Collegiate Schools of Planning and has been cited in several leading sociology journals.

Using Tax Returns to Study Mobility

In nascent research with Mark Ellis, I use linked, restricted-access IRS 1040 and ACS data to examine and improve migration measurement across Census products. This work is currently exploratory, with the goal of determining whether and to what extent these data are amenable to questions of sociological importance. The sociological value of IRS 1040 data depends in large part on the quality of address information. What share of tax returns are filed from PO Boxes or from preparers’ addresses instead of residence addresses? How do estimates of mobility between geographic units vary depending on the inclusion of these ambiguous filing addresses? To what extent are estimates of mobility pulled from IRS data consistent with comparable ACS estimates, and why do they differ? Most importantly, what populations are systematically missed in tax return data, and how does that affect our measures of mobility?

While questions about the restricted-use IRS 1040 data need to be answered, the potential insight the offer into residential mobility and migration in the US is unparalleled. Tax return data allow an annual, population-level measure of residential that is not subject to recall bias, with the potential to improve non-intuitive estimates of mobility found in the ACS. The 1040 data can also be linked to numerous other Census survey products to allow a longitudinal analysis of individual and family movement over time. As such, the data not only address theoretical questions associated with mobility decline in the US, but also allow insight into short-distance (i.e., intra-neighborhood) mobility, the shifting social and economic determinants of mobility, and the influence of mobility on neighborhoods and labor markets.

The Residential Mobility of Housing Choice Voucher Recipients

The effect of tenant-based housing subsidies has been hotly debated for decades, but the data brought to bear on that debate have been surprisingly limited. With respect to geographic scope, analysis is often limited to Chicago (site of the original Gautreaux Experiment) or a handful of other Moving to Opportunity sites (NY, LA, Boston and Baltimore). With respect to individual-level detail, studies are often constrained to cross-sectional, aggregated analyses (for example, with HUD’s Picture of Subsidized Households).

In collaboration with Kyle Crowder Erin Carll, and Chris Hess, I use restricted-access HUD administrative records on Housing Choice Voucher (HCV) subsidy recipients to provide a longitudinal and national-level look at the effects of tenant-based housing policy.