Previous studies have shown that intergenerational socioeconomic association becomes weaker as children's education level increases and is negligible among college graduates. A college degree is known as the great equalizer for intergenerational socioeconomic mobility. Recent studies, however, reported that the strong intergenerational association reemerges among advanced degree holders although it stays weak among BA-only holders. Despite the substantial theoretical importance and policy implications, the mechanisms behind the reemergence of the intergenerational association at the post-baccalaureate level have been less studied. In this paper, we examine the association between parents' education and children's earnings using the 2010, 2013, 2015, and 2017 National Survey of College Graduates data. Our results show that the strong intergenerational socioeconomic immobility among advanced degree holders is fully attributable to three educational sorting mechanisms: children from high-SES families (1) obtain expensive and financially rewarding advanced degrees, (2) attend selective institutions and major in hyper-lucrative fields of study such as law and medicine in graduate school, and (3) complete their education at a younger age and enjoy income growth over more years in the labor market. Implications of these findings are discussed.

A gap in childlessness rates between women with and without tertiary education in low-fertility settings has been well documented by scholars. However, in the United States, high rates of childlessness are declining for women with tertiary education. Will this current trend lead to a closing of the gap in childlessness across educational subgroups in this country? We answer this question using data from the Current Population Survey from 1976 through 2018. We present population-level trends in permanent childlessness by level of education and estimate the differences in the prevalence of childlessness across educational subgroups. Our findings indicate that the rates of childlessness for women aged 40–44 with tertiary education in the United States are the lowest they have been in over three decades and that rates of childlessness are converging among women with secondary and tertiary education. The declines in childlessness rates and the convergence in childlessness rates between women with secondary and tertiary education are observed for all of the three largest race/ethnicity sub-populations of American women: non-Hispanic white, non-Hispanic black, and Hispanic women. This report contributes to the emerging literature on the convergence of childlessness rates across sub-populations of women with different levels of educational attainment, which questions the well-established observation that there is a positive relationship between education and childlessness.

This research estimates a model of own-source revenue diversification and a model of changes in operational spending in municipal governments from 1997 to 2012 to determine how these governments have adapted to the two significant recessions that occurred during this time period. The first model examines factors that affect revenue diversification, focusing on the state–local fiscal context and how the level of urbanization of the area surrounding the municipalities impacts the effect of state–local context and other factors. The second model examines how municipal governments in the United States have adapted their spending to the two severe recessions of the 2000s, focusing on how state context, revenue diversification, and other factors affect changes in operational spending. Finally, this research also looks at the conditional effects of the size of government on the impact of state context, environmental pressures, and revenue structure on changes in operational spending.

This chapter explores the complicated dynamics that are inherent to the practice of data visualization involving issues of race and identity. We focus on data from the US Census and the profound questions that are raised as visual forms purport to represent groups. After reviewing historical context and related limitations and controversies, we present a project that explores a novel approach to visualizing US immigration patterns, an approach that relies on visual metaphors and algorithmic construction of visualization patterns based on massive sampling of Census microdata. The chapter suggests that the use of innovative expressive techniques to convey insights through poetic, and thus less literal, and limiting, forms is a way of grappling with underlying deficiencies in administrative population data.

This paper explores the causal impact of increased maternal education on children’s mortality by investigating the introduction of the 1994 Female Secondary School Stipend Program (FSSSP) in Bangladesh. This program made secondary level education (grades 6-10) free for young girls residing in rural regions of Bangladesh. Eligibility differences in the FSSSP based on area of residence and duration of exposure are two sources of exogenous variation I use to employ difference-in-difference methods. Using data from the Bangladesh Demographic and Health Survey my results indicate higher secondary school level education improves infant and child health, as evidenced by a smaller but statistically significant reduction in the total number of children who died per woman by age 26. This research has important policy implications, as the link between better secondary school education for women and a reduction in children’s death is another reason to provide widely accessible and affordable schooling for young girls in developing countries.

This thesis consists of three chapters. Chapter 1 studies the role of marketing in the economy. Using aggregate and firm-level data, I find that aggregate marketing intensity in the US increased sharply around the mid-1990s, which coincides with a rapid rise of elasticity between firm-level Marketing Produc- tion Cost Ratio and markup. To explain these facts, I develop a model with heterogeneous firms and endogenous markups in which firms engage in market- ing to signal their quality. I use a calibrated version of the model to quantify the impact of information frictions and marketing on aggregate productivity. I find that quality information revealed by marketing is valuable and access to marketing cannot undo the information frictions completely. Chapter 2 examines the impact of zombie firms on resource allocation. Using firm-level data in China, I show that zombie firms are larger, less productive, and receive a higher subsidy rate on average. The difference in average subsidy rate between zombies and non-zombies reflects both the selection criterion of zombies and the underlying joint distribution of subsidy rate and productivity. I develop a model with heterogeneous firms to quantify the impact of zombies on aggregate productivity. Quantitative exercise shows that reducing the dis- persion in subsidy rate across firms can lead to significant productivity gains, while policies that increase the exit rate of zombies have limited productivity effects. Chapter 3 establishes two facts along with the rise of information technol- ogy: (i) the output from the information sector is more intensively used as an intermediate input; (ii) the wage of information workers and their total em- ployment increase relative to those of non-information workers. To understand the causes, we develop a two-sector accounting framework with sector-factor specific technical changes. We find that labor-augmenting technical change is important in explaining the observed change in wage premium and intermedi- ate shares.

How do politics influence the geographic distribution of science funding? I investigate this question in the context of presidential politics. Science policy scholars endeavor to develop a systems-level understanding—using empirical data and quantitative analysis—of how governments make decisions about science. In the United States, one of the most important decisions that governments make is the allocation of federal funding from agencies such as the National Institutes of Health and National Science Foundation to researchers and universities. Science policy scholars typically explain the distribution of science funding through scientists’ or universities’ merit. I challenge these explanations’ assumption that presidential politics does not play a role. I use the theory of presidential particularism to examine the role that presidential politics plays in the distribution of science funding. The allocation of science funding is a . . .

While previous research often finds flood impacts outside of conventional flood risk zones such as FEMA's 100-year floodplain maps, we have less of a sense of the social and demographic composition of the areas outside of floodplains that experience these impacts, even though social inequalities in flood risk and impacts more broadly is well-documented in the United States. Using data on 100-year floodplains, flood impacts, socio-demographic characteristics, and residential parcels, this study focuses on race as a primary marker of socio-spatial inequality to examine flooding inside and outside of floodplains during Hurricane Harvey in Greater Houston. Descriptive findings show that a large majority of flooding occurred outside of 100-year floodplains. Regression models show that while there is limited evidence of racial inequalities in flood risk as conceptualized as location in 100-year floodplains, there are substantial racial inequalities in flood extent during Hurricane Harvey. Results further show that these overall racial inequalities in flood extent are primarily driven by impacts that occurred outside of 100-year floodplains. Conclusions center on how and why conventional delineations of flood risk can underestimate racial inequalities to natural hazards.

In the 1940 census, white males earned twice as much as black males. But white males born in 1900 had lifetime earnings 3.4 times as large as black males born in 1900. The black-white lifetime earnings gap is so much larger than the cross-sectional earnings gap because of the large black-white gap in life expectancy. 48% of black males born in 1900 died before the age of 30 as compared to just 26% of white males. Economists often use cross-sectional earnings gaps to measure inequality between groups, but a more complete measure of inequality combines income profiles and mortality risk. I develop a model of optimal consumption in a world with mortality and I calibrate the model separately for each cohort of black and white males born between 1900 and 1970. Using this model, I find that the black-white welfare gap shrank by 48% from the 1900 to the 1920 birth cohorts as black and white mortality rates converged, but the black-white welfare gap declined only modestly from the 1920 to 1970 birth cohorts.

The study of intergenerational mobility was once viewed as a quintessentially sociological topic that was widely investigated using occupational mobility tables. However, the popularity of that approach has been dwindling. This decline is associated with the increasing use of the economics model which is not encumbered by the shortcomings of occupational mobility tables. The first limitation of the latter is the contextual nature of occupation which provides an imprecise indicator of an individual's earnings. The second limitation is the focus on cross-sectional data in an era of increased labor market volatility. The third limitation is the dubious practice of distinguishing between structural mobility and circulation mobility. The fourth limitation is the failure of occupational studies to discern important empirical trends about rising inequality. The fifth limitation is that occupation is an inaccurate indicator of non-pecuniary rewards for individual jobs. The recognition of these limitations helps to explain why sociologists are abandoning occupational mobility tables—despite their once great popularity—in favor of the economics approach.

Much of current research on crime and education has focused on the effect of minimum dropout age on rates of crime. Combining the FBI’s uniform crime reporting database and district finance data, we study the longitudinal relationship between crime in every town/city (whose police department has reported crime statistics) and its school district spending in years 2003 to 2018. We combine over 213 datasets to control for population, density, wealth, education, employment, cost-of-living, race, law enforcement, and voting history. Additionally, we also look at teacher salary, teacher engagement, and student chronic absenteeism. Using linear mixed-effect modeling, we find an overall average of 2.35% percent decrease in property crime for every $1000 more a school district spends per pupil on education. Moreover, a $1000 increase in education spending decreased property crime nearly four times as much as a 10 percent increase in per capita income. We also looked at the range in district spending in towns/cities and counties whose students attend multiple districts. We find that for every $1000 difference in district spending within a city, property crime increases by an average of 3%; interestingly, violent crime decreases by 3%. When we lag variables of education quality, allowing these effects to playout, we also find that for every 10 percentage-point increase in chronic absenteeism among students, violent crime increases by 4%. Importantly, we find no such effect for property crime, suggesting a distinct mechanism of education on violent crime. Additionally, both law enforcement and unemployment explain little variance in crime. Our results demonstrate a robust relationship between education funding reduced crime across America with regard to amount spent per student as well as equity in spending.

We examine change in multiple indicators of gender inequality for the period of 1970 to 2018 for the United States, and post-1990 data on some of those indicators for the Republic of Ireland. We consider gender inequality and its trend over time in educational attainment, employment, fields of study in higher education, occupations, and earnings. We conclude that there has been dramatic progress in movement toward gender equality, but, in recent decades, change has slowed, and, on some indicators, stalled entirely. The slowdown on some indicators and stall on others suggests that further movement toward gender equality will only occur if there is substantial institutional and cultural change, such as an increase in men's participation in household and care work, governmental provision of childcare, and adoption by employers of policies that reduce gender discrimination and help both men and women combine jobs with family care responsibilities.

Inequality in the United States continues to increase in terms of earnings and household income (Semega, Fontenot, & Kollar, 2017). As is well known, this trend began several decades ago (Piketty, 2015; Semega, Fontenot, & Kollar, 2017). Compared to other developed nations, inequality in the United States is generally regarded as being among the highest (Brandolini & Smeeding, 2009). Inequality has become so high that it may be exacerbating social problems (Wilkinson & Pickett, 2009: Stiglitz, 2015). In part due to such concerns, researchers have sought to understand the sources of rising inequality. One potential source that has often been suggested is change in the job structure (Autor, 2014). If the distribution of jobs is becoming more unequal in regard to their typical skill levels and wages, then that trend might be an underlying cause of increasing income inequality. Some research has argued that the job structure is becoming more bifurcated and that this polarization is leading to a more unequal distribution of wages (Autor, Katz, & Kearney, 2006). In the following, we further investigate this issue.

In the U.S., households with less than one car per driver (auto-deficit households) are more than twice as common as zero-vehicle households. Yet we know very little about these households and their travel behavior. In this study, therefore, we examine whether car deficits, like carlessness, are largely a result of financial constraint or of other factors such as built environment characteristics, household structure, or household resources. We then analyze the mobility outcomes of car-deficit households compared to the severely restricted mobility of carless households and the largely uninhibited movement of fully-equipped households, households with at least one car per driver. Data from the California Household Travel Survey show that car-deficit households are different than fully-equipped households. They have different household characteristics, travel less, and are more likely to use public transit. While many auto-deficit households have incomes that presumably enable them to successfully manage with fewer cars than adults, low-income auto-deficit households are—by definition—income constrained. Our analysis suggests that low-income car-deficit households manage their travel needs by carefully negotiating the use of household vehicles. In so doing, they travel far more than carless households and use their household vehicles almost as much as low-income households with at least one car per driver. These results suggest that the mobility benefits of having at least one car per driver are more limited than we had anticipated. Results also indicate the importance of transportation and employment programs to ease the potential difficulties associated with sharing cars among household drivers.

The Bureau of Labor Statistics (BLS) of the U.S. Department of Labor is the principal Federal statistical agency responsible for measuring labor market activity, working conditions, and price changes in the economy. Its mission is to collect, analyze, and disseminate essential economic information to support public and private decision-making.1 Private and public decisions related to labor markets and working conditions are increasingly being influenced by technological considerations. Spurred by a wave of technological developments related to digitization, artificial intelligence (AI), and automation, governments around the world have declared that the creation and deployment of these technologies present both important opportunities and challenges to their citizens. Having better data related to the labor market and automation technologies could go a long way in helping address the concerns raised by technology. With these issues in the background, the BLS commissioned this report to identify constructs that would complement existing BLS products with a goal of ensuring that the necessary data exist that would allow stakeholders to assess the impact of automation on labor outcomes.

In this data essay, we analyzed the national and state-level shifts in racial and ethnic makeup of the United States electorate from 2000 to 2018. The term “eligible voters” refer to persons ages 18 and older who are U.S. citizens. All references to Asian, Black and White adults are single-race and refer to the non-Hispanic components of those populations. Hispanics are of any race. This data essay’s analysis is based on Pew Research Center tabulations derived from the following U.S. Census Bureau data: the American Community Survey (2018 and 2010), the 2000 U.S. decennial census, and the November Voting and Registration Supplement of the Current Population Survey (2000, 2004, 2008, 2012 and 2016).

This dissertation consists of two chapters. Chapter 1 This study estimates the causal effect of access to paid sick leave on worker mobility, by exploiting variation in the implementation of local paid sick leave mandates over time in the U.S. I use May 2004 - June 2019 Current Population Survey (CPS) basic monthly data, and by taking a Difference-in-Differences approach, I find that the local mandates significantly reduce private sector employees’ monthly job turnover. This study is, to the best of my knowledge, the first to present the effect of local paid sick leave mandates in the U.S. on worker mobility. Chapter 2 This study estimates the causal effect of access to paid sick leave at work on fertility. Using data from 2006 – 2018 American Community Survey (ACS), and the Synthetic Control method, I show that following the implementation of paid sick leave mandates in Connecticut, Washington DC and Seattle, there is a similar pattern: fertility rate sharply declines in the first few years, and then bounces back. Women are likely to postpone childbirth until they have accrued a certain amount of paid sick leave, in the absence of paid family leave or maternity leave. Access to paid sick leave also helps female workers balance career and family after childbirth. In addition, parents can take paid sick days to look after ill children under the paid sick leave mandates. Local paid sick leave mandate in the U.S. were adopted rather recently, and the number of sick leave mandates increased very quickly since 2014. Little has been known about their impacts on childbirth. This study is, to the best of my knowledge, the first to examine the impacts of local paid sick leave mandates in the U.S. on fertility

Background: Although health insurance coverage for adults in each of the Gulf Coast States and the rest of the country increased after implementing the Affordable Care Act, the coverage rates in the Gulf Coast region remained lower to rural residents, compared to those in the rest of the Nation. Purpose: This study aimed to update the changes of health insurance coverage in all states and the Gulf Coast states, confirm the significance of the health policy on insurance coverage by analyzing Louisiana, and examine the relationships between socio-demographic variables and rural/urban area by using interaction variables. Methods: This study used the American Community Survey, which is an annual survey of about three million U.S. households and collected social, demographic, and economic information, including health insurance coverage. Logistic regression was used to estimate the effects of the demographic and economic variables on health insurance coverage. Results: Florida and Texas increased health insurance coverage in the urban areas, while Alabama, Louisiana, and Mississippi present a more considerable increase in the rural area. However, Louisiana showed a significant increase in insurance coverage, rural areas in particular after joining the Medicaid expansion in 2016. A significant decrease in insurance coverage was found among young adults, African American, non-married, not in the labor force, and being poor for rural residents in Florida and Texas. In contrast, minorities in other races and unemployed decreased the likelihood of having insurance for rural residents in Alabama, Louisiana, and Mississippi. Discussion: Our examination of how socio-demographic variables interact with living in a rural area revealed a clear rural disadvantage pattern. The pattern, however, was varied between Florida and Texas and Alabama, Louisiana, and Mississippi. These findings have meaningful implications for the ongoing effort to reduce insurance coverage disparities in the Gulf Coast states and all Americans.

The county scale has thus far dominated rural demographic research—this descriptive profile of small town America is unique with its place-based lens. Another important extension is the nationwide application of the Community Capitals Framework which builds on the body of research examining capitals within case studies focused on one or more communities. Here, we examine place-based “community capitals” at the national scale through novel integration of data from a wide variety of sources. The goal is to identify tiny town socioeconomic and demographic patterns of change—or trajectories—over the past several decades—and contrast remote small towns with those proximate to metropolitan areas. Results reveal both commonalities and distinctions. Instead of differences in trends across time, the analyses suggest that what differs are the more general profiles of small places as contrasted with national data. For example, regardless of metro proximity, small town America has lower levels of human and financial capital. Still, distinction also appears in that rural population growth has focused on high-amenity regions, bringing some increases in community capitals but potentially also exacerbating inequalities. In all, the analyses presented here offer an important foundation for necessary work at the place scale to improve understanding of the nuances inherent in population shifts, and their implications, within rural communities.

This dissertation examines influences of gender essentialism on occupational mobility patterns, underlying occupational sex segregation in the contemporary United States (2011- 2015). Gender essentialism—the belief that men and women have fundamentally different skills, interests, and capacities—leads to gender-typed skills in the context of work: skills that are viewed as feminine (e.g., working with people) or masculine (e.g., working with machines). I examine the influence of requirements for these skills on workers’ occupational mobility, and in particular placement into sex-typical occupations. Previous studies have considered only the macro-level (gender essentialist forms of occupational sex segregation), or the micro-level (case studies of workers’ career decisions and work experiences). This dissertation addresses an intermediate gap: meso-level analysis linking essentialist structures of occupational sex segregation to individuals’ occupational mobility patterns. I employ conditional logit models (CLMs) to represent workers’ occupational movements in terms of occupational characteristics, rather than privileging workers’ individual characteristics. Privileging occupational characteristics (i.e., gender-typed skills) highlights their influence on workers’ probability of occupational placement. CLMs make pairwise comparisons between workers’ occupational destinations and their “alternatives”, i.e., other occupations they can reasonably access. CLMs are not widely used in studies of occupational sex segregation, and this dissertation builds on initial efforts: in particular, I make more realistic assumptions about the alternative occupations available to workers. I use data spanning 2011-2015 (inclusive), from two sources: the Annual Economic and Social Supplement to the March Current Population Survey, and the O*NET database. The former provides individual-level data on year-to-year occupational mobility; the latter provides occupational characteristics. Chapter 1 surveys literature on gender essentialism and occupational mobility, my own methodological approach, and the questions motivating each empirical chapter. In Chapter 2, I evaluate a hypothesis explaining why women with Bachelor’s degrees are less wellrepresented in female-dominated occupations (those where a majority of workers are female): feminine skills have a weaker influence on placing them there. I use CLMs to compare the influence of gender-typed skills on placement probabilities in sex-typical occupations (those where a majority of workers share the focal worker’s own sex), for women with and without Bachelor’s degrees, relative to their occupational alternatives. In Chapter 3, I test the hypothesis that gendered work rewards help place workers in sex-typical occupations, relative to their available sex-atypical alternatives. I examine influences of gendered work rewards on men’s and women’s probabilities of placement in sex-typical occupations, relative to their sex-atypical occupational alternatives. In Chapter 4, I test two hypotheses explaining why requirements for physical strength—a masculine skill—increase women’s placement probability in Professional occupations. Within Professional occupations, I analyze the influence of wages on workers’ placement probability, and examine the joint distribution of feminine skills and selected masculine skills. Chapter 5 discusses the findings of Chapters 2-4, limitations of CLMs and O*NET data, and further applications of CLMs, for future research on occupational sex segregation. I contribute to the literature on occupational sex segregation by demonstrating previously unexamined ways in which gender essentialism strongly influences workers’ placement in sex-typical occupations. This influence is heterogeneous across different groups of workers, e.g., by Bachelor’s degree attainment, occupational category, and sex. Chapters 2 and 3 suggest that skill development and work rewards play important roles in workers’ sex-typical occupational placement, and Chapter 4 suggests that gender-typed skills are important even in workers’ gender-atypical occupational placement.