This paper presents a proposal for learning users’ behavior patterns when they interactively analyse data. Users’ explorations (sequences of queries) are compared looking for subsequences of common actions or operations performed by the users during data analysis. We use a hierarchical clustering algorithm to retrieve groups of similar explorations. The main difficulty is to devise a similarity measure suitable to measure similarities between sequences of human actions. We propose to use a Contextual Edit Distance (CED), a generalization of Edit Distance that manages context-dependent edition costs. CED compares two users’ explorations, making special emphasis in the similarity of queries with nearby queries in the exploration, which determines a local context. We test our approach on three workloads of real users’ explorations, extracting common analysis patterns, both in explorations devised by students and expert analysts. We also experiment on an artificial workload, generated with CubeLoad [19], showing that our approach is able to identify the patterns imposed by the generator. To the best of our knowledge, this is the first attempt to characterize human analysis behavior in workloads of data explorations.

While many have documented the changing demographics of universities, understanding the effects of prohibiting mandatory retirement (“uncapping”) has proved challenging. We digitize detailed directories of all American law school faculty from 1971-2017 and show that uncapping in 1994 had dramatic effects. From 1971 to 1993, the percent of faculty above 70 – when mandatory retirement would typically have been triggered – remained stable at 1%, but starting in 1994, that proportion increased to 14%. We use a permutation test of moving cohorts to show that these increases are attributable to uncapping. Roughly 39% of faculty members would counterfactually have been subject to mandatory retirement. Effects were less pronounced at public schools, which were more likely to have defined benefits retirement plans. Second, we show that schools with the highest proportion of faculty over 70, and thus most impacted by uncapping, also exhibit the slowest integration of female and minority faculty members. Our study highlights cross-cutting effects of civil rights laws: preventing age discrimination can have collateral effects on racial and gender integration.

Can socializing with people who disagree with you reduce political polarization? I answer this question using a shock that induces us to socialize and discuss politics with a more ideologically diverse set of people: Thanksgiving. I use both Canada/U.S. comparisons (exploiting the fact that Thanksgiving occurs at different times in the two countries), as well as a series of Canadian electoral reforms affecting the timing of elections, to show that people converge towards the political viewpoints of their families after Thanksgiving and that this reduces polarization. These results suggest that increasing political diversity within our social circles can reduce polarization.

Previous research on state capacity has largely focused on external and internal elite competition. This paper describes a complementary explanation for differences in state capacity. I argue investments in state capacity entails a commitment problem. To build state capacity, legislators must form and maintain investment coalitions, but these are vulnerable to defections, which can undermine the investments. When legislators defect from investments, state capacity development is held back. I evaluate the theory in the setting of 19th century American state legislatures. Using new data on the curious practice of passing special legislation, I show that when legislators defected, state capacity development suffered. The research further indicates that the ability of legislatures to facilitate state capacity depends on rules and practices governing the behavior and ability of legislators to defect from their commitments.

The first two chapters of the dissertation study how local labor market adjusts to trade shocks. The last chapter explores the relationship between economic condition change and health outcomes. In what follows, I describe my three essays. The first chapter proposes a mechanism through which local labor markets adjust to trade shocks: immigrants’ mobility. I find that immigrants are more responsive than natives to trade shocks. A $1000 increase in the import exposure leads to a 2.6 percent decline in the immigrant population but has little change in the native population. Additionally, immigrant mobility reduces the negative effects of trade shocks on native employment and wages. The study ultimately shows that natives in areas with more immigrants experience smaller declines in employment and wages. The second chapter studies the disparate impacts of trade liberalization on U.S. workers according to gender and age. Focusing on US-China trade shocks that occurred between 1990-2007, I show that these trade shocks generated larger declines in manufacturing employment and wages for older women than for older men. In contrast to prior studies, I find that discrimination and gender differences in industrial employment play relatively small roles in explaining this pattern. Instead, I present evidence that vi women's career interruptions from marriage and motherhood provide a more promising explanation. Within an age cohort, trade shocks depress labor market outcomes more strongly for married women with children than their male counterparts. The last chapter estimates the impact on infant birth outcomes of the farm credit crisis that hit the U.S. Midwest in the 1980s. Exploiting county-level variation in agricultural loans before the crisis, I use a difference-in-differences methodology to show that counties with more pre-existing farmland loans (per acre) experienced relatively worse infant health outcomes as the crisis unfolded. My estimates indicate that a $100 dollar increase in farmland loan (per acre) increased the incidence of low birth weight by around 0.4 percentage points and reduced the birth weight by 19 grams. Other findings show that the credit crisis intensified financial distress and tightened financial constraints for affected households, economic pressures that potentially provide a mechanism for the impact on birth outcomes. Counties that had purchased more farmland prior to the crisis suffered larger declines in their farm earnings, higher delinquency rates, and more bank failures.

While many workers have been sent home amid the COVID-19 pandemic, those in our critical infrastructure industries are more essential than ever. In particular, it is vital that workers involved in the production and distribution of food continue to work to ensure ample food for our population. Here we outline three major concerns that might influence the ability of workers in these industries to continue showing up to work: new childcare obligations amid school closures, pre-existing conditions that put workers at-risk for developing complications if they contract COVID-19, and loss of wages if unable to work. We estimate that nearly 200,000 children in Colorado will require at-home parental supervision and have parents employed in essential critical infrastructure industries; 8,000 of these children have parents in essential food systems industries. We estimate that 157,000 Colorado workers in essential critical businesses are over 60 and have a pre-existing health condition, putting them at high risk of developing complications if they contract COVID-19, and 38,000 of these workers are in food systems industries. We estimate that at risk workers in Colorado’s essential critical industries have average weekly earnings of $1,137 and those within the food system have earnings of nearly $1,490.

We study whether a woman’s labor supply as a young adult is shaped by the work behavior of her adolescent peers’ mothers. Using detailed information on a sample of U.S. teenagers who are followed over time, we find that labor force participation of high school peers’ mothers affects adult women’s labor force participation, above and beyond the effect of their own mothers. The analysis suggests that women who were exposed to a larger number of working mothers during adolescence are less likely to feel that work interferes with family responsibilities. This perception, in turn, is important for whether they work when they have children.

The American Competitiveness in the Twenty-First Century Act of 2000 (AC21) eliminated the H-1B cap for foreign employees of academic, nonprofit and government research organizations. This act potentially affects the job preferences of newly graduated foreign Ph.D. students. Choosing a career in an uncapped H-1B-qualified entity means circumventing the risk of facing the fiercely competitive H-1B application process and possibly avoiding potential losses due to a visa rejection. We use data from the census of Ph.D. graduates to examine the causal effect of this policy change on academic and industry labor markets in the USA. We find that as a result of this policy, Ph.D. graduates with temporary visas are 5 percentage points more likely to pursue a job in academia, and 3–4 percentage points less likely to choose a job in industry. A series of robustness checks exclude other external factors around the same time period driving the results.

This study assesses the relationship between land use, measured as percent zoning designation per square kilometer in a census tract, and homicide counts in Portland, Oregon, while controlling for other neighborhood characteristics. Negative binomial models are implemented to account for the overdispersed homicide count indicator. Results suggest that some land use variables—specifically, mixed-use residential (positive association) and single-family residential (negative association)—have significant predictive value for homicide counts beyond neighborhood characteristics and socioeconomic variables deemed important by criminological theory and research.

Latinxs hold a variety of racial and ethnic identities, they have no homoge­neous culture, and have a variety of different experiences. In Visible Identi­ties, Alco££ (2008) argues that it is a metaphysical and political mistake to think of Latinxs as a monolithic racial or ethnic group that shares a set of intrinsic physical features and a common culture. Aspects of heterogeneity among Latinxs result in a diversity of in-group lived experiences (Haywood; 2017; Torres, 2003 ). Thus, in developing policy and practice to appropri­ately serve Latinx populations, particularly within educational contexts, we must better understand and address heterogeneity. The purpose of this work is to complicate the heterogeneity that exists within the Latinx community, centering around the influences of ethnicity, U.S. geographic differences, and race. We further seek to highlight the im­plications of overlooking Latinx heterogeneity within educational research, policy, and practice, focusing in particular on the higher education context. To do so, we draw connections between research conducted by historians, anthropologists, sociologists, and interdisciplinary scholars focused on the study of race, ethnicity, and Latinx communities. We supplement this ex­isting work with analysis of public data that provides the opportunity for presenting a more nuanced picture of Latinxs.

Historical inequality is difficult to measure, especially at the subcountry level and beyond the top income shares. This article presents new evidence on the level of inequality in Manhattan from 1880 to 1910 using housing rents. Rental prices and characteristics, including geocodable locations, were collected from newspapers and provide extensive geographic coverage of the island, relevant for the overwhelming majority of its population where renting predominated. This provides a measure of consumption inequality at the household level, which helps to develop the picture of urban inequality for this period, when income and wealth measures are scarce. For large American cities, but particularly for New York, housing made up a large share of consumption expenditure and its consumption cannot be substituted, so this is a reliable and feasible way to identify the true trends in urban inequality across space and time.

This paper provides new insights into the relationship between structural change and the fertility transition. We exploit the spread of an agricultural pest in the American South in the 1890s as plausibly exogenous variation in agricultural production to establish a causal link between earnings opportunities in agriculture and fertility. Households staying in agriculture reduced fertility because children are a normal good, while households switching to manufacturing reduced fertility because of the higher opportunity costs of raising children. The lower earnings opportunities in agriculture also decreased the value of child labor, which increased schooling, consistent with a quantity-quality model of fertility.

Information on the available supply of workers in a local job market is important when determining whether there are qualified workers to fill health care jobs in demand. The American Community Survey (ACS), a publicly available annual survey of over 3.5 million households conducted by the U.S. Census Bureau, has been a regular source of information for mapping the geographic distribution of a wide range of occupations, describing the time, duration and distance of workers' commutes, and identifying common forms of transportation for commuting. In this study, we explore what the ACS can tell us about commuting patterns among selected allied health occupations and registered nurses (RNs) as well as how these patterns may inform discussions of health workforce supply. We analyzed 13 occupations, which included 11 allied health occupations and 2 categories of RNs, those with less than a bachelor's degree and those with a bachelor's degree in nursing or higher. We restricted our analysis to individuals 18 years of age and above, currently employed in the United States. We excluded individuals who reported working from home. Our weighted sample represented 9,318,682 individuals. Our key findings are: n Most individuals commuted to a different geographic area from their home. n The average one-way commute time for these occupations ranged from 24.5 to 31.2 minutes compared to the nationwide average for all workers' commute time of 27.0 minutes, and most of these commuters drove alone. n Blacks, Asians/Pacific Islanders, and other non-White races experienced 1 to 2 minutes longer commute on average compared to Whites. On the other hand, Hispanics had about a 45-second shorter commute compared to non-Hispanics. n Public transportation was most commonly used in the Northeast Census region where public transportation is widely available, and it was associated with more minutes spent commuting compared to other forms of transportation. n One additional minute spent commuting by workers was associated with a 0.13% average increase in individual wages. Wage gains associated with spending more time commuting tended to diminish for the longest commutes in four occupations: clinical lab technicians/technologists, home health aides, nursing/psychiatric aides, and dental hygienists. n Compared to driving alone, carpooling, commuting by public transportation, and relying on other transportation lowered wages by 9.8%, 21.4%, and 31.8%, respectively. Our study found that commute patterns among allied health care workers and RNs generally followed national patterns across all occupations including average commute time and common modes of transportation. ACS is a valuable resource that provides detailed commuting information for a wide swath of health care workers although there are limitations and results should be interpreted with some caution. Despite these limitations, a key takeaway from this study is that when estimating supply of health care workers, researchers and workforce planners need to measure local supply based not only on where people report working but also where they live.

Medicaid enrollment increases during economic downturns which imply households using the public health insurance program during coverage gaps due to job loss. However, we provide new evidence demonstrating that the Medicaid program’s countercyclical protections against economic downturns are largely concentrated in states with more generous Medicaid eligibility criteria for adults. We exploit the timing of the 2007-2009 Great Recession to compare trends in recession-linked Medicaid enrollment between states with more generous Medicaid eligibility guidelines and states with more restrictive guidelines. For similar effects of the recession, Medicaid enrollment grew larger states in with more generous Medicaid programs. Our work suggests for every 100 people becoming unemployed in states with a restrictive Medicaid program, about 96 would be uninsured, and about 11 would enroll in Medicaid. Conversely, about 49 would be uninsured in a state with more generous Medicaid guidelines and 57 would enroll in Medicaid.

We provide nationally representative estimates of sexual minority representation in STEM fields by studying 142,641 men and women in same-sex couples from the 2009–2018 American Community Surveys. These data indicate that men in same-sex couples are 12 percentage points less likely to have completed a bachelor’s degree in a STEM field compared to men in different-sex couples. On the other hand, there is no gap observed for women in same-sex couples compared to women in different-sex couples. The STEM degree gap between men in same-sex and different-sex couples is larger than the STEM degree gap between all white and black men but is smaller than the gender gap in STEM degrees. We also document a smaller but statistically significant gap in STEM occupations between men in same-sex and different-sex couples, and we replicate this finding by comparing heterosexual and gay men using independently drawn data from the 2013–2018 National Health Interview Surveys. These differences persist after controlling for demographic characteristics, location, and fertility. Finally, we document that gay male representation in STEM fields (measured using either degrees or occupations) is systematically and positively associated with female representation in those same STEM fields.

There is a dearth of systematic information about the historical New York City housing market. We present a new sample containing rental price and characteristic data for almost 10,000 Manhattan units which was collected from historical newspapers for the period 1880 to 1910. These units were geolocated to the historical map of Manhattan Island to explore their geographic coverage, using Geographic Information System (GIS) software. We use this new sample to plot the evolution of the location and quality of available Manhattan housing units. This complements existing research on the growth of New York City and the evolution of the ethnic composition of neighborhoods across Census years, as we show information at annual frequency during this time of high growth for the city.

Low-skilled immigrants indirectly affect public finances through their effect on native wages & labor supply. We operationalize this general-equilibrium effect in the workhorse labor market model with heterogeneous workers and intensive and extensive labor supply margins. We derive a closed-form expression for this effect in terms of estimable statistics. We extend the analysis to various alternative specifications of the labor market and production that have been emphasized in the immigration literature. Empirical quantifications for the U.S. reveal that the indirect fiscal benefit of one low-skilled immigrant lies between $770 and $2,100 annually. The indirect fiscal benefit may outweigh the negative direct fiscal effect that has previously been documented. This challenges the perception of low-skilled immigration as a fiscal burden.

BACKGROUND Historical change in the availability of kin beyond the household has long interested scholars, but there has been little comparable evidence on long-run change. While generally accepted that individuals lived near kin historically, no systematic measures have been available to assess historical kin propinquity at the national level. METHODS With the release of historical complete count United States census data from the Integrated Public Use Microdata Series (IPUMS), a robust estimate of patrilineal kin propinquity for the United States nationally from 1790 to 1940 is calculated. Defined as the probability of non-random isonymy within an enumeration district, the estimate of patrilineal kin propinquity relies on the sequential ordering of households in the census. RESULTS The United States experienced a long-run decline in patrilineal kin propinquity from nearly 50% of households in 1790 to 17% of households in 1940. The age patterns of kin propinquity show substantial variation across the life course, and regional differences demonstrate the impact of economic and demographic conditions. The decline in kin propinquity reflected urbanization and the decline of agriculture, declining kin availability, growing distance between potential kin links, and a change in preferences of living near kin. CONTRIBUTION This is the first study to produce a systematic estimate of patrilineal kin propinquity at the national level for the United States between 1790 and 1940. Researchers can use this meaningful measure of patrilineal kin propinquity to better explain its relationships with other demographic behaviors and outcomes such as fertility, mortality, and migration choices.

Evidence shows that environmental quality shapes human capital at birth with long-run effects on health and welfare. Do these effects, in turn, affect the economic opportunities of future generations? Using newly linked survey and administrative data, providing more than 150 million parent-child links, we show that regulationinduced improvements in air quality that an individual experienced in the womb increase the likelihood that their children, the second generation, attend college 40-50 years later. Intergenerational transmission appears to arise from greater parental resources and investments, rather than heritable, biological channels. Our findings suggest that within-generation estimates of marginal damages substantially underestimate the total welfare effects of improving environmental quality and point to the empirical relevance of environmental quality as a contributor to economic opportunity in the United States.

Rising profitability and market valuations of US businesses, sluggish wage growth and a declining labor share of income, and reduced unemployment and inflation, have defined the macroeconomic environment of the last generation. This paper offers a unified explanation for these phenomena based on reduced worker power. Using individual, industry, and state-level data, we demonstrate that measures of reduced worker power are associated with lower wage levels, higher pr ofit shares, and reductions in measures of the NAIRU. We argue that the declining worker power hypothesis is more compelling as an explanation for observed changes than increases in firms’ market power, both because it can simultaneously explain a falling labor share and a reduced NAIRU, and because it is more directly supported by the data.