Survival to extreme ages clusters within families. However, identifying genetic loci conferring longevity and low morbidity in such longevous families is challenging. There is debate con- cerning the survival percentile that best isolates the genetic component in longevity. Here, we use three-generational mortality data from two large datasets, UPDB (US) and LINKS (Netherlands). We study 20,360 unselected families containing index persons, their parents, siblings, spouses, and children, comprising 314,819 individuals. Our analyses provide strong evidence that longevity is transmitted as a quantitative genetic trait among survivors up to the top 10% of their birth cohort. We subsequently show a survival advantage, mounting to 31%, for individuals with top 10% surviving first and second-degree relatives in both data- bases and across generations, even in the presence of non-longevous parents. To guide future genetic studies, we suggest to base case selection on top 10% survivors of their birth cohort with equally long-lived family members.

Few studies examine how school and neighbourhood composition in the US correspond over time, particularly in a context of neighbourhood change. As neighbourhoods diversify along racial and economic lines, do public schools also diversify or grow increasingly dissimilar from their surrounding areas? Drawing on novel data linking neighbourhoods and schools in the US in 2000 and 2010, I document: how racial composition corresponds over time between traditional public schools and the neighbourhoods they serve; how the compositional gap changes when greater school choice is available; and how the compositional gap varies between neighbourhoods experiencing various trajectories of socioeconomic change. I find an increasing mismatch in the white composition of public schools and their surrounding neighbourhoods, specifically that schools enrol fewer white students than the composition of the neighbourhood. The compositional mismatch grows the most in neighbourhoods experiencing socioeconomic ascent, particularly as the number of nearby non-neighbourhood schools increases.

This dissertation explores empirical issues concerning public libraries in the United States. In particular, it analyzes the direct and indirect effects to and from public libraries, on different issues such as charitable donations, labor market outcomes, institutions and technical efficiency. Chapter 1 discusses the economics of public libraries and its relevance. In addition, it provides an outline of the dissertation and its main contribution to the literature. Chapter 2, analyzes the determinants of donations to public libraries testing the crowding effect from government funding of public libraries. I find suggestive results of a crowd in effect with an inverted U shape in all levels of government. In other words, additional government dollars increase private donations to public libraries at a decreasing rate. Chapter 3 investigates the effect of public library programs and participation on unemployment and labor force participation in Appalachia. Using an instrumental variable approach, I show that neither adult nor children’s programs and participation affect local labor market outcomes. These results are robust across different specifications. Chapter 4 explores the institutional determinants of public library technical efficiency in West Virginia. The analysis is carried out in two steps. First, I document considerable cross-district variation in library efficiency. Then, I estimate the effect of institutional characteristics on the technical efficiency measures. I find consistent evidence that urban libraries are more inefficient than county and multi-jurisdictional libraries. In addition, I find revenue from local sources is associated with reduced efficiency, contrary to what would be predicted by local public goods producer theory. Chapter 5 concludes the dissertation by summarizing its main findings and discussing possible implications.

The retrenchment of court-ordered school desegregation has been more variable and incomplete than often acknowledged, challenging common accounts that blame changes in federal policy and legal precedent. This study supplements these accounts by examining local factors that influenced whether and when desegregation orders were dismissed between 1970 and 2013. After accounting for federal policy changes and districts’ variable success in desegregating schools, several ostensibly race-neutral organizational, financial, and political incentives appear to influence the survival of desegregation orders. Racial competition dynamics related to local racial composition also seem to play a role, as desegregation orders have been most vulnerable when and where black population shares surpass a tipping point of about 40%.

The primary health insurance coverage reforms of the Affordable Care Act took effect on January 1, 2014. Between 2013 and 2016, the uninsurance rate for nonelderly Americans from birth to age 64 fell every year,1 and 18.5 million more Americans had health insurance coverage in 2016 than in 2013.2 However, these coverage gains stalled in 2017. Using the American Community Survey, we find that uninsurance increased by 0.2 percentage points between 2016 and 2017, meaning 700,000 more uninsured Americans (Figure ES 1). This increase in uninsurance occurred despite a strong economy and accompanying increases in incomes and employer-sponsored insurance coverage.

Ohio’s labor market continues to strengthen, with the 2018 unemployment rate falling to 4.5% from its high of 11% in 2010.1 Despite these gains, the statewide unemployment rate masks important changes in labor market participation, health, and access challenges facing the Ohio workforce. This brief examines changes in the Ohio labor market from 2008-2017, the overall health of Ohio’s workforce, and implications for Ohio’s Medicaid program. Analyses focus on adults 19-64 years of age.

Housing and opportunity are inextricably linked in the United States, with homeownership long seen as a source of stability and wealth-building. However, homeownership has become harder to access, particularly for younger prospective homebuyers and households of color. This analysis seeks to tease out the role of supply-side factors in shaping homeownership trends across the country. Through an analysis of Zillow data, decennial census data, and American Community Survey Public Use Microdata for the nation’s 100 largest metro areas, we find the following:

Intimate partner violence (IPV) has serious health consequences, both for immediate victims and their children. Our research considers whether exposure to battles at any level, not only at the level of all-out war, is a risk factor for IPV and, if so, whether this relationship holds broadly across a range of countries. We hypothesize that even with relatively low levels of conflict, exposure to local conflict events will lead to an increase in intimate partner violence. We consider this question across five sub-Saharan African countries during periods when they experienced localized armed battles, but not civil or international wars. Using multilevel regression analysis with data from IPUMS DHS, we find that exposure to any level of battles is associated with a 15% increase in the odds of IPV (OR=1.146, p < 0.05) in our five countries.

Background Uninsurance for young adults (YAs) was greatly reduced by the Affordable Care Act (ACA). However, reforms may not be equally beneficial for all YAs and certain policies may exacerbate, rather than resolve, pre-existing disparities. Objective To investigate inequalities in YAs’ insurance coverage pre- (2000–2010) and post-federal health reforms (dependent coverage expansion, 2010–2013, and Medicaid/Marketplace expansions, 2014–2016), among a nationally representative sample. Design A difference-in-differences estimator (controlling for sociodemographics) was used to determine the effects of the ACA for young adults (ages 19–25) compared to adolescents (ages 13–18) and older YA (ages 26–30) counterparts; triple-difference estimators quantified differential policy effects by sociodemographics. Participants Three hundred eighty-seven thousand six hundred thirty-five participants in the 2000–2016 National Health Interview Survey. Main Measures Respondents reported their health insurance coverage types during the last 12 months, reasons uninsured, and detailed sociodemographics. Key Results An adjusted difference-in-differences estimator quantified a 12.3 percentage point increase (p < 0.0001) in full-year coverage post-ACA . . .

Since the early 1990s, the social safety net for families with children in the United States has undergone an epochal transformation. Aid to poor working families has become more generous. In contrast, assistance to the deeply poor has declined sharply, and what remains often takes the form of in-kind aid. A historical view finds that this dramatic change mirrors others. For centuries, the nature and form of poor relief has been driven in part by shifting cultural notions of which social groups constitute the “deserving” and “undeserving” poor. This line was firmly redrawn in the 1990s. Did the re-institutionalization of these categorizations in policy have material consequences? In this study, we examine the relationship between the decline of traditional cash welfare during the 2001-2015 period and two direct measures of wellbeing among households with children: household food insecurity and public school child homelessness. Using models that control for state and year trends, along with other factors, we find that the decline of cash assistance is associated with increases in these two forms of hardship.

Older Black women have always worked outside the home despite limited occupational opportunities. In 1940 more than three-fourths of Black women worked as either private household workers or farm laborers. Since then, Black women have increased their educational attainment and moved in large numbers into clerical and professional occupations. Despite these advances, Black women aged 40 and older remain concentrated in a small number of occupations-almost half of older Black women work in just 20 occupations out of more than 400. Many of the occupations that older Black women work in are disproportionately low wage occupations. Automation and other technologies threaten many of the jobs older Black women work in including low wage jobs, middle-skill jobs that pay well but do not require a bachelor's degree and professional jobs that require a bachelor's degree or more. The risks of automation threatens to increase economic inequality-either through the growth of low-wage occupations where older Black women are disproportionately employed, or by increasing the risk of automation or digital skills substituting for workers in middle-skilled and professional occupations. Older Black workers are trying to keep up with the digital skills, technological, and educational requirements necessary for the future of work but these efforts are resulting in increased student loan debt-especially for older Black women-but not necessarily better jobs or earnings. 2 Acknowledgements:

This paper investigates the spatial distribution of immigrants across US cities. We document that (a) immigrants concentrate in large, expensive cities, (b) the earnings gap between natives and immigrants is higher in these cities, (c) these patterns are stronger when price levels in the country of origin are lower, and (d) immigrants consume less locally than natives. We develop a spatial equilibrium model in which immigrants spend a fraction of their income in their countries of origin. Thus, immigrants care not only about local prices but also about price levels in their home countries, which gives them a comparative advantage for living in more productive cities, where they accept lower wages than natives. We rely on variation in the origin price level to estimate the model. Counterfactual simulations suggest that current levels of immigration have reduced economic activity in smaller, less productive cities, while they have expanded it in larger, more productive ones. This has increased total worker productivity by around 1 percent and aggregate native workers’ welfare by around 0.35 percent.

Objectives. To determine what role the 88 000 Housing and Urban Development–Veterans Affairs Supportive Housing (HUD-VASH) vouchers for permanent supportive housing among US veterans distributed between 2008 and 2017 played in the significant fall in veterans’ homelessness over the same time period. Methods. Using a panel data set at the Continuum of Care level over the 2007 to 2017 period, we correlated changes in vouchers with permanent supportive housing units and measures of homelessness. To reduce concerns about omitted variables bias, we used a 2-stage least-squares procedure. The instrument is a Bartik-type shift-share variable. Specifically, for the cumulative vouchers received at the local level, we used the share of the nation's homeless veterans from the local level in the year before the HUD-VASH program multiplied by the cumulative number of vouchers distributed at the national level up to that point. Results. For each additional voucher, permanent supportive housing units increased by 0.9 and the number of homeless veterans decreased by 1. Conclusions. Our results indicate the HUD-VASH program worked as intended and veterans’ homelessness would have risen substantially over the past decade without the program.

Asian Americans are the fastest growing and most diverse group in the country; they were 1 percent of the population in 1970, 6.4 percent today and are projected to be about 10 percent by 2060. Immigration has driven much of this growth. China and India have surpassed Mexico as the leading sources of new immigrants, and by 2055, Asians will become the largest immigrant group (Colby and Ortman 2015; U.S. Census Bureau 2015). The new face of immigration is Asian, but Asian is a catch-all category that masks tremendous diversity, heterogeneity, and inequality. In 1960, 80 percent of the U.S. Asian population was either Chinese or Japanese, but today their share is 20 percent. Immigrants and refugees from South and Southeast Asia have fueled the growth and diversity of the Asian American population U.S. Census Bureau 2016). And unlike other ethnoracial groups, most Asians are foreign-born (U.S. Census Bureau 2015). Two-thirds are immigrants and 90 percent are either immigrants or their children. Moreover, one in seven Asian immigrants is undocumented, and this group is growing at a faster rate than the Mexican and Central American undocumented populations (Ramakrishnan and Shah 2017).

Existing research shows that on average first-generation immigrants earn less than native-born workers in the United States, especially during their first decade in the country, but eventually overtake the native-born; second-generation immigrants (2g) tend to earn more than subsequent generations (3+g). However, the labor market outcomes of the “1.5 generation” (1.5g), the foreign born who migrate as children, have not been thoroughly analyzed. This paper hypothesizes that the 1.5g could have an earnings advantage relative to subsequent generations due to the higher ability of their parents; however, the accumulation of destination-specific human capital (i.e., acculturation) also declines with age at migration (AAM). Using a Mincerian earnings regression as applied to CPS data (1994–2016), this analysis tests whether there is a within-group “critical” AAM threshold after which the earnings advantage of the 1.5g becomes an earnings disadvantage. Results show that this threshold is approximately ages 5–9, varying by ethnicity and gender.

In December 2017, federal tax legislation ended federal enforcement of the Affordable Care Act's (ACA's) requirement that consumers who can afford health insurance must obtain it. Lawmakers took that step despite projections from the Congressional Budget Office and other non-partisan experts that, without an individual responsibility requirement, premiums would rise and the number of uninsured consumers would climb. Last year, New Jersey, Vermont, and other states 1 responded by passing laws that use state tax systems to enforce a requirement for people to obtain coverage if they can afford it. This approach builds on Massachusetts's favorable experience with such a law, in effect since 2006 under both Republican and Democratic governors. By increasing the incentive for young and healthy consumers to enroll, these laws lower average costs and cut premiums in the individual market. Moreover, when fewer uninsured seek hospital care, hospital uncompensated care costs decline, lowering hospital charges to private insurers and saving money on premiums for employers and workers alike. Maryland's down payment plan seeks to achieve even better results by using a more enrollment-oriented approach than federal lawmakers originally devised for the ACA. Instead of imposing tax penalties on the uninsured, the Maryland proposal helps the uninsured enroll into coverage whenever possible. As uninsured residents file state tax returns, they can avoid an insurance responsibility payment by agreeing to get health insurance and keep it through the end of the year. They can apply for financial assistance and immediately begin the enrollment process by simply checking a box on their tax return. That step sends information to the exchange, which reaches out and helps uninsured consumers sign up for coverage. Roughly 130,000 uninsured Maryland residents qualify for zero-premium health insurance but are not enrolled. More than 50,000 are eligible for Medicaid or the Children's Health Insurance Program. Nearly 80,000 qualify for private insurance that costs less than their federal premium tax credit plus, in some cases, their insurance responsibility payment. In effect, residents can turn their penalty payments into health insurance down payments, using their own money to help buy coverage for themselves and their families. The number of uninsured would decline by more than one-third, the individual market's risk pool would improve substantially, and premiums would decline below the levels otherwise charged. The remainder of this document answers frequently asked questions about aspects of the down payment plan, which has been introduced as SB 802 and HB 814.

The theoretical argument for understanding long-term trends of the association between feminisation and occupational pay rests on a distinction between two processes related to gender inequality, that have occurred in recent decades in the labour markets of most western democracies. The first process, which relates to women as individuals, addresses the upward occupational mobility of women, meaning that women are incrementally entering higher rungs of the occupational hierarchy. The second process-which is conceptualized as a structural process-refers to the criteria for rewarding occupations. The question in this regard is whether gender is one of the criteria and if so whether occupations are devaluated following the entry of women. The literature documenting long-term trends in gender inequality has tended to focus heavily on the former-i.e. the upward occupational mobility of women on the occupational ladder. The latter, i.e. the structural implications are largely overlooked. However, the two are inherently connected, as a notable consequence of the growing occupational attainments of women over recent decades is evident in the way occupational feminisation affects the pay level of occupations. Although a considerable amount of research has highlighted the negative association between the percentage of women in occupations and their rewards, most of these studies have focused on the causal mechanisms of the process rather than on the dynamics over an extended period of time. Prof Hadas Mandel sought to address this lacuna in the literature by examining trends in the effect of occupational feminisation on occupational pay over several decades in the US and exploring the mechanisms underlying these trends. Using integrated data on individuals and occupations from the US Census (1960-2010) and the ACS surveys (2001-2015), her findings show, similarly to previous studies, that in recent decades, and especially from 1980 onwards, a growing number of women in the US have approached the head of the occupational ladder (see the first figure below). This shift has been fuelled by women's growing educational attainments, which, together with the rising economic premium to education, have greatly contributed to the decline in gender wage gaps. Furthermore, based on these changes, the negative association between female percentage in Gender inequality: occupational devaluation and pay gaps The comparative research of long-term trends of gender inequality largely neglects structural mechanisms. As more women reach positions of power, structural elements will become more significant. Despite the growing body of literature in this area, the long-term effect of the changing gender composition of occupations on their relative pay has been largely neglected. Hadas Mandel, an Associate Professor in the Department of Sociology and Anthropology at Tel Aviv University, has addressed this gap in the literature by exploring the negative effect of occupational feminisation on occupational pay in the US and the mechanisms underlying these trends. Behavioural Sciences-occupations and occupational pay levels declines over time (see model 1 in the second figure below). This decline is most apparent from 1980 onward, a period in which US women witnessed a significant improvement in their occupational standing, and a period where occupations requiring higher education enjoyed a large wage premium. However, when examining the effect of gender composition of the occupation after accounting for women's higher education and for the level of education in occupations, the trend is reversed; the negative net effect of female percentage on occupational pay intensifies over time (see model 2 in the second figure below). These two opposite processes reflect the upward occupational mobility of women, on the one hand, and its gendered consequences, on the other hand. The major role education plays in explaining the divergent trends is twofold. The entry of women into occupations requiring higher education, and the growing economic reward to high education and to occupations with higher educational requirements, may both conceal the trend in the devaluation effect as they contribute to weakening the correlation between the percentage of women and pay across occupations over the course of time. Thus, the intensification of the devaluation effect is revealed only after controlling for education (at both the occupational and individual level), because the growing educational level of women, and the growing rewards to education, are processes that run counter to devaluation and thus conceal its intensification. Let's consider the example of industrial engineers and electrical engineers. Both occupations demand high education (more than 70% of incumbent workers in 2010 had an academic degree), and both enjoyed a wage premium during the period studied. However, while in both occupations the percentage of women in 1960 was negligible (2% and 1%, respectively), 50 years later only 10% of electrical engineers were women, compared to 19%-almost double-of industrial engineers. As both occupations enjoyed wage premiums, the devaluation effects may not be observed because the process of feminization was not followed by an absolute wage reduction. Rather, feminization is associated with a smaller wage premium relative to comparable highly educated occupations. Thus controlling for education is essential for revealing the devaluation process. Indeed, we see that while electrical engineers enjoyed a premium of 25% during the period studied, industrial engineers enjoyed a premium of less than 19%. The findings demonstrate the interrelationship between two opposing gendered processes and provide concrete evidence that gender stratification operates differently at the individual and at the structural/ occupational level. The split between individual and occupational forms of gender in/equality and the divergent trend of each are crucial for our understanding of gender inequality in theory as well as in practice. This is because structural mechanisms are not directed at any specific individual and thus are more ambiguous and more difficult to track empirically. The danger is Structural mechanisms of gender/race inequality are not directed at any specific individual and thus are more ambiguous and more difficult to track empirically. Professor Hadas Mandel Trends in female proportion in occupations by levels of average occupational pay. 0.6 0.5 0.4 0.3 0.2 0.1 Low pay Tertile Mid pay Tertile High pay Tertile Trends in the effect of gender composition (% female) on the average pay of occupations, before (Model 1) and after (Models 2) accounting for education.

We proxy for board members’ opinions and values using directors’ ancestral origins and show that diversity has costs and benefits, leading to high performance volatility. Consistent with the idea that diverse groups experiment more, firms with ancestrally diverse boards have more numerous and more cited patents. In addition, their strategies conform less to those of the industry peers. However, firms with greater ancestral diversity also have more board meetings and make less predictable decisions. These findings suggest that diversity may lead to inefficiencies in the decision-making process and conflicts in the boardroom.

There has been considerable growth and interest in industrial applications of machine learning (ML) in recent years. ML engineers, as a consequence, are in high demand across the industry, yet improving the efficiency of ML engineers remains a fundamental challenge. Automated machine learning (AutoML) has emerged as a way to save time and effort on repetitive tasks in ML pipelines, such as data pre-processing, feature engineering, model selection, hyperparameter optimization, and prediction result analysis. In this paper, we investigate the current state of AutoML tools aiming to automate these tasks. We conduct various evaluations of the tools on many datasets, in different data segments, to examine their performance, and compare their advantages and disadvantages on different test cases.

In order to investigate the role of immigrant artists on the development of artistic clusters in U.S. cities, we use the U.S. Census and American Community Survey, collected every 10 years since 1850. We identify artists and art teachers, authors, musicians and music teachers, actors and actresses, architects, and journalists, their geographical location and their status as a native or an immigrant. We look at the relative growth rate of the immigrant population in these occupations over a ten year period and how it affects the relative growth rate of native-born individuals in these artistic occupations. We find that cities that experienced immigrant artist inflows, also see a greater inflow of native artists by about 40%.