The majority of U.S. households that qualify for federal rental housing assistance do not receive it. In the absence of an entitlement to housing assistance, an underexplored cause of the shortfall is that higher rents in some areas driven by supply-constraining local regulations increase program costs, leaving fewer funds available to serve additional families. In this paper, we simulate the effect of increasing housing supply on the cost of Section 8 housing assistance programs in Los Angeles, as well as all 11 metropolitan areas most constrained by local regulations. If Los Angeles (all 11 metropolitan areas) produced new housing units at the same rate as the 90th percentile metropolitan area for a decade, market rents would fall by 18.1 percent (2.0 to 24.0 percent), and federal cost savings would equal $353 million ($1.8 billion), enough to increase the number of assisted families by 23.8 percent (18.6 percent). For comparison, doubling the number of units placed in service through the Low Income Housing Tax Credit for a decade—an alternative method for increasing housing supply—would lead to much lower cost savings of $18 million in Los Angeles, and $231 million across all 11 metropolitan areas.

The National Health Interview Survey (NHIS) data on self-reported trouble hearing and the use of hearing aids were examined for the 12 recent surveys from 2007 to 2018 for adults from 18 to 85+ years of age. The aggregate dataset for all years included data from 357,714 adult respondents. Sample size for annual data ranged from 22,058 (2008) to 36,798 (2014). The prevalence of self-reported trouble hearing and hearing aid use, both current use and ever-using hearing aids, are reported for males and females for each age decade. Measures of unmet hearing healthcare (HHC) need were derived from estimates of the prevalence of hearing aid use among those with self-reported trouble hearing. Logistic-regression analyses identified variables affecting the odds of having self-reported trouble hearing, of using or rejecting hearing aids, and of having unmet HHC needs. The results largely corroborate and extend the findings of recent analyses of data from the National Health and Nutrition Examination Survey (NHANES) for a similar period (2011-2020). Overall, for males, 18.5% (95% CI [18.2%-18.8%]) had self-reported trouble hearing and 76.6% [76.0%-77.2%] of these individuals had never used hearing aids and, for females 13.1% [12.9%-13.4%] had trouble hearing and 79.5% [78.9%-80.1%] of these individuals had never used hearing aids. Unmet HHC needs are highly prevalent in the United States and have been so for many years.

A person's gender is a crucial piece of information when performing research across a wide range of scientific disciplines, such as medicine, sociology, political science, and economics, to name a few. However, in increasing instances, especially given the proliferation of big data, gender information is not readily available. In such cases researchers need to infer gender from readily available information, primarily from persons' names. While inferring gender from name may raise some ethical questions, the lack of viable alternatives means that researchers have to resort to such approaches when the goal justifies the means-in the majority of such studies the goal is to examine patterns and determinants of gender disparities. The necessity of name-to-gender inference has generated an ever-growing domain of algorithmic approaches and software products. These approaches have been used throughout the world in academia, industry, governmental and non-governmental organizations. Nevertheless, the existing approaches have yet to be systematically evaluated and compared, making it challenging to determine the optimal approach for future research. In this work, we conducted a large scale performance evaluation of existing approaches for name-to-gender inference. Analysis are performed using a variety of large annotated datasets of names. We further propose two new hybrid approaches that achieve better performance than any single existing approach.

Conventional narratives of the 2020 presidential election typically overlook the growing importance of voters of color in suburbia, especially in battleground states such as Georgia and Virginia. Using novel precinct-level voting data for twenty-two states for the 2016 and 2020 presidential elections, we analyze the relationship between precinct-level Democratic vote shares and the racial composition of suburban precincts. Predominantly White suburban precincts turned modestly toward the Democratic Party in 2020 relative to 2016, but had the election been decided in the precincts in which most White suburban voters live, Donald Trump would have won both elections handily. The outcome in 2020 depended on turnout in heavily Black suburban precincts, which voted overwhelmingly for Biden, and in Asian and Latinx precincts that also supported Democrats, though less strongly. As the suburbs become increasingly diverse, the new racial demography of suburbs should change conventional understandings of voting behavior in these spaces.

Gender gaps in labor market activity are pervasive, longstanding, and a regular subject of policy debates. Relative to men, women tend to work fewer hours per week, more conventional hours, and fewer years over the course of their lives.2 These differences in the intensity and timing of work contribute to gender disparities in promotions and pay.3 But despite decades of research on this topic, little attention has been paid to the timing of work throughout the year. To motivate our inquiry, Figure 1 plots the monthly labor force participation rates of prime-age US women and men using non–seasonally adjusted data, with June, July, and August shaded gray. Summer after summer, women's labor force participation drops sharply while men's participation does not.

This StoryMap shows how racially segregated neighborhoods were created in San Bernardino from the 1920s through the 1960s, and how these patterns of segregation shaped schools, city planning and Black-white wealth disparities. It documents the ways restrictive covenants, white vigilante violence, real estate steering, and white flight shaped the color line in San Bernardino as the Black population increased from 1940 through 1970. The StoryMap uses census data newly compiled by a People's History of the I.E. Census Project from IPUMS Ancestry Full Count Data and from IPUMS USA data to visualize the changing patterns of racial segregation from 1940-1970. The StoryMap also draws heavily on the stories of Black elders from the Bridges That Carried Us Over Project: Documenting Black History in the Inland Empire, and on the People's History Map of the Archives which has created a geospatial database that enables us to trace the geography of real estate ads, white vigilante actions, and Black pioneers who crossed the color line.

This study shows that military attacks —through fear and panic— can distort political behavior and create a “conservative shift” in subsequent elections. Using the distance to the Ellwood bombardment in 1942, a shelling of civilian installations on the US mainland during WW2 which caused minimal damage but that created a large wave of panic, we find that support for Republican candidates increased in subsequent Gubernatorial, Presidential and House elections in Californian counties in the vicinity of the incident. Interestingly, the effect appears to persist for a long time, even after WW2 ended. Using a large corpus of articles from Californian newspapers and text analysis, we provide evidence that the event led to a persistent shift in conservative beliefs of local communities. We conclude that attacks, through their psychological effects, might have long-run consequences through preference-shifting and changes in voting behaviors.

Economic lives have become fluid in the United States during recent decades, with individual or family incomes often varying from month to month, from quarter to quarter, or across years. The income instability that hourly workers and their families are facing is mostly driven by frequent earnings changes, which could be a result of unpredictable work hours or involuntary job churn. This, along with insufficient work hours (another understudied workplace issue), may jointly affect low-income workers’ economic security. The current economic recovery, with strong job growth and high rates of jobs switches, provides a unique opportunity to examine work-hour insecurity issues and better understand the extent to which the volatility of work hours has leveled off in the past two years and who does or does not benefit from the ongoing economic recovery. Using a nationally representative sample of workers, this brief provides a portrait of intra-year work-hour volatility (or instability) and patterns over time, from 2016 to 2022, by wage level, parental status, race and ethnicity, educational attainment, and age. It then explores whether specific demographic groups exhibit higher variability in working hours in the phase of economic recovery, net of other workers’ characteristics. For comparison, I also examine levels of work hours across groups. Results reveal that greater volatility is marked among workers from the bottom wage quintile, the less educated, and the young. Unmarried parents and workers of color experience relatively high volatility in hours. Despite regaining hours as the economy reopened, these groups consistently worked fewer hours over the period examined. Prior to the pandemic, the gaps in volatility across groups narrowed; although COVID-19 erased some of this progress, the decline in the level of volatility workers faced in 2022 suggests a return to pre-pandemic conditions. The extent to which those gains will be sustained remains to be seen as economic recovery progresses.

Immigration to Hawai‘i between 1870 and 1930 led to a more than six-fold increase in population and resulted in high and rapidly varying sex ratios in the Chinese, Japanese, Korean, Filipino, and Caucasian populations of marriageable age. Using complete populations of the 1910, 1920, and 1930 Territorial Censuses of Hawai‘i, we estimate how male-biased sex ratios, both within a particular ethnic group and in all other ethnic groups, affected important household choices of second-generation men and women of marriageable age. Econometric results indicate that within-group and extra-group sex ratios impact the likelihood of males and females to marry, to marry a spouse from another ethnic group, to have children, and to live in larger households.

We investigate the effect of personal income tax (PIT) rates on the number of hours entrepreneurs work weekly. Using the rotating panel data from the Annual Social and Economic Supplement of the Current Population Survey from 2003 to 2019, we estimate instrumental variable regressions in first differences to exploit changes in the tax code for identification. We distinguish between self-employed owners of incorporated versus unincorporated businesses and examine their differential responses. The findings reveal that higher individual-specific marginal PIT rates increase the hours worked among entrepreneurs with incorporated businesses, which could be explained by the availability of tax avoidance strategies. Among unincorporated entrepreneurs, we find a significant response to PIT rates in hours worked only for those who work 50 or more hours per week.

This StoryMap traces one step on a long fight for educational justice and racial equality in San Bernardino California. Black and Mexican parents and organizers had long struggled against racial segregation and unequal education in the city as housing segregation had created deeply segregated and unequal schools in the city. But throughout the 1960s, demands for change grew louder as new groups like CORE and the Community League of Mothers organized to press for change. The school district made incremental changes throughout the 1960s and early 70s in response to community pressure, but by 1976 the NAACP won a major lawsuit against the school district which forced more substantial plans to desegregate the schools including busing and magnet schools which transformed the geography of segregation in the city.

The purpose of this study was to examine factors associated with negative job outcomes for nurses during the initial phase of the COVID-19 pandemic, with a focus on nurses of color. The study used data from 3,782 nurses in the Current Population Survey to examine the relationship between nurse characteristics and COVID-19-related inability to work or look for work during May through December 2020. The analysis showed that race and gender did not significantly impact nurses’ job outcomes. The odds of a negative impact were increased by age (1.5% per year, p <.05), having a child in the home (43%, p <.01), having no spouse present (36%, p <.01), and working in an outpatient role (48%, p <.001). While race alone was not linked to negative outcomes, nurses of color had higher rates of other factors that were associated with negative outcomes, indicating a need for a more nuanced examination of their work and life contexts and job outcomes throughout the pandemic.

Background: To date, several studies have shown that parenting stress, a kind of role strain, is related to child maltreatment. However, few studies have examined how the effects of crises, such as negative pandemic-related events on the household, may be related to parenting stress and maltreatment-related behavior. Objective: This study examines the impact of negative Covid-related events on parenting stress and parenting behaviors during a period that was likely to have been a peak point of stress for many parents during the Covid-19 pandemic. Participants: Respondents were female caregivers (N = 720) of children under the age of six located in the Midwestern United States. Results: Three or more Covid-related impacts on the household were positively associated with parenting stress (B = 0.177, p < 0.05). Parenting stress fully mediated the weak relationship between these impacts and maltreatment-related behavior. Mothers of different employment statuses, including those who were recently laid off or who chose to stay at home, did not have significantly different probabilities of parenting stress or maltreatment-related behaviors. Contrary to theory, similar null results were found across other socio-demographic variables. Conclusions: These null findings suggest that crises have effects that encompass family systems, potentially raising parenting stress levels in many groups that are typically considered low-risk for child maltreatment. Results have implications for scholarship on parenting stress, the targeting of social supports to mothers of young children, and rapid interventions to reduce stress, such as the stimulus check relief program.

We document systematic differences in wage and earnings inequality between and within occupations and show that these differences are intimately related to systematic differences in labor supply across occupations. We then develop a variant of a Roy model in which earnings are a non-linear function of hours, with the extent of this non-linearity differing across occupations. In our theory, the interplay between heterogeneity in tastes for leisure and occupational differences in non-linearities affects the sorting of workers. Moreover, this interplay is crucial to account for the facts on the distributions of hours, wages, and earnings within and across occupations.

We use linked parent-child administrative data for five countries in North America and Europe, and detailed survey data for two more, to investigate methodological challenges in the estimation of absolute income mobility. We show that the commonly used “copula and marginals” approximation methods perform well across countries in our sample, and the greatest challenges to their accuracy stem not from assumptions about relative mobility rates over time but from the use of non-representative marginal income distributions. We also provide a multi-country analysis of sensitivity to specification decisions related to age of income measurement, income concept, family structure, and price index.

We examine the determinants of the consequences of COVID-19 on employment and wages in the United States. Guided by a pre-analysis plan, we investigate whether the economic consequences of COVID-19 were larger for certain occupations, using four indexes: workers relatively more exposed to disease, workers that work with proximity to coworkers, essential/critical workers and workers who can easily work remotely. We find that individuals that work in proximity to others are more affected while individuals able to work remotely and essential workers are less affected by the pandemic. We also present suggestive evidence that our indexes are likely explanations why certain demographic groups such as younger and minority workers have worse labor market outcomes during the pandemic.

We use linked historical US censuses to study the empirical relationship between inequality and intergenerational mobility. We first confirm that the "Great Gatsby Curve" already existed in the early 20th century. We then study a "dynamic" version of the curve that relates changes in equality to changes in intergenerational mobility. Surprisingly, we find that this relationship is unstable over horizons of two decades for income, but not for education. Finally, we propose novel unitless measures of intergenerational mobility and inequality to show that the "Great Gatsby Curve" result re-emerges over the long run, for the period 1920 to 2011.

This dissertation explores the role federal policy can play in rectifying gender inequities in labor market outcomes. The first chapter shows that federal anti-discrimination policy helped to increase women’s representation in medical education, contributing to a stark decline in occupational segregation in the second half of the 20th century. The second chapter shows that equal pay policy in the 1960s led to marked increases in women’s earnings, staving off an increase in the gender pay gap, especially for lower-earning women. The third chapter makes a theoretical contribution to our understanding of why firms offer parental leave, an amenity at the heart of discussions surrounding the differential labor market impact of childbirth on women and men. In Chapter 1, I consider the role of federal policy in increasing women’s access to medical education. In the 1960s, women comprised under 10% of all medical students, resulting in a lopsided gender imbalance in the medical profession. I find that a host of federal anti-discrimination policies, implemented in the late 1960s and early 1970s, increased women’s enrollment by successfully applying pressure to curb sex discrimination in admissions though the threat of revoking federal funding. In addition, this policy amplified the impact of a massive expansion in enrollment through Health Manpower policy on women by allowing them to capture a higher fraction of newly created seats. In Chapter 2, Martha J. Bailey, Bryan A. Stuart and I consider the success of two landmark statutes—the Equal Pay and Civil Rights Acts—in targeting the long-standing practice of employment discrimination against U.S. women. At the beginning of the 1960s, the gender earnings ratio at the median had dipped below 60%, and the stability of this statistic over the next decade suggested that policy had not done much to alleviate this gap. We revisit this conclusion using two separate causal designs. In our first design, we find that women’s wages grew more quickly after 1964 in states that did not have pre-existing equal pay laws, where we would expect the effects of federal policy to be the strongest. Then, in our second design, we find larger wage growth for women working in jobs with a higher wage gap, where pay discrimination is more likely to be present. In Chapter 3, I consider the question of why firms voluntarily offer parental leave as a benefit to employees. The federal government requires covered employers to provide only 12 weeks of job protected unpaid leave through the Family and Medical Leave Act, but many employers provide additional wage replacement and leave time beyond this statutory requirement. I consider a setting where workers and firms search over a collection of submarkets characterized by the posted wage and likelihood of finding an employment match. Firms are homogenous but are able to choose whether or not they offer a job with parental leave. Workers differ from one another in the amount of time they would spend out of the labor market on leave after childbirth. I find that firms will only offer leave when the duration of expected leave is below a particular threshold. This threshold is given at the point where the benefit of leave, given by hiring savings, is higher than the cost of retaining a worker on leave net of pass through to the employee.

Background: Asian Americans (AA) young adults face a growing non-communicable disease burden linked with poor dietary behaviors. Family plays a significant role in shaping the diet of AA young adults, although little is known on the specific types of family structures most associated with different dietary behaviors. Aim: This analysis explores the changes in dietary behaviors across different AA young adult family structural characteristics. Methods: Nationwide data of 18–35-year-old self-identified Asians surveyed in the 2015 National Health Interview Survey (NHIS) was analyzed. Family structure was measured through family size, family health, and family members in one's life. The Dietary Screener Questionnaire (DSQ) measured the average intake of 10 food and nutrient groups. Published dietary guidelines were used to calculate the number of dietary recommendations met. Results: 670 AA young adults with dietary data were analyzed (26.1% Asian Indian, 26.1% Chinese, 19.3% Filipino, 28.5% other Asian). Participants had an average family size of 2.3. In weighted analyses, 19% of AA young adults met none of the examined dietary recommendations, and only 14% met 3–4 guidelines. Living with a child was associated meeting more dietary recommendations (adjusted odds ratio [AOR]: 1.22; 95%CI: 1.05, 1.42). The adjusted association between living with an older adult and lower odds of meeting dietary recommendations approached significance (AOR: 0.70; 95%CI: 0.49, 1.00). Conclusions: Findings revealed the important role of children and older adults in influencing the diet of AA young adults. Further mixed-methods research to disentangle mechanisms behind the influence of family structure on diet is warranted.

This paper revisits the cyclical nature of births and infant health and investigates to what extent the relationship between aggregate labor market conditions and birth outcomes is mitigated by unemployment insurance (UI). We introduce a novel empirical test of standard neoclassical models of fertility that directly tests the prediction of opposite-signed income and intertemporal substitution effects of business cycles by examining the interaction of the aggregate unemployment rate with a measure of potential income replacement from UI. Our results show that as UI benefit generosity reaches 100 percent income replacement, there is no effect of the unemployment rate on births. This implies that the well-documented cyclical nature of births is about access to liquidity. We also provide novel evidence that infant health is countercyclical based on timing of conception, but procyclical based on time in utero. The negative relationship between the in utero aggregate unemployment rate and infant health also disappears when potential UI replacement rates reach 100 percent. Our results imply that the social insurance provided by UI has a pro-natalist effect and improves the health and economic well-being of the next generation.