The Food Stamp program is currently one of the largest safety net programs in the United States and is especially important for families with children. The existing evidence on the effects of Food Stamps on children’s and families’ outcomes is limited. I utilize a large, recent source of quasi-experimental variation–changes in documented immigrants’ eligibility across states and over time from 1996 to 2003–to estimate the effect of Food Stamps on children’s health. I find loss of parental eligibility has large effects on program receipt, and an additional year of parental eligibility before age 5 improves health outcomes at ages 6-16.

Differentially private learning tackles tasks where the data are private and the learning process is subject to differential privacy requirements. In real applications, however, some public data are generally available in addition to private data, and it is interesting to consider how to exploit them. In this paper, we study a common situation where a small amount of public data can be used when solving the Empirical Risk Minimization problem over a private database. Specifically, we propose Private-Public Stochastic Gradient Descent, which utilizes such public information to adjust parameters in differentially private stochastic gradient descent and fine-tunes the final result with model reuse. Our method keeps differential privacy for the private database, and empirical study validates its superiority compared with existing approaches.

Background: We know diseases such as the 2019 Novel Coronavirus (COVID-19) are spread through social contact. Moreover, interventions to control social contacts such as stay-home orders are required to stop disease spread in pandemics for which vaccines have not yet been developed. However, existing data on social contact patterns in the United States (U.S.) is limited. Method: Consequently, we use American Time Use Survey data from 2003-2018 to describe and quantify the number and duration of social contacts occurring at home and in nonhousehold locations. For household locations we also estimate age contact matrices (who spends time with whom by age). This is the first study to describe variation in U.S. social contact patterns across space, time, and based on demographic characteristics. Findings: We find that gender differences in social contact patterns exist. In the home, they appear to be driven by caretaking responsibilities. Non-Hispanic Blacks have a shorter duration and fewer social contacts than non-Hispanic Whites. However, they are more likely to work in jobs that require close physical proximity, therefore the nature of their contacts is riskier. Hispanics have the highest number of household contacts and are also more likely to work in jobs requiring close physical proximity compared to non-Hispanic whites. Seasonal differences in contact patterns are not large; they appear to be driven by the school term and therefore are chiefly present in school-aged respondents. Consequently, if the main mechanism driving infectious disease seasonality is seasonality in contact patterns, then we should not expect to see large seasonal differences in disease incidence when the young are not very susceptible or infectious. Spatial differences in contact patterns are small. Conclusion: In addition to age, demographic characteristics, particularly race and ethnicity and gender, are associated with differences in social contact patterns. In contrast, seasonal differences seem to be associated with school participation and therefore less consequential for older respondents.

Since 1990, U.S. policymakers have worked to close gaps in academic achievement by income and race (e.g. with school finance reform and school accountability systems) even as rising income inequality and income-based residential segregation have threatened to widen them. Using estimates of the mean and variance in household income for sampled schools, we reconstruct the student-level relationship between achievement and household income in the National Assessment of Educational Progress (NAEP) from 1990 to 2015. We find that achievement at all levels of parental income rose substantially in 4th and 8th grade. In contrast to Reardon (2011), we find that achievement gaps narrowed substantially in 4th grade reading and math and in 8th grade math, while the gaps remained stable in 8th grade reading. As a robustness check, we used the March Current Population survey to impute income for dependent children by race, mother’s education, urbanicity and state and then calculated mean achievement for those same groups in the NAEP. Again, we found gaps in achievement narrowing between groups with high and low predicted mean household incomes. Our results challenge the prevailing understanding that income-based achievement gaps have widened in the United States over the last 30 years.

Local economies benefit from attracting in-migration either as workforce or as consumers. To compete for constituents, local economies need to provide attractive tax policy and expenditure bundles. An important consideration in this regard is the relative natural advantage of some locations in terms of its climate and geographical features, among other things. In this three-paper dissertation, I explore how natural amenities affect the variations in local government public goods and how people choose their residential locations as they trade-off between natural amenities and local government public goods as they go through phases in their life cycle. In Article 1, I propose that one of the reasons locations differ in their stock of local government public goods is because of differences in existing natural amenities by testing the hypothesis that some goods are substitutes while others are complements using spatial autoregressive random effects model estimation. In Article 2, I explore how local government expenditures and population vary in two contiguous areas that are similar in all but one natural amenity using border-matching methodology to determine how local government expenditures differ between counties sharing a border within a state that have the same level of natural amenities except for one natural feature. In Article 3, I use fixed effects panel data regression to test whether age and life milestones shape preferences and budget constraints of people when they choose residential locations as they trade-off between natural amenities and local government-provided public goods. My results indicate that some natural amenities are complements to local public goods while others are substitutes. Some expenditures are not affected by natural amenities because they have to be provided regardless of what are naturally available. Moreover, age and marital status are consistent predictors of moving. Natural amenities and certain per capita tax revenue and expenditure items also affect the likelihood of moving.

This paper examines the evidence from randomized evaluations of sector-focused training programs that target low-wage workers and combine upfront screening, occupational and soft skills training, and wraparound services. The programs generate substantial and persistent earnings gains (11 to 40 percent) following training completion. Theoretical mechanisms for program impacts are explored for the WorkAdvance demonstration. Earnings gains are generated by getting participants into higher-wage jobs in higher-earning industries and occupations not just by raising employment. Training in transferable and certifiable skills (likely under-provided from poaching concerns) and reductions of employment barriers to high-wage sectors for non-traditional workers appear to play key roles.

Almost 33 million working people in the United States—about one in five—work part time. Many people work part time to support their families while caring for loved ones, going to school, or attending to other obligations, but are penalized for choosing part-time work in terms of pay, benefits, stability, and opportunities to advance. And for many others, working part time isn’t a choice at all: some employers, particularly in low-wage service industries, rarely offer full-time positions, and some employees—especially women—find that caregiving or other responsibilities preclude full-time work.

The implementation of a proposed differential privacy algorithm to 2020 US Census data releases, and other census products has brought about discussions about the consistency and reliability of the data produced under the proposed disclosure avoidance system. We test the potential impact of this change in disclosure avoidance systems to the tracking of population growth and distribution using county-level population counts. We ask how population counts produced under the differential privacy algorithm might lead to different conclusions regarding population growth for the total population and three major racial/ethnic groups in comparison to counts produced using the traditional methods. Our results suggest that the implementation of differential privacy, as proposed, will impact our understanding of population changes in the US. We find potential for overstating and understating growth and decline, with these effects being more pronounced for non-Hispanic blacks and Hispanics, as well as for non-metropolitan counties. These findings draw attention to the potential local consequences of the implementation of differential privacy for tracking demographic changes of the US population, which is bound to have implications for our understanding of the transformations the nation is going through.

This dissertation examined the roles of maternal characteristics and state-level immigrant policies on youth uninsurance and receipt of family centered care (FCC) among US-born Latino youth. For the analyses of this dissertation, we relied on three nationally representative databases (National Health Interview Survey, Medical Expenditure Panel Survey, and American Community Survey). The first aim examined the direct effect of maternal citizenship and health insurance coverage status on Latino youth health care access. Using generalized structural equation modeling, compared to citizen Latina mothers, noncitizen Latina mothers had higher odds of being uninsured (aOR, 4.75). Maternal uninsurance was associated with youth uninsurance and there was an indirect pathway in the association between maternal citizenship and youth uninsurance that operates via maternal uninsurance. The second aim assessed the association between Latina maternal-provider ethnic concordance and youth receipt of FCC. Relying on a series of multivariable logistic regressions, maternal-provider ethnic concordance was positively associated with reports of the medical care provider listening carefully to the parent (aOR, 1.71), explaining things in a way the parent could understand (aOR, 1.75), showing respect for what the parent had to say (aOR, 1.98), and spending enough time with the patient (aOR, 1.45). Lastly, the third aim studied the association of maternal citizenship and state-level integration and criminalization immigrant policies on uninsurance for US-born Latino youth. We performed average marginal effects methods and found that youth uninsurance for youth with noncitizen mothers is 3.3% higher than youth with citizen mothers within states with few inclusive immigrant policies. For youth who have noncitizen mothers, the difference in the predicted probability in uninsurance between those who resided in states with few inclusive versus high level of inclusive immigrant policies is 2.1%. Youth with noncitizen mothers who resided in states with a high level of criminalizing immigrant policies had a 2.6% higher probability of being uninsured compared to youth with citizen mothers. The difference in the predicted probability in uninsurance among youth with noncitizen mothers between those who resided in states with a low level versus a high level of criminalizing immigrant policies is 1.7%.

Postsecondary education provides an opportunity to increase the economic potential of individuals. Earnings for individuals with cognitive disabilities are a major concern, as occupational outcomes are often dire. The prevalence of individuals with cognitive disabilities in postsecondary education settings is increasing, but little is known about how postsecondary attendance may relate to post-graduation earnings for this population. This article presents findings from the 2017 U.S. Census Bureau’s American Community Survey on the prevalence of individuals with cognitive disabilities who have attended various levels of postsecondary education and a series of linear regressions identifies the relationship between varying degrees and earnings while controlling for age, sex, race, ethnicity, public support systems, weeks worked each year, and hours worked each week. Results indicated that some college without a degree (14.815%), an associate’s degree (35.831%), a bachelor’s degree (68.267%), and advanced degrees (106.063%) all provide substantial earnings increases over individuals who received a high school degree or less. Findings include policy and practice implications to improve supports and services to increase access to postsecondary education for individuals with cognitive disabilities.

Objective. Cardiovascular disease (CVD) risk is elevated in patients with systemic lupus erythematosus (SLE) and diabetes mellitus (DM), but whether risk of CVD in patients with SLE is as high as in those with DM is unknown. The present study was undertaken to compare CVD risks between patients with SLE and DM and general population US Medicaid recipients. Methods. In a cohort study, we identified age-and sex-matched adults (1:2:4) with SLE or DM and those from the general population using Medicaid Analytic eXtract, 2007-2010. We collected data on baseline sociodemographic factors, comorbidities, and medications. We used Cox regression models to calculate hazard ratios (HRs) of hospitalized nonfatal CVD events (combined myocardial infarction [MI] and stroke) and MI and stroke separately, accounting for competing risk of death and adjusting for covariates. We compared risks in age-stratified models. Results. We identified 40,212 SLE patients, 80,424 DM patients, and 160,848 general population patients; 92.5% were female, and the mean ± SD age was 40.3 ± 12.1 years. Nonfatal CVD incidence rate per 1,000 person-years was 8.99 for patients with SLE, 7.07 for those with DM, and 2.36 for the general population. Nonfatal CVD risk was higher in SLE compared to DM (HR 1.27 [95% confidence interval (95% CI) 1.15-1.40]), driven by excess risk at ages 18-39 years (HR 2.22 [95% CI 1.81-2.71]). Patients with SLE had higher risk of CVD compared to the general population (HR 2.67 [95% CI 2.38-2.99]). Conclusion. SLE patients had a 27% higher risk of nonfatal CVD events compared to age-and sex-matched patients with DM and more than twice the risk of the Medicaid general population. The highest relative risk occurred at ages 18-39 years. These high risks merit aggressive evaluation for modifiable factors and research to identify prevention strategies.

The coordination of transportation and land use (also known as “smart growth”) has been a long-standing goal for planning and engineering professionals, but to this day it remains an elusive concept to realize. Leaving us with this central question -- how can we best achieve transportation and land use coordination at the corridor level? In response, this report provides a review of literature and practice related to sustainability, livability, and equity (SLE) with a focus on corridor-level planning. Using Caltrans’ Corridor Planning Process Guide and Smart Mobility Framework as guideposts, this report also reviews various principles, performance measures, and place typology frameworks, along with current mapping and planning support tools (PSTs). The aim being to serve as a guidebook that agency staff can use for reference, synergizing planning insights from various data sources that had not previously been brought together in a practical frame. With this knowledge and understanding, a key section provides a discussion of tools and metrics and how they can be used in corridor planning. For illustration purposes, this report uses the Smart Mobility Calculator (https://smartmobilitycalculator. netlify.app/), a novel online tool designed to make key data easily available for all stakeholders to make better decisions. For more information on this tool, see https://transweb.sjsu.edu/research/1899-Smart-Growth-Equity-Framework-Tool. The Smart Mobility Calculator is unique in that it incorporates statewide datasets on urban quality and livability which are then communicated through a straightforward visualization planners can readily use. Core sections of this report cover the framework and concepts upon which the Smart Mobility Calculator is built and provides examples of its functionality and implementation capabilities. The Calculator is designed to complement policies to help a variety of agencies (MPOs, DOTs, and local land use authorities) achieve coordination and balance between transportation and land use at the corridor level.

The state of Illinois has had one of the highest rates of outbound migration in America. This paper evaluates the impact of out-migration on the communities these people leave behind, in particular the financial hub of Chicago and the small city of Bloomington. These cities are compared to the growth city of Phoenix, whose population has exploded in this decade and is one of the most popular destinations for those migrating out of Chicago and Bloomington. Human capital theory suggests that highly educated people with high wage potential are more likely to migrate than less educated people. This paper uses the American Community Survey Census database to test the hypothesis that communities with outbound migration (Chicago and Bloomington) face increases in poverty and lower levels of educational attainment on average than the cities with inbound migration (Phoenix). This hypothesis is explored through differencein-difference and OLS regression analysis of poverty, education, and standard of living variables.

Over the past half century, public preschool has emerged as a leading policy remedy for opportunity and achievement gaps. Disparities between low- and high-income students are large—at least three years of learning—and persistent (Hanushek et al. 2020; reardon, Robinson-Cimpian, and Weathers 2015). Yet, socioeconomic gaps at kindergarten entry—before the start of formal schooling—have declined in recent years (reardon and Portilla 2016). Researchers have proposed several explanations to explain this paradox, including more enriching home environments; expanding federal, state, and local investments in public preschool; and rising preschool quality and effectiveness (Bassok, Latham, and Rorem 2016; reardon and Portilla 2016; Yoshikawa et al. 2013). To take advantage of these expanding investments, rising quality, and growing effectiveness to improve children’s early educational experiences at scale, families first need access. The evidence base on preschool choice and access, as well as enrollment and continued participation, is growing (Tang, Coley, and Votruba-Drzal 2012; Greenberg, Michie, and Adams 2018). A recent study of application and enrollment behavior in Boston suggests that children of color, children from low-income households, and dual language learners apply to public preschool at lower rates than their peers, even in a universal preschool context (Shapiro et al. 2019). This study raises concerns about barriers to access and highlights a basic challenge in meeting public equity goals.

Before Covid-19, US unemployment was at a 50-year low. But even an historically tight labor market had notably more slack in it for brown and black populations, in cities across the country. In Minneapolis, for example, the unemployment rate declined from 3.3% to 3.1% between 2000 and 2018, according to data from IPUMS, a database housed at the University of Minnesota that integrates census and survey data. Among white people, it dropped from 2.5% to 1.9%. But among black people, unemployment rose from 6.8% to 8%.

The theoretical literature has predicted that inequality affects long-run growth by reducing human and physical capital, particularly in the presence of imperfect credit markets and other contractual frictions. We test these four mechanisms using measures of inequality at the country-level, dating as far back as the 1700s, and the 1800s, and data for 27 manufacturing industries across 88 countries during 1981-2015. Our findings show industries that are more dependent on financial markets experience lower long-run growth in real output, number of firms and real salaries in more unequal countries compared to more egalitarian. Similarly, industries intensive in physical capital experience lower growth in salaries in highly unequal countries. However, there is no evidence that industries intensive in human capital experience any differential growth in output, the number of firms, average number of employees or salaries in unequal countries compared to more egalitarian, suggesting that the progress made in public schooling provision could have lessened the effect of inequality. Moreover, industries with complex contractual arrangements experience lower growth in the number of firms and paradoxically higher growth in the number of employees hired in more unequal countries, in line with the predictions of the theoretical literature. These findings are robust to using contemporaneous indicators of inequality and instrumental variable specifications. Abstract The theoretical literature has predicted that inequality affects long-run growth by reducing human and physical capital, particularly in the presence of imperfect credit markets and other contractual frictions. We test these four mechanisms using measures of inequality at the country-level, dating as far back as the 1700s, and the 1800s, and data for 27 manufacturing industries across 88 countries during 1981-2015. Our findings show industries that are more dependent on financial markets experience lower long-run growth in real output, number of firms and real salaries in more unequal countries compared to more egalitarian. Similarly, industries intensive in physical capital experience lower growth in salaries in highly unequal countries. However, there is no evidence that industries intensive in human capital experience any differential growth in output, the number of firms, average number of employees or salaries in unequal countries compared to more egalitarian, suggesting that the progress made in public schooling provision could have lessened the effect of inequality. Moreover, industries with complex contractual arrangements experience lower growth in the number of firms and paradoxically higher growth in the number of employees hired in more unequal countries, in line with the predictions of the theoretical literature. These findings are robust to using contemporaneous indicators of inequality and instrumental variable specifications.

The demand for a skilled workforce is increasing even faster than the supply of workers with college degrees – the result: rising wage inequality by education levels, and firms facing a skills gap. While it is often assumed that increasing the number of college graduates is required to fill this gap, this paper explores the extent to which workers without BA college degrees can help fill this gap. To find workers without BA degrees who are potentially skilled through alternative routes (STARs), we use data on the skill requirements of jobs to compute the “skill distance” between a worker’s current occupation and higher wage occupations with similar skill requirements in their local labor market. Based on our calculations, of the 16 million non-college educated workers with skills for high-wage work (> twice median earnings), 11 million whom we term “Rising STARs” are currently employed in middle-to low-wage work. We propose a general taxonomy for STARs to identify potential job transitions to higher wage work within their current earnings category and across earnings categories.

U.S. labor markets have experienced transformative change over the past half century. Spurred on by global economic change, robotization, and the decline of labor unions, state labor markets have shifted away from an occupational regime dominated by the production of goods to one characterized by the provision of services. Prior studies have proposed that deterioration of employment opportunities may be associated with the rise of substance use disorders and drug overdose deaths, yet no clear link between changes in labor market dynamics in the U.S. manufacturing sector and drug overdose deaths has been established. Using restricted-use vital registration records between 1999-2017 that comprise over 700,000 drug deaths as well as data on opioid-related hospitalizations, I test two questions. First, what is the association between manufacturing decline and drug overdose mortality rates and opioid-related hospitalizations? Second, how much of the increase in these drug-related outcomes can be accounted for by manufacturing decline? The findings provide strong evidence that restructuring of the U.S. labor market has played an important upstream role in the current drug crisis. 93,000 overdose deaths for men and up to 34,000 overdose deaths for women are attributable to the decline of state-level manufacturing over this nearly two-decade period. These results persist in models that adjust for other processes changing at the same time, including the supply of prescription opioids. Critically, the findings signal the value of policy interventions that aim to reduce persistent economic precarity experienced by individuals and communities, especially the economic strain placed upon the middle class.

Entrepreneurship is sometimes portrayed as a cure-all solution for poverty reduction. Proponents argue it leads to job creation, higher incomes, and lower poverty rates in the cities in which it occurs. Others, by contrast, posit that many entrepreneurs are actually creating low-productivity firms serving local markets. Yet, despite this debate, little research has considered the impact of entrepreneurship on poverty in cities. This paper addresses this gap using a panel of US cities for the period between 2005 and 2015. We hypothesise that the impact of entrepreneurship depends on whether it occurs in tradeable sectors-and, therefore, is more likely to have positive local multiplier effects-or non-tradable sectors, which may saturate local markets. We find that entrepreneurship in tradeables reduces poverty and increases incomes for non-entrepreneurs. The result is confirmed using an instrumental variable approach, employing the inheritance of entrepreneurial traits as an instrument. In contrast, while there are some economic benefits from non-tradeable entrepreneurship, we find these are not large enough to reduce poverty.

Using variation in crop prices induced by large swings in demand World War I, we examine the fertility response to increases in crop revenues during the period 1910-1930. Our estimates from samples utilizing both complete count decennial census microdata and newly collected county- level data from state health reports indicate that a doubling of the agricultural price index reduced fertility by around 8 percent both immediately and in the years following the boom. We further document that this effect was more pronounced in more agrarian areas and where the labor intensity of agriculture was more intense. Extensive robustness checks and analysis of potential mechanisms indicate that the decrease in fertility was driven by increased female opportunity costs which dominated any household income effects resulting from the price boom.