Using customized survey data from the US and Germany, we study how respondents perceive the impact of recent technological advances on the labor market. We document that a majority views digital technologies and automation as a major threat to overall employment and as a cause of rising inequality, while a quarter is concerned about their own labor market prospects. Providing scientific evidence on the likely labor market implications of automation in a randomized experiment reduces these concerns (on average). Yet, treatment responses depend on prior beliefs about the future of work. This translates into heterogeneous and even opposing treatment effects of policy demand.

Recent work has shown that the words used in the Storm Prediction Center’s convective outlook are not easily understood by members of the public. Furthermore, Spanish translations of the outlook information have also been shown to have interpretation challenges. This study uses survey data collected from the Severe Weather and Society Spanish Survey, a survey of Spanish speakers across the United States, to evaluate how U.S. residents receive, understand, and respond to weather forecasts and warnings. For this experiment, respondents were tasked with ranking the words and colors used by the SPC convective outlook. They were randomly assigned 1) the words originally used by the SPC for Spanish translations or 2) a set of words suggested by linguistic experts familiar with Spanish dialects in the United States. We find Spanish speakers have similar challenges to English speakers when ordering the words the SPC uses. When using the translations proposed by the linguistic experts, we find the majority of Spanish speakers ranked the words in the intended order of associated risk. Spanish speakers also displayed similar ranking distributions for the colors in the outlook as English speakers, where both groups ranked red as the highest level of risk. These findings suggest the original translations used by the SPC convective outlook create barriers for Spanish speakers and that the expert translations more effectively communicate severe weather hazards to Spanish speaking members of the public.

This research integrated and evaluated emerging user data sources (Strava Metro, StreetLight, and hybrid docked/dockless bike share) of bicycle activity data with conventional “static” demand determinants (land use, built environment, sociodemographics) and measures (permanent and short-duration counts) to estimate annual average daily bicycle traffic (AADBT). We selected six locations (Boulder, Charlotte, Dallas, Portland, Bend, and Eugene) covering varied urban and suburban contexts and specified three sets of Poisson regression models: city-specific models, an Oregon pooled model, and all cities pooled. Static variables, Strava, and StreetLight complemented one another, with each additional data source tending to improve the model performance. Sites with lower volumes were more difficult to predict, with considerable error in even the best-performing models. City-specific models in general exhibited improved fit and prediction performance. Expected prediction error increased by a factor of about 1.4 when using Strava or StreetLight alone, but without static adjustment variables, to predict AADBT. Combining Strava plus StreetLight, but without static variables, increased error slightly less; by 1.3 times. We also found that transferring the model specifications from one year to the next without re-estimating the model parameters resulted in a 10% to 50% increase in error rates across models, so such transfer is not recommended. The findings from this study indicate that rather than replacing conventional bike data sources and count programs, old “small” data sources will likely be very important for big data sources like Strava and StreetLight to achieve their potential for predicting AADBT.

Texas Realtors analysis shows nearly 20% of people relocating to Texas came from California in 2021; nearly half of all newcomers were under 35

While rising house prices benefit existing homeowners, we document a new channel through which price shocks have intergenerational wealth effects. Using panel data from school zones within a large U.S. school district, we find that higher local house prices lead to improvements in local school quality, thereby increasing child human capital and future incomes. We quantify this housing wealth channel using an overlapping generations model with neighborhood choice, spatial equilibrium, and endogenous school quality. Housing market shocks in the model generate large intra- and intergenerational wealth effects, with the latter accounting for over half of total wealth effects.

This paper studies the manipulation of electoral maps by political parties, known as gerrymandering. At the core of our analysis is the recognition that districts must have the same population size but only voters matter for electoral incentives. Using a novel model of gerrymandering that allows for heterogeneity in turnout rates, we show that parties adopt different gerrymandering strategies depending on the turnout rates of their supporters relative to those of their opponents. The broad pattern is to "pack-crack-pack" along the turnout dimension. That is, parties benefit from packing both supporters with a low turnout rate and opponents with a high turnout rate in some districts, while creating districts that mix supporters and opponents with intermediate turnout rates. This framework allows us to derive a number of empirical implications about the link between partisan support, turnout rates, and electoral maps. Using a novel empirical strategy that relies on the comparison of maps proposed by Democrats and Republicans during the 2020 redistricting cycle in the US, we then bring such empirical implications to the data and find support for them.

Objectives To compare 2 frequently used area-level socioeconomic deprivation indices: the Area Deprivation Index (ADI) and the Social Vulnerability Index (SVI). Methods Index agreement was assessed via pairwise correlations, decile score distribution and mean comparisons, and mapping. The 2019 ADI and 2018 SVI indices at the U.S. census tract-level were analyzed. Results Index correlation was modest (R = 0.51). Less than half (44.4%) of all tracts had good index agreement (0–1 decile difference). Among the 6.3% of tracts with poor index agreement (≥6 decile difference), nearly 1 in 5 were classified by high SVI and low ADI scores. Index items driving poor agreement, such as high rents, mortgages, and home values in urban areas with characteristics indicative of socioeconomic deprivation, were also identified. Conclusions Differences in index dimensions and agreement indicated that ADI and SVI are not interchangeable measures of socioeconomic deprivation at the tract level. Careful consideration is necessary when selecting an area-level socioeconomic deprivation measure that appropriately defines deprivation relative to the context in which it will be used. How deprivation is operationalized affects interpretation by researchers as well as public health practitioners and policymakers making decisions about resource allocation and working to address health equity.

What is the correct interpretation of average moving costs? I show that in the steady-state of a standard moving cost model, average moving costs are proportional to the difference between the Shannon entropy of next period's location and the Shannon information of staying in the same location. Therefore, moving costs are correctly interpreted as a statistic about the modeler's lack of information regarding future moves, but not about the actual cost of moving in the real world. This alternative interpretation helps make sense of the wide range of moving costs in the literature.

Objective: Granulomatous mastitis (GM) is a benign breast disease that can have an extended clinical course impacting quality of life and causing breast disfigurement. Granulomatous mastitis has been studied throughout the world; however, less is known about GM patients in the United States. We aim to identify demographic and socioeconomic factors associated with GM in the United States. Methods: An IRB-approved retrospective case-control study was performed of 92 patients with biopsy-proven GM at two institutions in Los Angeles, California: a safety-net hospital and an academic institution. Age-matched controls were selected from patients presenting for diagnostic breast imaging. Demographic and socioeconomic characteristics were collected. Data were analyzed using univariable test for odds ratios (ORs) with 95% confidence intervals (CIs) and multivariable conditional logistic regression. Results: Patients with GM were more likely to prefer Spanish language (OR 6.20, 95% CI: 2.71%-14.18%), identify as Hispanic/Latina (OR 5.18, 95% CI: 2.38%-11.30%), and be born in Mexico (OR 3.85, 95% CI: 1.23%-12.02%). Cases were more likely to have no primary care provider (OR 3.76, 95% CI: 1.97%-7.14%) and use California Medicaid for undocumented adults (OR 3.65, 95% CI: 1.89%-7.08%). In the multivariable analysis, participants who preferred Spanish language had four times higher odds of GM versus those who preferred English language (OR 4.32, 95% CI: 1.38%-13.54%). Conclusion: Patients with GM may have barriers to health care access, such as preferring Spanish language, being an undocumented immigrant, and not having a primary care provider. Given these health care disparities, further research is needed to identify risk factors, etiologies, and treatments for this subset of GM patients.

The Universal Child Care Act recently proposed in the New York State Senate would begin the creation of a universal child care system in New York State. This would involve a large-scale up of child care service supply, precipitating a large increase in employment in the child care sector, while increasing wages for jobs in that sector. Who will benefit from these new jobs and wages? We use the Levy Institute Micro Model (LIMM) to simulate the distribution of these new jobs and wages to the population of New York State. Two econometric contributions are made to the LIMM which improve dispersion and result in allocations which are more representative of predicted likelihood distributions. The distribution of jobs and wages is found to be highly income-progressive, making it an effective pro-equity and anti-poverty measure. The distribution of jobs and wages favors women, especially women of color, across the state.

Every year, 2 million first-time, full-time undergraduate students enter a degree-granting post-secondary institution in the United States, but more than one-third leave college before obtaining a college degree. While some students drop out of college for personal reasons or to pursue a different career goal, during this report’s sample period, nearly 40 percent of young adults left college because they could no longer afford to stay. The objective of this study is to gain better insight into the relationship among employment, credit constraint, and the college persistence of US 18- to 24-year-old working college students, who represent more than 50 percent of the undergraduate population. To do so, we investigate two interrelated research questions: (1) Does involuntary job loss affect the college-dropout decision of working students, and (2) does access to credit through credit card loans buffer against the liquidity effect of job loss? This report’s analysis shows that job loss has an adverse effect on college persistence for 18- to 24-year-old US working students, that is, whether those students remain in college. Although the effect was minimal during the 2000–08 period, it became significantly magnified after 2008. It is estimated that from the 2009–10 through 2019–20 academic years, involuntary job loss led to a 17 percentage point increase in a working student’s probability of dropping out of college in the next academic year. We find supporting evidence that the stronger effect in the later period reflects college students’ more restricted access to credit card loans after the passage of the Credit Card Accountability Responsibility and Disclosure (CARD) Act of 2009, which imposed tight restrictions on credit extension to individuals who are younger than 21 or older but enrolled in college. For many working students who have difficulty acquiring alternative forms of credit, credit card loans serve as a crucial means of smoothing consumption when income fluctuates. Tightening of the credit card market has a direct impact on these students. This report’s findings suggest that employment stability plays a pivotal role in the retention of young working students, and a small contingency fund goes a long way in preventing college dropout due to temporary employment disruptions. While the underlying analysis was conducted using national data, the findings are relevant to New England, where higher education employs 4 percent of the region’s workforce, more than twice the national average. Student retention therefore carries implications not only for the individual students seeking a college education, but also for the vitality of the region’s labor market. An important caveat is that the report’s findings do not imply that credit card loans improve college students’ net welfare. While access to credit card loans improves persistence in the short run for unemployed college students, a large credit card debt leads to other adverse consequences and is unlikely the optimal solution to liquidity issues. Instead, the significance of credit card loans in the personal finance of working students reflects a dearth of alternative income assistance to compensate for short-term earnings loss. Extending timely unemployment assistance to college students through either unemployment insurance or student financial aid programs could potentially insure these students against unforeseen job-loss risks and yield retention benefits. Policymakers concerned with the retention of working college students should consider these options and explore them to a greater degree.

This paper studies the distribution of benefits from the EPA’s NOx Budget Program, a seasonal cap-and-trade market for NOx emissions. We use electricity generating unit-level data together with a triple di↵erence estimator, a new empirical method to estimate heterogeneous treatment e↵ects, and an air pollution transport model to estimate the market’s impacts on ambient concentration of fine particulate matter. We find that the market improved equity outcomes by reducing the Black-white disparity in electricity sector air pollution. Replacing the market with a counterfactual policy where all coal units reduce their emissions by the same percentage leads to fewer avoided deaths and worse equity outcomes compared to the market based policy.

BACKGROUND Segmented assimilation theory asserts that children born to immigrants experience divergent paths of incorporation. While some exhibit substantial gains in well-being, others may fare worse than US-origin whites or their own parents. It is certainly true that contemporary immigrants find themselves living in a different context than those who arrived in the United States during the early 20th century. However, it remains an empirical question whether the incorporation process has suddenly become segmented. METHODS We select five of the top European sending regions to ask whether socioeconomic outcomes varied between immigrant-origin populations between 1910 and 1930. We use the Integrated Public Use Microdata Series Multigenerational Longitudinal Panel to link men over a 20-year period. Logistic regression is used to predict probabilities of school enrollment in 1910 among US- and immigrant-origin youths. We then rely on a series of OLS specifications to predict the socioeconomic standing of these men in 1930 as well as differences in father–son status. We also compare relative rates of occupational mobility across country of origin. RESULTS We find evidence of intergenerational mobility as well as convergence in economic success. Though some immigrant-origin groups fare better than others (e.g., the Irish and those from the United Kingdom versus Italians and Germans), our results largely align with classical theories of assimilation. To the extent that segmented assimilation occurs, it emerges in the especially low levels of attainment among German-origin youths. CONTRIBUTIONS Our findings raise important questions about studies that investigate segmented assimilation among immigrant-origin youths. We argue that more work is needed to determine whether downward assimilation is a sign of permanent disadvantage or a shortterm consequence from which youths can recover.

The COVID-19 pandemic has changed the way time is allocated towards market and non-market activities. This paper looks at the effect of the pandemic on a key category of non-market activity – leisure. We use micro-data from the American Time Use Survey to see how the pandemic affected leisure in the United States. A seemingly unrelated regression system is deployed to examine how the pandemic affects various leisure activities such as television viewing, socialising, reading/writing, active sports, games, and computer use. Apart from socialising, which was actively discouraged during the pandemic, all major leisure activities saw an increase, which resulted in overall leisure going up in the Covid year 2020 by 5%, compared to 2019. However, the impact is statistically significantly different for different types of leisure activities. Active leisure is being consumed less while passive leisure is being consumed more – pointing towards an increased sedentarization of lifestyles. Since previous studies have shown passive leisure to be negatively associated with health, our result has important implications for the health and well-being of Americans, and if the trend continues, policymakers should actively incentivise the consumption of active forms of leisure.

This study uses Zillow’s ZTRAX property transaction database to investigate variation in hedonic price effects of water clarity on single-family houses throughout the United States. We consider five spatial scales and estimate models using different sample selection criteria and model specifications. Our results indicate considerable spatial heterogeneity both within and across the four U.S. Census regions. However, we also find heterogeneity resulting from different types of investigator decisions, including sample selection and modelling choices. Thus, it is necessary to use practical knowledge to consider the limits of market areas and to investigate the robustness of estimation results to investigator choices.

Working college students represent 57 percent of the 18- to 24-year-old US undergraduate population. On average, compared with their non-working peers, they have lower family income and receive less parental support, and more than half depend on their own earned income to pay for their college education. Furthermore, most working students have no access to commercial loans but hold one or more credit cards, and in many cases, they carry a credit card balance over months. This paper studies the effect of job loss on the college-persistence decision of US 18- to 24-year-old working college students and whether access to credit through credit card loans buffers this liquidity effect of job loss.

It has been shown that individuals socioeconomic status influences their health outcomes. When looking at socioeconomic status an important factor to consider is educational attainment. The higher an individual’s educational attainment the higher socioeconomic status they are in. In this study I evaluate the correlation between educational attainment and quantity of physician visits for a given individual. Investigating the relationship between physician visits and educational attainment provides further insight into the health behaviors and outcomes of individuals. I found that the higher one’s educational attainment the more likely they are to visit the physician. These findings can be used to create health care policies that address disparities within the health care system.

The most powerful solutions to today’s problems can be identified when data is analysed to tell their own stories. This potential can be better exploited when data originating from heterogeneous domains are linked together to improve data quality and facilitate effective decision-making in data mining applications. Entity resolution, the process of linking records that refer to the same entity across one or more data sets, has provided proven advancements in domains ranging from health research, social sciences and genealogy, national censuses, and online shopping, to the domain of crime and fraud detection and prevention. For example, linking the birth, marriage, and death certificates of all individuals in a population to generate family pedigrees can help the health sector with the early identification of hereditary diseases. In recent years, supervised learning methods have propelled to the forefront of entity resolution research because of the high linkage quality that can be obtained with sufficient labelled data. A recent study has, however, found that a random forest based classifier required at least 1.5 million training labels to link two fairly clean data sets of consumer products with a precision and recall of 99%. The process of obtaining such labelled data is often manual and requires extensive human efforts. Therefore, the practical difficulties of applying supervised learning methods usher in the requirement to consider alternatives with limited labelled training data. In this thesis, we propose two main frameworks for addressing the problem of limited training data in entity resolution. We start with a background study and a comprehensive literature review to understand current gaps and research directions. The first is a solution based on transfer learning which allows to exploit labelled data available in a semantically related domain for the classification task of entity resolution. The novelty of our framework is that it can provide improved linkage quality for short structured data, whereas existing transfer learning frameworks for entity resolution are based on deep learning models that provide better results only for unstructured textual data. The second framework is an unsupervised method that does not require any labelled data. Existing unsupervised methods only consider data sets containing basic entities that have static attribute values and static relationships, such as publications in bibliographic data sets. These methods cannot achieve high linkage quality on data sets with complex entities, where an entity (such as a person) can change its attribute values over time while having different relationships with other entities at different points in time. Therefore, we propose an unsupervised graph-based entity resolution framework that is aimed at linking records of complex entities. We then propose a novel method to geocode historical addresses to support our unsupervised entity resolution framework. We finally describe two application case studies of our frameworks that unveil the benefits of the practical impact of our frameworks.

During the COVID-19 pandemic, the U.S. Department of Agriculture waived the certification interview for the Supplemental Nutritional Assistance Program (SNAP), substantially reducing the administrative burden associated with SNAP application for both applicants and agencies. Using primary policy data collected from ten county-administered states, we find that only 27% of counties implemented the interview waiver. Further, models of local decision-making indicate that public health risk, demographic vulnerability and economic need, and political orientation in the county were not statistically significant predictors of waiver use. Finally, we find that the waiver choice did affect SNAP caseloads: using difference-in-difference models that make use of the natural experiment, we find that counties that adopted the SNAP interview waivers experienced a 5% increase in SNAP caseloads.

While it is commonly believed that teachers take more absences than other professionals, few empirical studies have systematically investigated the prevalence of teacher absences in the US. This study documents the level of teacher absences and compares it with other college-educated workers. Using the Monthly Current Population Survey between the 1995 and 2019 school years, we conduct descriptive and regression analysis to estimate the level of teacher absences and the absence gaps between teachers and other college-educated workers. Additional regression analysis using data from the Leave Module of the American Time Use Survey is conducted to explain the gaps in absences between teachers and other observationally similar college-educated workers. The analysis reveals that 7% of teachers are absent at least once weekly, accounting for around 4% of their weekly working time. Compared to observationally similar college-educated workers, teachers take the same, if not less, amount of absences. Further investigation of teachers' absence behaviour indicates that teachers report fewer demands for absences, have fewer paid leaves, and are more likely to attend work despite needing to be absent. We also find that individuals who prefer fewer absences tend to enter the teaching profession. This study adds to the emerging group of research examining the nature, determinants, and consequences of teacher absences using national-level data. Our findings imply that policymakers may be able to use more support programs to increase teacher attendance.