In a world where we routinely overexploit our resources in an unsustainable way, the development of human talent has endless potential. When a person is given an education, it both improves that individual and develops the society as a whole. All humans can contribute to economic, social, and political development; therefore, educational opportunities must be provided to as many people as possible. The importance of education for all is clearly expressed in the United Nation’s Sustainable Development Goal, SDG 2 – Quality of education. The value of education touches upon nearly all aspects of our societies. From the point of view of economic progress, it is centred on our abilities to become more productive. This would include finding information in books and websites, making mathematical calculations, innovating and adopting new technologies, passing on our knowledge and experience to others, and so on. The basis for it is the ability to read (literacy) and count (numeracy). Education, skills, experience, character, and creativity together correspond to what the economists call ‘human capital’. When combined with new technology and modern machines, this human capital makes each worker more productive. The lesson is clear: To grow richer, we need to improve education and skills. In Africa, the history of education and the development of specialised skills in literacy and numeracy can be traced back to ancient Egypt. But although education has a long history on the continent, many African countries today struggle to offer good quality education to their citizens, and the value of the region’s human capital is low overall. Considering the substantial current and future population growth, human talent is a force that could drive socio-economic development in the region. In this chapter, we will discuss how Africans acquired new skills, in the past and in the present. To do so, we begin with a short review of the economic benefits of education and then proceed to a discussion of education in Africa today and in the past. We consider the role of Islam, Christianity and colonisation, and follow the development of education in Africa during the period of independence.

This chapter focuses on the spatial variation of non-local brewery names in the United States, i.e., breweries that are named after places that they are not located in. Yenne (The American brewery, MBI, St. Paul, MN, 2003) suggested that more breweries in the Eastern United States were named for places in Europe, while more breweries in the Western United States were named for Eastern US Cities. Using data compiled from lists of breweries, the author examined all US breweries founded prior to prohibition to determine if this is true. A Getis-Ord Gi* function was used to analyze the data. The analysis found that breweries with European names are clustered in the Northeastern and Midwestern United States, while breweries with non-local US names are clustered in the Western United States.

This thesis employs statistical modeling to answer research questions on the topic of race-mixing (interracial marriage and sex) and crime victimization. First, I used event history analysis of historical data from 1620 through 1959 to examine predictors of the passage of anti- miscegenation laws, with State Identity emerging most consistently as an important factor. Then I used logistic regression of the National Crime Victimization Survey 1992-2019 to test the hypothesis that victims of intimate partner violence (IPV) in mixed-race relationships have a lower risk of reporting their assault to the police compared to victims of IPV in same-race relationships, and found support for it. Finally, I analyzed the data from Wave 1 (1994-1995) of the National Longitudinal Study of Adolescent to Adult Health (Add Health), and found support for my hypotheses that mixed-race students are less-centrally located in their social networks than single-race students (though not for all centrality measures), and also at higher risk of victimization, even after controlling for centrality.

In the latest report on homeownership released by RentCafe, the most recent data from the Integrated Public Use Microdata Series (IPUMS) was analyzed, showing a boom of millennial homeownership nationwide. “After years of facing challenges in a volatile housing market, millennials have finally become a mainstream generation of homeowners, with 52% of them now owning their homes,” cited RentCafe.

Climate change induced extreme weather events will increase in intensity and frequency, leading to longer and widespread electricity outages. As an example, Winter Storm Uri in Texas left over 4.5 million customers without power between 14 and 18 February, 2021. The social justice consequences of these events remain an outstanding question, as outage data are typically only available at the county level, obscuring detailed experiences. We produce a first-of-its-kind unique spatially resolved dataset of interruptions using satellite data on nighttime lights (NTL) to track blackouts at the census block group (CBG) level. Correlating this dataset with demographic data reveals that minority CBGs were 1.5-3 times more likely to suffer from interruptions compared to predominantly white CBGs, whereas income status was positively correlated with the likelihood of interruption. The presence of critical facilities-including police and fire stations, hospitals, and water treatment facilities-in a CBG reduced the chances of interruptions by around 16%, a small difference that does not otherwise explain the disparity among communities. We suggest explanations, test a subset of them, and propose further work needed to explain what drives these disparities.

An analysis of national-level data shows that Muslim nonprofits are relatively younger, smaller, located in more diverse urban settings, and growing more rapidly both in number and in contributions received compared with other religious and nonreligious nonprofits. No prior studies appear to have summarized national-level data on giving to US charitable organizations affiliated with Islam. As a first attempt to address this gap, this descriptive study compares the characteristics of Muslim-affiliated nonprofits to those of Christian-affiliated, Jewish-affiliated, and all other nonprofits using a dataset of e-filed IRS Form 990s with classifications using keywords appearing in organizational names or mission statements. Muslim nonprofits grew in number at a faster rate, were newer, spent less on fundraising as a percentage of total contributions, and received less in total contributions. However, when controlling for other organizational factors, Muslim nonprofits experienced significantly greater growth in total contributions than did Christian, Jewish, or other nonprofits. Finally, Muslim-affiliated organization were more likely to be headquartered in demographically younger, more diverse, urban settings.

For decades, researchers have sought to understand the separate contributions of age, period, and cohort (APC) on a wide range of outcomes. However, a major challenge in these efforts is the linear dependence among the three time scales. Previous methods have been plagued by either arbitrary assumptions or extreme sensitivity to small variations in model specification. In this article, we present an alternative method that achieves partial identification by leveraging additional information about subpopulations (or strata) such as race, gender, and social class. Our first goal is to introduce the Cross-Strata Linearized APC (CSL-APC) model, a re-parameterization of the traditional APC model that focuses on cross-group variations in effects instead of main effects. Similar to the traditional model, the linear cross-strata APC effects are not identified. The second goal is to show how Fosse and Winship’s (2019) bounding approach can be used to address the identification problem of the CSL-APC model, allowing one to partially identify cross-group differences in effects. This approach often involves weaker assumptions than previously used techniques, and in some cases can lead to highly informative bounds. To illustrate our method, we examine differences in temporal effects on wages between men and women in the United States.

This study examines breast cancer screening behavior among subpopulations of Black women in the United States. Binary logistic regression was used to analyze breast cancer screening among a nationally-representative sample (n = 9,783) of Black women—US-born, non-US-born Caribbean, and non-US-born African—from the 2011-2017 National Health Interview Survey dataset. Non-US-born African Black women were less likely to have breast cancer screening, compared to US-born Black women. Among non-US-born Black women, non-US-born Caribbean Black women were more likely to have had breast cancer screening. Differential healthcare access associated with nativity differences among Black women can be informing their breast cancer screening behaviors. Efforts to improve breast cancer outcomes among Black women can benefit from interventions that account for differential access to healthcare and breast cancer screening behaviors among subgroups of Black women.

This thesis examines the impact of a large income-mixing redevelopment program, specifically the HOPE VI public housing demolitions, on neighborhood public safety externalities in Chicago. To identify the causal effect of demolitions on crime reported by local law enforcement, I leverage the variation in the timing of program applications. Using a two-way fixed-effects specification, I find a significant reduction in crime following demolitions in the post-millennium years between 2001 and 2015. This result is generally robust to the inclusion of neighboring demolitions as spatial lags and addressing selection bias through matching. An event study further shows that the decrease in crime is dependent on time elapsed since the first tract-level housing demolitions and largely driven by implementations in entities with relatively higher socio-economic status. Overall, this study provides evidence that the demolitions of public housing units under the HOPE VI program induce a positive change in reported crime within neighborhoods.

I evaluate whether incorporating sub-national trends improves macroeconomic fore-casting accuracy in a deep machine learning framework. Specifically, I adopt a computer vision setting by transforming U.S. economic data into a ‘video’ series of geographic ‘images’ and utilizing a recurrent convolutional neural network to extract spatiotemporal features. This spatial forecasting model outperforms equivalent methods based on country-level data and achieves a 0.14 percentage point average error when forecasting out-of-sample monthly percentage changes in real GDP over a twelve-month horizon. The estimated model focuses on Middle America in particular when making its predictions: providing insight into the benefit of employing spatial data.

More than half of millennials now own homes, according to a recent study conducted by Rent Cafe. Millennial homeowners now outnumber renters in their generation by nearly 52%. In 2022, there were 18.2 million millennial homeowners across 260 U.S. metropolitan areas, compared with 17.2 million renters. The largest share of millennial homeowners was found in Midland, Texas, followed by Provo, Utah.

The first chapter examines the connection between gentrification and urban violence. I demonstrate a positive and plausibly causal relationship between urban redevelopment and gun violence in Philadelphia. As the underlying mechanism, I focus on gentrification’s displacement effect on local drug markets. Treating the city as a spatial network of city blocks and using two-way fixed effects differences-in-differences estimators, I show the gentrification of one block increases violence across the surrounding neighborhood. I find that some 2,400 (8%) of Philadelphia’s shootings between the years 2011 and 2020 can be attributed to spillover effects from the gentrification of drug blocks. This effect is nearly ten times stronger than that observed on blocks without high levels of drug crime. This study also contributes a new empirical measurement of gentrification drawn primarily from property sales, along with building, zoning, and alteration permit issuance and utilizes a novel nearest-neighbor network approach to identify spatial spillover effects. The second chapter formalizes the synthetic difference-in-differences estimator for staggered treatment adoption settings, as briefly described in Arkhangelsky et al. (2021). To illustrate the importance of this estimator, I use replication data from Abrams (2012). I compare the estimators obtained using SynthDiD, TWFE, the group time average treatment effect estimator of Callaway and Sant’Anna (2021), and the partially pooled synthetic control method estimator of Ben-Michael et al. (2021) in a staggered treatment adoption setting. I find that in this staggered treatment setting, SynthDiD provides a numerically different estimate of the average treatment effect. Simulation results show that these differences may be attributable to the underlying data generating process more closely mirroring that of the latent factor model assumed for SynthDiD than that of additive fixed effects assumed under traditional difference-in-differences frameworks. The third chapter is joint work with Dr. Bryan McCannon. In it, we exploit a novel data set of criminal trials in 19th century London to evaluate the impact of an accused’s right to counsel on convictions. While lower-level crimes had an established history of professional representation prior to 1836, individuals accused of committing a felony did not, even though the prosecution was conducted by professional attorneys. The Prisoners’ Counsel At of 1836 remedied this imbalance and first introduced the right to counsel in common law systems. Using a difference-in-difference estimation strategy we identify the causal effect of defense counsel. We find the surprising result that the professionalization of the courtroom led to an increase in the conviction rate, which we interpret as a consequence of jurors perceiving the trial as being fairer. We go further and employ a topic modeling approach to the text of the transcripts to provide suggestive evidence on how the trials changed when defense counsel was fully introduced.

This dissertation consists of three chapters that study the interactions between health insurance and labor markets in the United States, especially in light of the Affordable Care Act (ACA) of 2010. The first chapter examines how labor market mobility affects equilibrium risk composition, premiums, and costs on the ACA health insurance marketplaces. I find that the labor market transition channel reduces adverse selection on these marketplaces: the participation from transitioners lowers average costs by 5% and effective premiums by 2%. However, total government spending on subsidies increases by 34% due to this interaction. The second chapter studies the impacts of the ACA on job mobility and how they vary by worker type. Using two complementary research designs that separately leverage MSA-level and individual-level variations in exposure to employer-sponsored insurance, I find that the ACA increased job mobility, with greater gains for less educated workers. The third chapter analyzed a survey conducted in June 2020 to provide timely insights on the effects of the Covid-19 pandemic on health insurance offering among small U.S. businesses.

This paper uses new data on copyright registration title pages from the Library of Congress (LOC) to analyze the intellectual and cultural development of the United States over 1790-1870. I construct national time series of book production over this period which show an uptake in per-capita terms in 1830, well before the start of the Second Industrial Revolution and the era of "knowledge based progress" (Abramovitz & David, 1996). Matching authors to locations (at the county level) using declassified census data reveals that the spatial distribution of intellectual production in the early 19th century is strongly correlated with inventive activity over 1860-1940 and the evolution of the manufacturing sector. Identification is based on a shift-share type instrument exploiting the large internal migration patterns occurring in this time period. I then use topic modeling to classify books according to subject matter. Contrary to commonly held beliefs, scientific works are not the strongest predictors of the economic trajectories of US counties. Their correlation with manufacturing activity is relatively large in the short-run but disappears over a few decades whereas non-scientific works show an enduring relationship with economic development well into the 20th century. A theoretical model is briefly sketched which rationalizes these results.

Rationale: Black adults experience worse cognitive function than their White peers. Although educational attainment is an important predictor of cognitive function, other aspects of education, including school desegregation, may also shape this relationship. For Black adults who grew up in the U.S. South in the 1950s–1970s, exposure to school desegregation may have altered life course pathways critical for later cognitive function. Objective: We determined if state variation in exposure to school desegregation in the U.S. South was associated with cognitive function at mid-life, if the association varied by race, and if the association remained after adjustment for state-level education quality and respondents’ educational attainment. Methods: We linked historical data on state-level school desegregation to the Health and Retirement Study, a nationally representative sample of U.S. adults aged 50 and older. We restricted our sample to Black (n = 1443) and White (n = 1507) adults born between 1948 and 1963 who resided in the U.S. South during primary school. We assessed three cognition outcomes: total cognitive function, episodic memory, and mental status. We estimated race-stratified linear regression models with cluster adjustment and a final model using state fixed effects. Results: Greater exposure to desegregated primary schooling was associated with higher cognitive function and episodic memory among Black but not White adults. Among Black adults, the association between school desegregation and cognitive function and episodic memory remained after adjustment for state-level education quality and educational attainment. Conclusions: Our findings suggest that state-level school desegregation efforts played a consequential role in shaping the cognitive function of Black adults who grew up in the U.S. South.

Objectives Some state veterans homes (SVHs) received media attention in the wake of the COVID-19 pandemic because of allegations of poor infection control and excess mortality. However, little research has investigated how these facilities differ from community nursing homes (CNHs) and what the geographical trends of these infection and mortality differences are. We aimed to test (1) whether infection was overall lower in SVHs than CNHs, (2) whether mortality was overall lower in SVHs than CNHs, as well as the geographic distribution of nursing home infection and mortality across the United States. Design Retrospective nationwide cohort study. Setting and Participants Skilled nursing facilities in the United States from May 2020 to July 2022 during the COVID-19 pandemic. Methods Using multilevel negative binomial regression, we modeled COVID-19 infection and mortality rates in skilled nursing homes, testing for overall SVH differences from May 2020 to July 2022, placing random effects on counties to calculate adjusted county-level infection and mortality rates. Results SVHs experienced 18% fewer cases but 25% more deaths overall compared with CNHs. Counties with the highest levels of facility infection were situated mainly in Midwestern, Atlantic, and Southern states, with the majority of counties with low infection levels in Central and Western states. Counties with the highest levels of facility mortality emerged in Rust Belt and Midwestern states down to Southern states as well as westward to Central and Western states to a lesser degree. Conclusion and Implications SVHs experienced lower infection levels but higher mortality levels than CNHs, and fewer extremely high infection and mortality rates in counties containing SVHs despite higher mortality risk in SVHs, suggesting unobserved facility-level differences such as gender and age distributions and future research opportunities using more granular geographical aggregations to better understand facility-level SVH risk within the broader neighborhood context.

Purpose: Smokers can spend a substantial amount on cigarettes, potentially constraining their ability to purchase food. We tested the association of smoking cessation and household food security. Methods: Using the Current Population Survey (2001–2019), we longitudinally linked the Tobacco Use Supplement and the Food Security Supplement (n = 71,278). Among adult smokers (n = 13,144), we used modified Poisson regression to model household food insecurity as a function of quit status (continuing smokers vs. recent quitters), adjusting for sex, age, household size, children in the household, and other household smokers. We also used multinomial logistic regression to examine more detailed food security status (high, marginal, low, very low). Results: The adjusted probability of household food insecurity at follow-up was 11% (95% CI: 8.7%–13%) for recent quitters and 20% (95% CI: 19%–21%) for continuing smokers. Continuing smokers had a lower adjusted probability of high food security (69% vs. 80%) and a higher adjusted probability of marginal (11% vs. 9.8%), low (12% vs. 7%), and very low food security (7.8% vs. 3.6%) compared to recent quitters. Conclusions: Cigarette cessation is associated with a lower risk of household food insecurity. Therefore, promoting tobacco cessation alongside food assistance and poverty reduction policies may help alleviate food insecurity.

We examine the role of pay transparency in the productivity of firms’ and inventors’ innovation activities. Our test exploits the staggered adoption of state-level pay secrecy laws, which enhance pay transparency in the workplace. We find a significant increase in inventor productivity of firms located in states that have passed such laws relative to firms elsewhere. This relation is more pronounced for firms in states with lower levels of pre-existing pay transparency. We further show that pay secrecy laws promote inventor productivity by motivating inventors—especially minority inventors—to exert more effort, enhancing the diversity of inventor teams, and encouraging all inventors to pursue promotions.

In the late 19th and early 20th centuries, official policy and doctrine of The Church of Jesus Christ of Latter-day Saints encouraged the “gathering of the faithful.” In fact, Brigham Young declared in 1860 that emigration “upon the first feasible opportunity directly follows obedience to the first principles of the Gospel we have embraced.”1 As a result, thousands of converts to the Church gathered to “Zion” in Utah. During the 19th century, more immigrants came to Utah from Denmark than any other country except Great Britain.2 Not far behind the Danes were the Swedes and a good number of Norwegians. Together, by 1880, these Scandinavian immigrants made up nearly 9% of the young Utah Territory’s total population.3 In some counties, the foreign-born Scandinavians constituted up to nearly a quarter of the population, and in a few communities they were the major-ity.4 This influx of people who came from a foreign culture and spoke a language other than English affected Utah in many ways. This chapter will focus on the influence they had on the onomastics of Utah.In light of the demographic facts mentioned above (and explained in greater detail below), one might expect a strong Scandinavian influence on both personal names and place names in Utah. Nevertheless, while there was certainly some influence, it seems to have been disproportionately small, and thus it is often neglected, forgotten, and unstudied by those interested in Western American and Latter-day Saint naming patterns. For instance, in the chapter on “Proper Names in America” in his monumental work The American Language, H. L. Mencken writes of Scandinavian names elsewhere in America (e.g., Minnesota) but makes no mention of any Scandinavian onomastic influence in Utah.5 Likewise, modern investigators of personal names in Utah have noted the potential influence of French (“The quintessential Utah name often has a French-sounding prefix such as Le-, La-, Ne- or- Va-”6) and religious scriptures (e.g., Book of Mormon names like “Mahonri or Nephi or Moroni”7) on Utah LDS given names. Mansfield even draws parallels between Latter-day Saint and African-American naming patterns.8Nevertheless, these authors make no mention of any Scandinavian influence on anthroponyms in Utah. Regarding Utah toponyms, the situation is similar. The introduction to the 450-page Utah Place Names mentions the influence of the native Shoshone, Ute, Paiute, Goshute, and Navajo Indians; the Spanish explorers; and “the mountain men, Mormons, and military”9 on toponyms in Utah, but it fails to make any reference to the Scandinavians and their influence.

Each year the Internal Revenue Service distributes over $8 billion in low-income housing tax credits (LIHTC) for rental housing development. I map LIHTC housing to school attendance zones to explore program impact on K-12 public schools, employing a rich dataset of campus level expenditures, enrollments, teacher counts, and neighborhood incomes. I estimate that LIHTC housing causes absolute school spending to increase nearly 9% and instructional spending to increase by 5%. My results imply the average LIHTC investment of $1, 332, 222 generates a $452, 328 school spending spillover, equating to 35 cents for every dollar of housing tax credits. The primary mechanism is a mean enrollment increase of nearly 66 students, half of which are likely subsidized under federal lunch programs, and the whole requiring two additional teachers on average. Overall, the absolute spending increase is efficient as relative spending per-pupil does not change after LIHTC construction.