Industrialized swine facilities adversely affect the health and well-being of Eastern North Carolina residents in the U.S.and are an issue of environmental racism. Concentrated animal feeding operations (CAFOs) emit various harmful and noxious airpollutants, including ammonia (NH3). There are limited measurements of CAFO-related air quality, contributing to disputes aroundits severity. We use NH3 vertical column densities from the space-based Infrared Atmospheric Sounding Interferometer (IASI) toreport systematic, distributive inequalities in NH3 column enhancements (ΔNH3 columns), equal to NH3 columns less anobservationally determined tropospheric background. Population-weighted block group-scale ΔNH3 columns are higher by 27 ± 3%for Black and African Americans, 35 ± 3% for Hispanics and Latinos, and 49 ± 3% for American Indians compared to non-Hispanic/Latino whites in Eastern North Carolina (April−August 2016−2021). Surface winds and air temperature influence block group-scaleNH3 distributions, with higher absolute NH3 inequalities for all groups on calm days and for Black and African Americans andHispanics and Latinos on hot days, consistent with effects from NH3 volatization downfield of facilities from, e.g., manure-coveredfields, particles, and other surfaces. ΔNH3 columns correspond spatially with permitted swine facilities, with residents living multiplekilometers from swine CAFOs chronically exposed to elevated NH3. Trends in NH3 columns over 2008−2023 are driven byregional-scale atmospheric processes rather than localized NH3 changes in CAFO emissions. Results are discussed in local decision-making contexts that have broad relevance for air quality issues without protective federal regulatory standards.

The archipelago of Puerto Rico has faced multiple natural disasters, including hurricanes and earthquakes, disrupting the mental health and daily lives of its residents. These disasters, combined with socio-political abandonment, have led to the deterioration of the electrical grid, exacerbating health disparities. This study aimed to explore the linkages between natural and structural disasters, mental health, and energy independence in Puerto Rico. In this setting, the community organization Casa Pueblo has installed solar panels in the homes of people with chronic health conditions. Through ethnographic fieldwork in the town of Adjuntas and qualitative interviews with 45 individuals with different levels of access to solar power, we explore the implications of ongoing natural and structural disasters on the mental health of individuals and how access to solar energy could help mitigate their effects. Results indicate that participants with direct and indirect access to solar panels held positive and hopeful narratives regarding their mental well-being and quality of life. This innovative model of communal energy security through solar power for climate adaptation holds promises for enhancing mental health in the town of Adjuntas and potentially in other regions.

The housing unaffordability and cost-of-living crisis is affecting millions of people in US cities, yet the implications for urban dwellers’ well-being and social reproduction remain less clear. This Article presents a longitudinal analysis of household access to running water—a vital component of social infrastructure—in the 50 largest US cities since 1970. The results indicate that water access has worsened in an increasing number and typology of US cities since the 2008 global financial crash, disproportionately affecting households of color in 12 of the 15 largest cities. We provide evidence to suggest that a ‘reproductive squeeze’—systemic, compounding pressures on households’ capacity to reproduce themselves on a daily and societal basis—is forcing urban households into more precarious living arrangements, including housing without running water. We analyze the case study of Portland (Oregon) to illustrate the racialized nature of the reproductive squeeze under a housing crisis. Our insights reveal that plumbing poverty—a lack of household running water—is expanding in scope and severity to a broader array of US cities, raising doubts about equitable progress towards Sustainable Development Goals for clean water and sanitation for all (SDG 6) and sustainable cities (SDG 11) in an increasingly urbanized United States. Meehan and colleagues study access to running water in large US cities since 1970, finding that the 2008 financial crisis worsened household ‘plumbing poverty’ in many cities. This disproportionately impacted households of color and generally squeezed lower-income households into more precarious living situations.

Single parents are critical members of communities across the country, fundamental to their children’s well-being and the overall economy. In fact, 35 percent of parents in the United States are single parents. Moreover, these proposed policies would disproportionately harm women, who are more likely than men to be single parents and who are more economically insecure to begin with. Those most at risk are the children impacted by these policies, which are based on outdated ideas about families and gender roles.

Importance Abortion bans may lead to births among those who are unable to overcome barriers to abortion. The population-level effects of these policies, particularly their unequal impacts across subpopulations in the US, remain unclear. Objective To assess heterogeneity in the association of abortion bans with changes in fertility in the US, within and across states. Design, Setting, and Participants Drawing from birth certificate and US Census Bureau data from 2012 through 2023 for all 50 states and the District of Columbia, this study used a bayesian panel data model to evaluate state-by-subgroup-specific changes in fertility associated with complete or 6-week abortion bans in 14 US states. The average percent and absolute change in the fertility rate among females aged 15 through 44 years was estimated overall and by state, and within and across states by age, race and ethnicity, marital status, education, and insurance payer. Exposure Complete or 6-week abortion ban. Main outcome and Measures Fertility rate (births per 1000 reproductive-aged females) overall and by subgroups. Results There were an estimated 1.01 (95% credible interval [CrI], 0.45-1.64) additional births above expectation per 1000 females aged 15 through 44 years (reproductive age) in states following adoption of abortion bans (60.55 observed vs 59.54 expected; 1.70% increase; 95% CrI, 0.75%-2.78%), equivalent to 22 180 excess births, with evidence of variation by state and subgroup. Estimated differences above expectation were largest for racially minoritized individuals (≈2.0%), unmarried individuals (1.79%), individuals younger than 35 years (≈2.0%), Medicaid beneficiaries (2.41%), and those without college degrees (high school diploma, 2.36%; some college, 1.58%), particularly in southern states. Differences in race and ethnicity and education across states explain most of the variability in the state-level association between abortion bans and fertility rates. Conclusion and Relevance These findings provide evidence that fertility rates in states with abortion bans were higher than would have been expected in the absence of these policies, with the largest estimated differences among subpopulations experiencing the greatest structural disadvantages and in states with among the worst maternal and child health and well-being outcomes.

The material well-being of veterans and their families is an important, policy-relevant topic. I research this topic through two crucial measures. In Chapters 1 and 2, I focus on wages, and in Chapter 3, I focus on poverty. In Chapter 1, I estimate the effect of veteran status on unconditional quantiles of wages for males across the period 1979-2017. I show there are important changes to the veteran wage differential over time across the entire wage distribution. In earlier periods, veterans enjoyed wage premiums across the entire wage distribution. Beginning in the early 1990s, wage premiums systematically declined across the wage distribution and largely became slight wage penalties across the wage distribution. Starting around 1997 and going forward, a pattern emerged showing that veterans at the low end of the wage distribution experience wage premiums of around 2% to 4%, whereas veterans at the upper end of the wage distribution experience wage penalties of around 4%. I then conduct wage decompositions of these differentials at both the beginning and end of the period of study. The results show that veterans in more recent times have had their wage premiums reduced or are now earning wage penalties compared to nonveterans. The primary reason for this is that more recent veterans have less favorable compositional differences as compared to nonveterans, especially in the upper-end of the wage distribution. Whereas veterans in the past maintained an advantage in characteristics the labor market valued throughout the wage distribution. In Chapter 2, I examine the effect of potential non-random selection into employment on the female veteran-nonveteran median wage gap over the period 2006-2021. I find no evidence iii that selection is contaminating estimates of the female veteran-nonveteran median wage gap over this period. In Chapter 3, I estimate both the effect of veteran status on households’ likelihood of poverty and deep poverty and the effect of service-connected disability on veteran households’ likelihood of poverty and deep poverty. I construct two measures of service-connected disability. First, I construct an indicator for whether a veteran household has any service- connected disability. Next, I construct a categorical measure of the severity of service-connected disability. The latter allows me to test whether service-connected disability has heterogeneous effects depending on its severity. I estimate these effects over the period 2009-2019. The results are consistent with prior literature, indicating that veteran households have a much lower likelihood of poverty as compared to nonveteran households in all periods. Generally, veteran households enjoy a 1.7 to 3.3 percentage point lower likelihood of poverty depending on the year. This advantage does not transfer to the likelihood of deep poverty. Among impoverished households, veteran households generally maintain around a 2 percentage point higher likelihood of being in deep poverty. Veteran households with the presence of service-connected disability have a lower likelihood of poverty by 2 to 2.5 percentage points as compared to veteran households with no service-connected disability, depending on the year. A different pattern emerges for deep poverty. Among impoverished veteran households, those with the presence of service-connected disability have around a 4 percentage point higher likelihood of poverty. This effect is consistent across the period of study. I also find evidence for heterogeneity in the effect of service-connected disability on the likelihood of poverty and deep poverty. There is a strong pattern whereby as the severity of service-connected disability increases, the likelihood of poverty among veteran households declines. The most severe category of service-connected iv disability is associated with a consistent reduction in poverty by around 3 to 4 percentage points (depending on the year) as compared to veteran households with no service-connected disability or a rating of 0%. Veteran households in the second most severe category analogously experience around a 2 to 3 percentage point reduction in poverty. Veteran households in the second lowest severity category analogously experience around a 1 to 2 percentage point reduction in poverty.

This study aims to investigate the impact of financial inclusion and household decision-making on child health outcomes in Nigeria. Specifically, it explores how financial inclusion—measured through mobile phone usage for financial transactions, bank account ownership, and mobile digital payment methods—interacts with women's decision-making power regarding earnings and household purchases to influence child health. Using data from the 2018 IPUMS-NDHS dataset, the analysis considers both rural and urban divides to assess disparities in financial inclusion and its effects on child mortality. The study employs logistic regression analysis to examine the relationship between financial inclusion, household decision-making, and child health. The findings reveal that households with mobile phones for financial transactions have significantly lower odds of child mortality, while bank account ownership has no significant effect. Regional disparities are apparent, with women in rural areas more likely to face financial exclusion and associated negative health outcomes for their children. Additionally, the study finds that women who have decision-making power over household purchases are less likely to experience child mortality in their households. In conclusion, the study emphasizes the need for financial empowerment of women as a critical strategy to improve family well-being and reduce child mortality. Expanding access to mobile financial platforms and enhancing women's decision-making authority are recommended to promote better health outcomes for children in Nigeria.

The COVID-19 pandemic has exposed individuals to various risks, including job loss, income reduction, deteriorating well-being, and severe health complications and death. In Brazil and the U.S., as well as in other countries, the initial response to the pandemic was marked by governmental underestimation, leading to inadequate public health measures to curb the spread of the virus. Although progressively mitigated, this approach played a crucial role in the impacts on local populations. Therefore, the principal aim of this paper is to evaluate the impact of COVID-19 and, indirectly, of the policies adopted by the U.S. and the Brazilian governments to prevent pandemic diffusion on income distribution. Utilizing available microdata and employing novel econometric methods (RIF-regression for inequality measures) this study shows that growth in COVID-19 prevalence significantly exacerbates economic disparities. Furthermore, the impact of COVID-19 on inequality has increased over time, suggesting that this negative impact has been intensifying. In the U.S., results indicate that working from home, the inability to work, and barriers to job-seeking significantly increase inequalities. Although further data are necessary to validate the hypothesis, this preliminary evidence suggests that the pandemic has significantly contributed to increased inequality in these two countries already characterized by increasing polarization and significant social disparities.

Our report focuses on the current state of the American middle class. First, we examine changes in the financial well-being of the middle class and other income tiers since 1970. This is based on data from the Annual Social and Economic Supplements (ASEC) of the Current Population Survey (CPS), conducted from 1971 to 2023. Then, we report on the attributes of people who were more or less likely to be middle class in 2022. Our focus is on their race and ethnicity, age, gender, marital and veteran status, place of birth, ancestry, education, occupation, industry, and metropolitan area of residence. These estimates are derived from American Community Survey (ACS) data and differ slightly from the CPS-based estimates. In part, that is because incomes can be adjusted for the local area cost of living only with the ACS data. (Refer to the methodology for details on these two data sources.) This analysis and an accompanying report on the Asian American middle class are part of a series on the status of America’s racial and ethnic groups in the U.S. middle class and other income tiers. Forthcoming analyses will focus on White, Black, Hispanic, American Indian or Alaska Native, Native Hawaiian or Pacific Islander and multiracial Americans, including subgroups within these populations. These reports are, in part, updates of previous work by the Center. But they offer much greater detail on the demographic attributes of the American middle class.

Rural areas have experienced dramatic economic and demographic change over the past half-century. Since the 1970s, globalization trends have moved manufacturing industries away from rural areas and sometimes outsourced their labor to other countries, while agricultural industries have increasingly used technological advances that have lessened the need for workers (Flora et al. 2015). Industrial restructuring patterns that became notable in the 1990’s also moved many “newer” jobs in specialized and high-paying industries away from rural areas (Center on Rural Innovation 2022; Muro 2020). Concurrently, these trends have led to shrinking economic opportunities within much of rural America and high levels of out-migration among native-born rural residents (Flora et al. 2015). Despite these trends, an increase in foreign-born in-migration to these areas started in the 1990s and continues today (e.g., Farmer and Moon 2009; Lichter and Johnson 2009). More recently, the COVID-19 pandemic created widespread disruptions to employment in the U.S., but little attention has been given to how these disruptions have affected rural areas or the immigrant workers who live there. This two-article dissertation uses American Community Survey data to provide a descriptive picture of industrial prevalence by nativity status in both urban and rural areas five years before COVID-19, and Current Population Survey data to investigate the onset of COVID- 19 and its implications for rural workers (U.S.- and foreign-born) participating in Retail and Hospitality industries and Professional industries. In addition to providing a novel analysis of rural immigrant labor by industry, I find evidence of regional shifts in rural workers’ participation in retail and hospitality work by nativity status during the COVID-19 years. Moreover, this research highlights the enduring importance of race/ethnicity for rural labor opportunities. These findings extend the literature on rural immigrant well-being by examining labor and industry trends across spatial contexts and build on existing industrial restructuring and segmented assimilation theories.

By concentrating large populations in compact geographical regions, cities have emerged as potent catalysts for socioeconomic development and progress. For cities to thrive and prosper, it is imperative that they provide alluring living and working environments, establishing an atmosphere conducive to growth, innovation, and overall well-being. Among the essential factors that directly influence the potential productivity and prosperity of these environments is the availability of clean, uncontaminated air that enables individuals to thrive both physically and mentally. In order to shed light on the impact of urban pollution on the growth of urban output and employment, a comprehensive comparative statics model has been meticulously developed.

This paper delves into the nuanced impact of the COVID-19 pandemic on time use activities in the United States, particularly examining gender and parental status differences. Drawing on data from the 2018–2022 American Time Use Survey, the study analyzes trends in six distinct time use categories: unpaid care work, leisure, employed activities, personal care, childcare, and household activities. The research sheds light on the evolving dynamics within households during the pandemic, emphasizing the potential implications for economic stratification and societal well-being. The COVID-19 pandemic completely disrupted established time use patterns, forcing a reconsideration of ‘traditional’ gender roles and caregiving responsibilities in certain countries. While early studies hinted at a temporary shift toward more equitable distribution of household activities, particularly childcare, this paper scrutinizes these trends over a more extended period. Despite the short-lived increases in fathers’ involvement in childcare during the pandemic, the study finds that the caregiving burden remained disproportionately on mothers and women. Notably, the analysis reveals persistent gender disparities in unpaid care work, with women and mothers spending a disproportionate amount of time on household activities, housework, and caring for children. This unequal distribution of caregiving responsibilities limits a women’s ability to engage in paid work activities, contributing to economic stratification and constraining their financial resources. While some activities demonstrated slight reductions in gender gaps, the overall picture suggests that the pandemic may not bring about lasting changes in how time is allocated within households.

Background: This study assessed the associations between children’s experiences of discrimination based on race/ethnicity and sexual orientation/gender identity and their parents’ anxiety, depressive symptoms, and self-rated health. Methods: Our sample included 3910 parents with at least one child between 12 and 17 years of age. Data were pooled from the 2021 and 2022 waves of the National Health Interview Survey. Path analysis with maximum likelihood (ML) estimation and bootstrapping were used to examine the indirect pathways from child-experienced discrimination to parental self-rated health through parental anxiety and depressive symptoms. Results: Child-experienced discrimination based on race/ethnicity was associated with worse parental self-rated health via pathways through higher parental anxiety (p = 0.003) and depressive symptoms (p < 0.001). Child-experienced discrimination based on sexual orientation/gender identity was also associated with worse parental self-rated health via pathways through higher parental anxiety (p = 0.002) and depressive symptoms (p = 0.001). Conclusions: The results show a strong association between children’s experiences of discrimination and parental self-rated health through pathways of poor parental mental health. Findings demonstrate the need to reduce discrimination directed at children and to facilitate mental health support for parents of children experiencing discrimination to improve family well-being.

Background: Transportation and health are undeniably related topics. Transportation choices impact our physical and mental health, affecting our activity levels and access to different opportunities. In recent years, research on the transportation and health domain has mainly focused on urban settings, while the rural population has been overlooked. Research is needed to explore the complex relationships across transportation-related health determinants in the rural context and support policies prioritizing health in these areas. Methods: A conceptual framework is proposed based on an extensive literature review to identify the factors representing the transportation and health relationships in the rural U.S. Confirmatory factor analyses are employed to identify the variables representing each transportation-related health determinant. A structural equation model (SEM) is then estimated to quantify the relationships between these determinants in the rural U.S. For this endeavor, we collected individual data from an online survey and available secondary spatial data. A policy approach, the ABC PATH, is also proposed to support community health through transportation changes. Results: We find strong and significant associations across all transportation-related health determinants, particularly with physical activity. Decreasing car usage and providing alternative transportation options can increase health and well-being. The proposed ABC PATH policy approach comprehensively promotes a healthy lifestyle and identifies potential interventions, with particular attention to vulnerable groups. Conclusions: This paper presents a conceptual framework and quantifies the complex interrelationships between transportation and health in rural areas. The policy implications suggest that decision-makers must identify all-inclusive solutions and promote active lifestyles, considering the interlinked connections of the transportation-related determinants examined herein.

Employment has grave implications for workers’ well-being. In this work, we explore how Black women are sorted into what we call “vulnerable work.” Vulnerable work includes work characteristics such as wages, hours, health insurance and retirement benefits, as well as work arrangements that often come with less security, predictability, and economic renumeration—namely independent contracting, contract work, temp agency work, and on-call/day work. We use occupational crowding methodology (Bergmann 1974; Gibson, Darity, and Myers 1998; Hamilton 2006) to understand if Black women are under- or overrepresented in occupations that are more vulnerable, taking into account educational attainment and educational requirements of the work. We compare Black women’s representation with multiple groups, acknowledging marginalization at work across multiple identities (Crenshaw 1989): within race (compared with Black men), within gender (compared with white women), and across both race and gender (compared with white men). A few findings are as follows:  Black women are underrepresented in higher-paying occupations across both race and gender—in contrast to white men, white women, and Black men.  Black women tended to be underrepresented in occupations with higher rates of employer-sponsored health insurance compared with white men but overrepresented in occupations with higher retirement coverage rates compared with white women.  Black women are also overrepresented in occupations that have fewer hours compared with white men and Black men.  Findings on alternative work are below: » As shares of independent contractors and contract workers in occupations rise, the representation of Black women falls significantly more compared with white men and Black men (and with white women as independent contractors). Large diversity exists in the types of work that independent contractors and contract workers perform, and future studies could examine these forms of employment. » The share of temp agency workers in occupations (an arrangement with the lowest benefit rates and pay) has a positive and significant relationship with crowding rates for Black women compared with white men. These findings suggest the need for policies that prevent employer discrimination, boost collective bargaining, and provide a greater social safety net, regardless of connection to an employer.

It is well established that living in a high-poverty area often leads to lower levels of well-being for residents. While these deleterious effects of place-based poverty are well-documented, the conceptual mechanisms linking poverty of place to negative outcomes remain debated, and the our understanding of the spatial patterning of poverty remains underdeveloped. In this paper, we problematize simple conceptualizations of the negative impacts of poverty exposure by illustrating the dynamic patterns poverty displays across cities on a daily basis. The vast majority of prior research on poverty of place has relied upon data anchored to place of residence. Thus, poverty rates broadly reflect poverty as it exists at night. This bias toward nighttime statistics leaves us with an incomplete understanding of spatial inequalities because daytime poverty rates can differ markedly from nighttime poverty rates due to work-related commuting patterns. Here, we use novel data from the Census Transportation Planning Products to fully illustrate diurnal patterns in poverty at the census tract level in metropolitan America. Through a combination of descriptive, spatial, and statistical analyses, we show that the majority of census tracts experience changes in poverty throughout the day. Through a series of regression models, we also show that diurnal patterns in poverty are unevenly distributed along the lines of suburbanization, race, economic status, age composition, and industrial structure. Overall, our findings provide analytic insights into properly documenting poverty across space, while further problematizing lingering culture of poverty frameworks.

Older adults are often overlooked in decision making processes despite the detrimental effects on their well-being. The representation of older adults in investigations and initiatives is needed so that the issues they face, today and in the future, can be resolved through partnership and their active involvement. The aim of this article is to describe a participatory process for conducting a needs assessment (NA) and the contributions and opportunities of partnering with older adults to shape community services and support for older adults. Data were collected from 1,863 participants in a midwestern county via focus groups, interviews, and surveys. We describe how participatory principles were enacted, the involvement of older adults influenced the design and contributed to making sure hard-to-reach residents’ voices were included. Finally, we discuss the ways in which a community inclusive multi-method strategy can optimize resource allocation, identify pathways to more effective policymaking that is matched to the needs and interests of its oldest residents, and lead to unanticipated benefits. Single method, non-inclusive approaches can obscure the critical context and exclude perspectives of the most burdened and vulnerable, who are most in need of support from their community.

The 6th Amendment to the United States constitution guarantees all defendants in criminal trials the right to an impartial jury of their peers. The U.S. Supreme Court has ruled impartiality to mean that the pool of potential jurors must reflect a fair cross-section (FCS) of the community, for instance, that the proportion of Black jurors approximates the proportion of Black residents in the jurisdiction in which a case is adjudicated. In this study, we show that the prevailing standards for evaluating the constitutionality of racial and ethnic minority representation in jury pools are impossible to meet in over three quarters of U.S. counties. Further, we demonstrate striking systematic differences in indicators of social and economic well-being between counties where the FCS requirement could be met versus where it would not. We conclude by discussing the implications of current FCS jurisprudence, highlighting how a legal standard meant to reduce bias in the criminal legal system functions to further racial and ethnic inequality.

This study explores how work- and family-sphere gender structures spill over into individuals’ leisure for personal well-being and self-development. Guided by theories of perpetuated gender inequality in work, family, and leisure, this study further brings in theories of labor and leisure and human development to highlight the significance of leisure to the person and conceptualizesthis domain asthe person sphere.Thisstudy expands the bi-sphere work-family framework of gender inequality into a tri-sphere work-family-person framework aiming to further structural-level understandings of how work- and family-sphere gender structures shape individuals’ ostensibly free person sphere through the lens of activity-time allocations. It draws on time diary and socio-demographic data from a representative sample of 31,000 individuals aged 25 to 54 from the 2014- 2019 American Time Use Survey. The findings reveal significant and systematic gender gaps in personal time. Gendered occupations, whether male- or female-concentrated, result in more constrained personal time than gender-neutral occupations. Structural work demands function as a significant mechanism through which male-concentrated occupations, as well as gender-neutral occupations to a lesser extent, influence individuals’ personal time, while not for female-concentrated occupations. Family gender structure shapes inequalities in personal time among typical fathers and mothers, with typical mothers facing more constraints than fathers, household-specialized in-union mothers facing additional constraints than those not, and single mothers facing particular challenges. These findings highlight how gender inequality is reproduced in the ostensibly free person sphere, calling for greater attention to systematic gender inequality beyond the work and family spheres.

Climate change is expected to undermine population health and well-being in low- and middle-income countries, but relatively few analyses have directly examined these effects using individual-level data at global scales, particularly for reproductive-age women. To address this lacuna, we harmonize nationally representative data from the Demographic and Health Surveys on reproductive health, body mass index (BMI), and temporary migration from 2.5 million adult women (ages 15 to 49) in approximately 109,000 sites across 59 low- and middle-income countries, which we link to high-resolution climate data. We use this linked dataset to estimate fixed-effect logistic regression models of demographic and health outcomes as a function of climate exposures, woman-level and site-level characteristics, seasonality, and regional time trends, allowing us to plausibly isolate climate effects from other influences on health and migration. Specifically, we measure the effects of recent exposures to temperature and precipitation anomalies on the likelihood of having a live birth in the past year, desire for another child, use of modern contraception, underweight (BMI < 18.5), and temporary migration, and subsequently allow for nonlinearity as well as heterogeneity across education, rural/urban residence, and baseline climate. This analysis reveals that exposures to high temperatures increase live births, reduce desire for another child, increase underweight, and increase temporary migration, particularly in rural areas. The findings represent clear evidence that anthropogenic temperature increases contribute to temporary migration and are a significant threat to women's health and reproductive autonomy in low- and middle-income countries.