Abstract
The COVID-19 pandemic has exacted a historic toll on Americans’ health and longevity. It has also shaped socioeconomic inequalities along the lines of gender, race, ethnicity, nativity, and class in America. The effects of COVID-19 are evident in the stratified experiences of Americans in work, unemployment, and unpaid labor; in stark inequalities in wealth and income; in the historic expansions and retrenchments in social welfare spending; and in the increase in violence and changes in the criminal justice system. While there has been an outpouring of research on the social and economic consequences of COVID-19, far less work draws together research across these varied, but interrelated, domains. In this introduction, we provide a broad narrative of how the COVID-19 pandemic unfolded in America and reshaped, in some instances fleetingly and in others more permanently, the landscape of socioeconomic inequality in America.
By now, the timing of the coronavirus pandemic in the United States is familiar. The first COVID-19 case in the United States was confirmed by the Centers for Disease Control and Prevention (CDC) in late January 2020. President Donald Trump declared a national health crisis on March 13, after which several states implemented mandatory stay-at-home orders, and businesses, government agencies, and schools across the country closed to in-person activity. The initial wave of infections crested in July 2020, followed by a much larger second wave of infections that peaked in January 2021 (CDC 2021; New York Times 2021).
Meanwhile, the Food and Drug Administration granted emergency use authorization for first the Pfizer and then the Moderna COVID-19 vaccine in December 2020, and a mass vaccination effort commenced shortly thereafter. The combination of vaccinations, continued social distancing, and warmer weather caused new infections to decline to low levels by summer of 2021 (Syal and Miller 2021). This relative lull then gave rise to a third wave of infections and deaths driven by the Delta variant, and largely focused among the unvaccinated (Smith and Bosman 2021) and then the fourth and largest wave through the beginning of 2022 driven by the Omicron variant. As of October 2022, more than one million people in the United States have died from COVID (CDC 2022a). Cumulative crude death rates are generally highest for racial and ethnic minorities. Adjusting for age differences reveals large disparities in the risk of death: the highest age-adjusted death rates are among Native Americans, who are followed by Latinos, Pacific Islanders, African Americans, White Americans, and Asian Americans.1
Incorporating excess deaths driven by delayed medical care, stress, and other residual consequences of the pandemic yields an even greater toll in terms of excess mortality (Findling, Blendon, and Benson 2020).
We know that the pandemic has had a disproportionate impact on institutionalized and marginalized populations who are resource poor and in some instances politically disenfranchised. African Americans and Latinos are disproportionately represented among documented COVID-19 cases and fatalities, in part because of existing disparities in health problems, differential access to health care, and differential exposure to essential work (Wrigley-Field et al. 2020). Many of the largest and deadliest early outbreaks occurred in institutionalized settings such as nursing homes, rehabilitation facilities, state and federal prisons, and local jails (Dosa et al. 2020; Nelson and Kaminsky 2020). The pandemic has hit Native American communities particularly hard because they tend to be located in rural areas with poor access to adequate health services (Hathaway 2021).
Alongside these epidemiological trends were major changes to employment and work arrangements that broke down along the lines of race-ethnicity, class, and gender. Disparate impacts of employment and of an expanded social safety net also reshaped poverty, income inequality, and wealth inequality. We observed differential resistance to public health measures intended to reduce the reproduction rate of the disease as well as differential vaccination hesitancy that appears to break along socioeconomic, racial, and political lines. Finally, the increase in lethal violence in the United States has disproportionately affected specific demographic groups.
The sudden onset of the COVID-19 pandemic and these changes in economic and social outcomes laid bare many of the structural inequalities in American society. Within a few months of the first documented community transmission, nearly one-quarter of the workforce filed for unemployment benefits; low-income workers and those with less flexibility in scheduling and the ability to work remotely disproportionately lost their jobs. Meanwhile, workers deemed essential, from health-care providers to supermarket employees to delivery workers, bore the brunt of the early exposure to infection; others sheltered in place under state and local orders. Early in the pandemic, these unequal labor-market experiences exacerbated inequalities in material hardship, household economic insecurity, and poverty. Over the following year and a half, expansionary fiscal policy (operating through expansions of the unemployment insurance system), enhanced in-kind benefits and transfers (such as earned income and child tax credits), federal renter assistance funds, and direct and repeated stimulus payments likely dulled the adverse effects of the pandemic on the poorest U.S. households.
Beyond differential impacts by socioeconomic status (SES), the pandemic brought into sharp relief the difference in home and work responsibilities by gender and the extent to which women still bear disproportionate responsibility for childcare and other household tasks. The disruption to childcare and K–12 schooling led to larger employment declines among women relative to men and shifted the amount and division of housework and carework.
These rapid transformations have been dramatic and may portend durable change in American society. Looking back over the years since the onset of the pandemic, the contributions in this issue collectively document the socioeconomic impacts, the nature and efficacy of federal and local responses, and the distributional consequences of what is arguably the largest economic and health shock in generations. The articles in this issue provide both a broad portrait of the causal effects and of the lived experiences of the pandemic across a set of key socioeconomic inequality domains.
The articles paint a portrait of early economic distress that disproportionately affected the most vulnerable households and was followed by a massive federal response that largely countered the adverse effects of the pandemic on material hardship. Despite massive increases in unemployment and declining labor-force participation, total disposable income increased considerably, largely in response to increases in transfer income and consequent declines in overall poverty and childhood poverty. Moreover, measures of housing distress, though still common among renter households, such as being behind on rent or being subjected to eviction proceedings, did not appear to increase and in some instances declined over the course of the pandemic largely because of the moratorium on evictions and federal rental assistance. These developments illustrate the power and capacity of the federal government to respond to a crisis of this magnitude when the political will to do so is sufficient.
Simultaneously, we see evidence that the pandemic disproportionately affected working women with children, appears to have permanently altered spatial arrangements of work for relatively higher educated workers, and has at least temporarily buttressed the labor-market power of lower-SES workers. How these changes ultimately play out and reshape living arrangements, internal migration, and power relations has yet to be determined. It does appear, however, that for knowledge workers place of work and place of residence are becoming increasingly decoupled. We have also observed a sizable increase in lethal violence and gun violence more generally that loosely coincides with the onset of the pandemic, especially following the social upheavals unleashed by the murder of George Floyd in May 2020. This increase in lethal violence has been largely driven by increase in homicide rates where African American males are the victims.
The collection of articles in this issue provides a comprehensive assessment of the socioeconomic consequences of the pandemic in the United States. In this introduction, we provide a broad overview of these developments, most of which are developed further in the quantitative and qualitative studies that follow. Specifically, we document and discuss broad developments in work arrangements and employment, income inequality and other measures of economic insecurity, and the increase in violence. Finally, we offer some speculative thoughts on the likely longer-term and even permanent impacts of the pandemic on the U.S. economy.
REMOTE WORK, ESSENTIAL WORK, AND UNEMPLOYMENT
Initially hailed as a “great equalizer” (Adams-Prassl et al. 2020), for much of 2020, the COVID-19 pandemic appeared to be anything but that. The labor-market dynamics of the period exemplify the stark inequalities that emerged along the lines of class, gender, and race-ethnicity. We present a stylized description of the three faces of work during the pandemic—remote work, essential work, and unemployment.
Remote Work
Even though professional workplaces had experimented with telework and flexible schedules (Moen et al. 2017), such arrangements remained relatively uncommon before the COVID-19 pandemic (Galinsky, Bond, and Friedman 1996). In the first months of the pandemic, however, white-collar workplaces made a rapid shift to work-from-home arrangements (Dunatchik et al. 2021). We expect that perhaps one of the most enduring effects of the pandemic will be to reorient the spatial organization of work.
The necessity to socially distance encouraged a wave of organizational, managerial, and technological innovation aimed at facilitating remote work, asynchronous collaboration, and high-quality video communication. Anecdotes abound concerning the profound impacts of these developments on business travel, formal court proceedings, work meetings, and other aspects of the workplace in many sectors (Pearson, Patel, and Wilkes 2021; Tashea 2020). Moreover, these changes have greatly affected the economies of areas such as central business districts that thrive on spatially concentrated workplaces (Florida 2021; Ramani and Bloom 2021).
Of course, from day one of the pandemic it was apparent that not everyone was able to work from home. Frontline workers ranging from health-care workers to grocery store clerks worked throughout and bore much of the risk of providing essential services and keeping the economy moving. Moreover, many retail trade employees and key service employees, given the nature of their jobs, could not work from home.
Data from the U.S. Census Bureau’s Household Pulse Survey (a weekly survey fielded during the pandemic to provide quick gauges of the impact of COVID on the American public) clearly illustrate large SES differences in the proportion of people able to work from home. Figure 1 presents the proportion of respondents indicating that someone in their households worked from home between September 2020 and late February 2021. Regarding interracial and interethnic differences (panel A),2 Asian respondents had the highest proportion of household members working from home, followed by white respondents; black and Hispanic respondents have the lowest values. By educational attainment (panel B), college graduates are by far more likely to live in homes where someone is telecommuting; the least educated adults are the least likely.
Proportion Indicating That Someone in Household Works from Home
Source: Authors’ tabulations based on U.S. Census Bureau Household Pulse Survey (U.S. Census Bureau 2022).
Although the question in the Pulse survey about working from home changes over the waves of this data collection effort (and thus comparability qualifications should be kept in mind), the tabulation for this version of the question for the last available week of the survey provides some indication of the durability of the change. Table 1 presents the proportions of respondents in households where someone works at home, stratified by education, stratified by race-ethnicity, and stratified by education and race-ethnicity. Interestingly, for all racial-ethnic groups, we see a strong relationship between educational attainment and the likelihood of working at home, and the education gradients in working at home are quite similar across racial-ethnic groups. Hence the average differences in the likelihood of working at home should reflect intergroup differences in the educational attainment distributions.
Someone in Household Works from Home, Week of September 29, 2021
Clearly, those who work from home were less exposed to virus transmission, generating clear inequalities in the impact of the pandemic along the socioeconomic and demographic dimensions used in table 1. This was especially important during the prevaccination phase of the pandemic, when the potential health consequences of contracting COVID-19 were more serious. Although we cannot use the Pulse survey to chart the time path of COVID-19 infections, we can estimate the percentage of different subpopulations that indicate a COVID-19 diagnosis by a medical professional. Table 2 presents tabulations by education, race-ethnicity, and the interaction of education and race-ethnicity of the percentage of respondents indicating that they have had COVID-19 as of the end of September 2021. Overall, 18 percent of respondents indicated that they have had COVID. The percentage is the highest for Hispanic respondents, followed by other, black, white, and Asian respondents. Although the relationship is not monotonic, we see that people with more education generally have lower percentages of infections. Comparing tables 1 and 2 suggests that infections rates are lower among demographic and educational groups that indicate they are more likely to be in households where adults are working from home.
Have Had COVID, Week of September 29, 2021
This pivot to remote work effectively protected white-collar workers from furlough and layoff during the pandemic. Although the unemployment rate rose to 21 percent among workers with less than a high school degree and to 17 percent among those with only a high school degree, it rose to 8.4 percent among college graduates and to 7 percent among those with a professional degree. Moreover, unemployment rates had essentially returned to pre-pandemic levels for workers with at least a college degree by October 2021, yet remained elevated above their pre-pandemic lows for less-educated workers (FRED 2021b).
Even though the ability to work from home shielded many professional workers from job loss and may have protected them from infection, remote work was a double-edged sword. Professional workers, long subject to intense cultures of office-face-time and limited work-life flexibility (Schieman, Glavin, and Milkie 2009), now reported greater flexibility and schedule control (Parker, Horowitz, and Minkin 2020). With widespread remote work, however, also came exposure to the well-documented drawbacks of spillover, intensification, and the blurring of work-life balance (Chesley 2005; Kelly and Moen 2020; Schieman, Glavin, and Milkie 2009). During COVID-19, it appears that work hours increased for remote workers (Fan and Moen 2022), making it difficult to maintain a family-life balance and leading mothers to exit the labor force (Alon et al. 2020; Collins et al. 2020, 2021; Landivar et al. 2020)
Essential Work
The story for lower-SES workers is more complex and more precarious. At the outset of the pandemic, many frontline workers were deemed essential, including those in health care, transportation, food processing, and public safety, as well as service-sector workers in grocery, pharmacy, and logistics (Kane and Tomer 2020). Although some frontline workers, such as physicians, are socioeconomically advantaged, essential workers are disproportionately low-wage earners (Kane and Tomer 2020) and people of color (Fremstad, Brown, and Rho 2020; Huang et al. 2020). Initially a fairly clear distinction, as the pandemic wore on, the line between essential workers and all other workers whose jobs forestalled remote work became increasingly blurred. Against a backdrop of widespread household economic insecurity, workers’ whose establishments had reopened, whether casual dining establishments or retail stores, often had no choice but to return to in-person work.
These workers faced significant direct risks of COVID-19 infection from staying on the job (Chen et al. 2021), and in the early days of the pandemic, struggled to access enough personal protective equipment, such as masks (Harknett and Schneider 2020; Kamerow 2020). Unlike many white-collar workers who were able to work remotely, frontline staff worked in close proximity to others and had little capacity for social distancing (Hammonds and Kerrissey 2020b; Ho, Harknett, and Schneider 2020). Staffing the frontline, these workers also contended with emotional, contentious, and at times dangerous interactions with clients and customers. Frontline workers helped the sick, the very young, and the very old in carework settings, enforced mask mandates in stores and restaurants, and in general dealt with a polarized and often angry public (Hammonds and Kerrissey 2020a).
Despite being hailed as frontline heroes, workers received only modest material recognition of their efforts. Some large firms introduced “hero pay” or bonus pay (Stateler and Kinder 2021), but these compensation supplements were generally short lived (Du, Stateler, and Kinder 2020). Before the pandemic, service-sector workers were among the least likely to have access to paid sick leave through their employers (Schneider and Harknett 2020) and outside a patchwork of state and local laws, paid leave for workers was not mandated (Drie 2021). This lack of paid leave posed both a personal risk to in-person workers and a public health risk because sick workers without paid leave could not stay home while they were sick. The federal response, the Family First Coronavirus Response Act (FFCRA), was historic in being the first federally mandated paid sick leave in the United States and helpful in reducing COVID-19 infections (Pichler, Wen, and Ziebarth 2020), but it was limited. The FFCRA exempted employers with fewer than fifty or more than five hundred workers, leaving frontline service-sector workers at the nation’s largest firms without guaranteed coverage (Schneider and Harknett 2020). Voluntary expansions of personal sick leave, such as at Olive Garden, helped workers to stay home while sick, but such expansions were unusual (Schneider, Harknett, and Vivas-Portillo 2021).
Unemployment
In parallel to the experiences of essential and in-person workers, as Americans heeded public health guidance to stay home and public health orders restricted business operations in order to arrest the spread of the virus, millions of other workers saw dramatic reductions in work hours and spikes in furloughs and unemployment, which were concentrated among generally low-wage workers in hospitality, food service, and retail (Bartik et al. 2020; Gould and Kassa 2021).
Figure 2 presents monthly unemployment rates by race and ethnicity (panel A) and by broad levels of educational attainment (panel B) for the period from January 2019 through October 2021. The onset of social distancing measures in March 2020 followed several years of historically low unemployment rates for all racial-ethnic groups as well as groups stratified by level of education. By April 2020, we observe sharp increases in unemployment for everyone, though the increases were particularly large for Hispanic and African American participants as well as for those with less than a college education. The onset of the pandemic also led to a reversal of racial-ethnicity disparities in unemployment. Although the black unemployment rate exceeded the Hispanic in all months before April 2020, the Hispanic rate rose to the top for a month before dropping below the black rate once again. This likely reflected greater representation of Hispanic workers among industries hardest hit by social distancing measures. Increases in the unemployment rates were particularly large for those with less than a high school education, exceeding 20 percent in April 2020, and also very high for those with a high school degree. Rates have declined considerably since, but for racial and ethnic minorities as well as the less educated, current unemployment rates still exceed pre-pandemic levels.
Unemployment Rates
Source: BLS 2021a.
Note: Race-ethnicity (sixteen and older); educational attainment (twenty-five and older).
Figure 3 further documents disparities in job loss by households employing data from the Household Pulse Survey. The figure displays the proportion of respondents indicating that someone in the household experienced job loss since March 2020, for each week that this question was asked. Panel A presents results by race-ethnicity of the respondent while panel B presents results by level of educational attainment. Households of Hispanic respondents are by far the most likely to experience some job loss with more than 60 percent indicating as much in several of the survey weeks. This is followed by African American households, Other-multiracial, Asian, and white households. We see higher incidence of job loss among less-educated households, though surprisingly, households in which the respondent had less than high school diploma experienced less job loss than those who self-identify as having some college education but less than a bachelor’s degree. This may reflect a higher proportion of frontline workers among the former relative to the latter.
Proportion Indicating That Someone in Household Lost a Job Since March 13, 2020
Source: Authors’ tabulations based on U.S. Census Bureau Household Pulse Survey (U.S. Census Bureau 2022).
The immediate effects of the COVID-19 crisis were also sharply demarcated by gender. Here we focus on total employment rather than unemployment because Liana Landivar and her colleagues (2023, this issue) present compelling evidence of a disproportionate impact of school closures on labor-force participation among mothers of school-age and younger children. In figure 4 we see that on the eve of the pandemic (February 2020), male and female employment stand at 83.9 and 74.9 million respectively. Both drop sharply between March and April 2020. The decline for women, from approximately 74.9 million to 61.4 million, however, is larger than the comparable decline for men, approximately 83.9 million to 73.7 million.
These inequalities were the reverse of those of the Great Recession, during which male unemployment rates rose from 5 percent in February 2008 to 11.1 percent in October 2009, even as the female rate rose only from 4.7 percent to 8.7 percent over the same period. The result is that female labor-force participation rate fell to the lowest levels since 1985 (FRED 2021a). This decline was most pronounced among black women, whose unemployment rate was 16.4 percent (Holder, Jones, and Masterson 2021).
This reversal in gender inequality in unemployment is in part explained by compositional factors. Although the employment shocks of the Great Recession were most pronounced in male-dominated industries, including construction, management, and manufacturing (Goodman and Mance 2011; Meade 2012), the COVID-19 employment shocks were concentrated in female-dominated industries, such as hospitality, childcare, health care, and education (Alon et al. 2020). It appears that this compositional difference accounts for approximately one-third of the gender gap in unemployment that emerged in the first months of COVID-19 (Albanesi and Kim 2021).
In addition to occupational segregation, widespread school and childcare closures and existing gender inequalities in childcare (Bianchi et al. 2012) have interacted to significantly depress women’s labor-force participation. We certainly see evidence in the Household Pulse Survey suggesting that childcare responsibilities have disparately impacted labor-force participation rates for women. Specifically, for people who indicate that they are currently not working for pay, the Census Bureau asks whether they are not working because they have children at home and not in school. Figure 5 presents the proportion of respondents by gender who are not working for pay and who indicate that they are not working for this reason for all weeks when this question is asked. Between 6 and 8 percent of women responding to this question between April 2020 and January 2021 indicate that having to care for children not in school was the reason they were not working for pay. Among men, fewer than 2 percent in each week said the same.
Respondents Not Working Because Children Are at Home
Source: Authors’ tabulations based on U.S. Census Bureau Household Pulse Survey (U.S. Census Bureau 2022).
Note: Respondents between eighteen and fifty; children at home and not in school or daycare.
Social scientists have quickly amassed a detailed literature on the role of school and childcare closures, gender inequality in domestic labor, and inequities in COVID-19 labor-force participation. Looking across the early months of the pandemic, it appears that mothers’ work hours fell more than fathers’ (Collins et al. 2020), perhaps especially among the college educated (Zamarro and Prados 2021). These overall changes in work hours appear to mask even larger gender differences in taking temporary leave from work (Heggeness 2020; Landivar et al. 2020). Although research conducted early in the pandemic found that the decline in employment for mothers was on par with that of childless women and childless men (only fathers experienced smaller drops) (Dias and Buchanan 2020), recent research that looks across the first nine months of the pandemic, through December 2020, in fact finds that mothers stand out as experiencing the largest declines in employment and in hours worked, on the order of a 5 percentage point reduction in the employment-to-population ratio (Fairlie, Couch, and Xu 2021).
Other research contributes to the case that women’s disproportionate increase in labor-force exit is driven at least in part by rising childcare and housework demands in the context of durable inequalities in domestic labor (Bianchi et al. 2012; Sayer 2005). When surveyed, both men and women reported increasing their time spent on childcare and housework; further, many men and even some women reported increasingly egalitarian divisions of domestic labor, at least initially (Adams-Prassl et al. 2020; Carlson, Petts, and Pepin 2021; Dunatchik et al. 2021). However, where men did somewhat more housework and childcare, it appears that women did much more, shifting housework gaps in levels, but not necessarily narrowing them much (Adams-Prassl et al. 2020; Carlson, Petts, and Pepin 2021; Dunatchik et al. 2021; Zamarro and Prados 2021). In fact, in structuring a shift toward more highly gendered domestic production, the COVID-19 pandemic may have shifted gendered attitudes about parenting in a more regressive direction (Mize, Kaufman, and Petts 2021). Correspondingly, among those age twenty-five to forty-four, women were significantly more likely (33 percent) than men (12 percent) to self-report that childcare demands were the main reason for not working during the early months of COVID-19 (Heggeness and Fields 2020). Mothers of young children who used at least forty hours of nonparental childcare before the pandemic and lost that time (only about 5 percent of the sample) were more likely to report not working for pay during the pandemic, whereas such shocks to childcare did not affect fathers’ employment (Petts, Carlson, and Pepin 2020)
In an attempt to derive a more precise estimate of the effects of school closures on employment, Landivar and her colleagues (2023, this issue), construct novel measures of the geography of school operating status, which they merge with American Community Survey and Current Population Survey (CPS) data. They find that mothers’ employment was significantly lower when schools were operating remotely. These effects were most pronounced for mothers with less education as well as for black and Hispanic mothers.
It is important to distinguish this question—whether school and childcare closures disproportionately affected women—from the broader question of whether the lingering reduction in the employment-to-population ratio seen through the end of 2022 is in large part due to the inability of mothers to return to work given ongoing childcare problems. Although Landivar and colleagues (2023, this issue) convincingly show disproportionate impacts of school and childcare closures on women’s work, other research makes a strong case that these impacts do not ultimately explain the persistent fall-off in the employment-to-population ratio (see Furman, Kearney, and Powell 2021).
WEALTH, INCOME, AND ECONOMIC INSECURITY
At the onset of the pandemic, it appeared that all households in the United States were likely to experience losses in both wealth and income. A precipitous decline in the value of equities in the first quarter of 2020 likely reduced wealth levels for the highest income and highest wealth households and the sharp employment declines clearly caused a reduction in wage and salary income for everyone, especially lower-income households. However, over time equities recovered their value with stock prices and equity wealth appreciating considerably over the course of the pandemic. Moreover, massive federal stimulus in the form of direct cash payment, expanded unemployment insurance benefits, an expanded Child Tax Credit, enhanced Supplemental Nutrition Assistance Program (SNAP) benefits, and rental assistance caused an unprecedented increase in transfer payment that more than offset declines in wage and salary income. Because declines in spending were associated with stay-at-home orders, savings actually increased for many U.S. households throughout the income distribution (Babson 2021).
Michael Batty, Ella Deaken, and Alice Henriques Volz (2021) document that over the year 2020, U.S. households experienced a collective increase in wealth of roughly $18 trillion, most of which took the form of higher equity valuations, a source of wealth disproportionately realized by higher-income and higher-wealth households. Using historical relationships between macro-aggregate wealth totals and household wealth by income percentile, the authors project an increase in net worth for the top one percent of households of roughly $1 trillion, while wealth for the bottom 50 percent of households increased by roughly $0.8 trillion. Compositionally, higher equity prices account for the largest share of wealth gains among the richest households; meanwhile, bank deposits, higher pension valuations, and real estate appreciations account for greater proportions of the wealth increases among the bottom 50 percent.
Equity and real estate valuations will have little impact on the poorest households, which often have little savings, little equity wealth, and overwhelmingly rent their homes. Given the large sustained increases in unemployment, one might have expected increased hardship among the poorest households. To be sure, the nation’s unemployment insurance system should partially offset lost wage and salary income, as would other elements of the social safety net, such as the SNAP program, for which, given the contraction in economic activity, more of the population would quality. However, unemployment insurance (UI) coverage before the pandemic was far from universal, benefits are typically limited in time absent emergency extension, and benefit replacement ratios are below one for all workers, and far below for higher-wage workers. Even in conjunction with greater eligibility and take-up of SNAP, the net effect of the sharp increase in unemployment would have squeezed the budgets of many households absent federal intervention.
At the outset of the pandemic, unemployed workers did suffer economically, reporting greater food insecurity, delayed rent or mortgage payments, problems paying utility bills, deferred medical care (Karpman and Acs 2020), and significantly more household financial fragility (Schneider, Tufano, and Lusardi 2020). These shocks were particularly strong for service-sector workers who had contended with low wages before the pandemic (Osterman and Shulman 2011) and generally had few financial reserves to draw on (Schneider, Harknett, and Gailliot 2020). These workers experienced significantly more hardship than those who remained employed (Gassman-Pines, Ananat, and Fitz-Henley 2020; Schneider, Harknett, and Gailliot 2020) and experienced reductions in health and well-being (Schneider, Harknett, and Gailliot 2021).
However, the federal government did respond to the crisis with several rounds of stimulus payments. UI benefits were greatly expanded early in the pandemic to include an additional $600 per week through July 2020, raising the benefit replacement ratio for many well above one. In addition, UI benefits were extended to gig workers and the self-employed, and an additional $300 per week in benefits authorized in March 2021 through the end of summer 2021.
At least initially, many of these workers struggled to access unemployment insurance, contending with administrative burdens (Herd and Moynihan 2020) and limited state UI administrative capacity in the face of a historic flood of claims (Carey et al. 2021). Moreover, as Alex Bell and his colleagues (2023, this issue) demonstrate, access to UI tends to be greater in more affluent states (and affluent areas within a state), is correlated with broadband access and the ability to negotiate the application process online during a pandemic when both offices shut and the number of applications soared, and tended to be lower in areas where the residents are predominantly black and Hispanic. Difficulty accessing UI was higher for low-wage workers and varied across states depending on the strength of the unemployment system before the pandemic (Carey et al. 2021). For those workers who were able to access UI, however, these benefits were substantial, often more than replacing lost wages at the bottom of the earnings distribution (Ganong and Noel 2019) and effectively buffering many workers from both the economic and the health consequences of unemployment (Karpman and Acs 2020; Schneider, Harknett, and Gailliot 2021, 2020). Collectively, this research suggests a considerable role for policy in increasing accessibility to this particularly important element of the social safety net.
Unemployment assistance provided crucial economic support for workers. This sudden generosity in assistance, however, may have also posed a challenge to workers’ identity and self-concept. Alexandrea Ravenelle and Susannah Knoble (2023, this issue), take up precisely this question. Drawing on surveys and on in-depth interviews with nearly two hundred gig and precarious workers, they find that workers were profoundly unsettled by making more on UI than they did when working and that the security the Coronavirus Aid, Relief, and Economic Security (CARES) Act and the Pandemic Unemployment Assistance offered was tempered by deep and, as it turns out, well-warranted uncertainty about the continuity of the programs in the context of fears of future unemployment.
Along with augmented UI, the vast majority of Americans received stimulus checks from the federal government. Three separate stimulus packages—the Coronavirus Aid Relief and Economic Security Act of March 2020 (CARES Act), the December 2020 Stimulus Act, and the American Rescue Plan of March 2021—provided direct cash stimulus to individuals, from $600 to $1,400 per eligible person. In addition, the Child Tax Credit was expanded from $2,000 to $3,000 for children six to seventeen and to $3,600 for children younger than six in March 2021, made fully refundable, made free of any requirement for labor earnings for eligibility, and could be paid in advance in monthly installments. Most parents received Advance Child Tax Credit Payments beginning in July 2021. Together, these programs successfully delivered vital economic support to American families. The scope of these transfers was significant, totaling $819 billion over approximately eighteen months.
At the beginning of the pandemic, demand for food assistance and palpable material hardship were undeniable (Ayllón and Lado 2022; Waxman, Gupta, and Gonzalez 2021), but this constellation of stimulus packages appears to have more than offset any potential losses in income from declining employment. Figure 6 draws on National Income and Product Account (NIPA) data from the Bureau of Economic Analysis. The figure presents the sum of wage and salary income and personal income received via transfer programs for each quarter between the first of 2019 and the third of 2021. Despite a notable break in trend and decline in wage and salary income between the first and second quarters of 2020, the large increase in transfer income more than offsets the decline in wage and salary income. In fact, quarterly wages and salaries plus transfers during the pandemic exceeds pre-pandemic levels for all of the quarters displayed in figure 5. The same is true if one looks at total disposable personal income by quarter (effectively accounting for all other income sources in the NIPA data).
Quarterly Wage and Salary and Personnel Transfer Income
Source: U.S. Bureau of Economic Analysis 2021.
Given the income trends displayed in figure 6, it is not surprising to see relatively stable patterns in measures of distress such as not having enough food or not being current on rent or mortgage payments. Figure 7 presents the proportion of respondents indicating that over the last seven days they sometimes or often did not have enough to eat. This question is asked in a consistent manner in all weeks of the Household Pulse Survey. Despite notable differences across race-ethnicity groups (black and Hispanic respondents are the most likely to indicate that they do not have enough food and white and Asian respondents are the least likely), no notable trends in the response to this question are evident. Similarly, for educational attainment, we see much higher food insufficiency for households in which the respondent has less than a high school education or a high school diploma and lower incident of food insufficiency among households in which the respondent is more educated. However, we do not detect notable trends in this outcome.
Respondents Indicating Not Enough Food
Source: Authors’ tabulations based on U.S. Census Bureau Household Pulse Survey (U.S. Census Bureau 2022).
Despite little evidence of overall trends in these data, figures 6 and 7 do show some evidence of increases in hardship around December 2020, the period after the expiration of the CARES Act, but before the passage of the second relief bill, which is consistent with earlier analysis (Berkowitz and Basu 2021; Cooney and Shaefer 2021; Waxman, Gupta, and Gonzalez 2020).
Similar patterns are observed for the proportion of people who are current on their mortgage payments (figure 8) and the proportion of people who are current on their rent (figure 9). Again, we see that racial and ethnic minority homeowners are somewhat less likely to be current on their mortgage relative to white homeowners, and relatively less educated homeowners are less likely to be current than more educated homeowners. Figure 9 documents similar relative patterns for renter households.
Homeowner Households Indicating Mortgage Payments Up to Date
Source: Authors’ tabulations based on U.S. Census Bureau Household Pulse Survey (U.S. Census Bureau 2022).
Renter Households Indicating Rent Payments Up to Date
Source: Authors’ tabulations based on U.S. Census Bureau Household Pulse Survey (U.S. Census Bureau 2022).
The Household Pulse Survey, however, tracks material hardship only beginning in April 2020, after the onset of the pandemic. If the pandemic significantly and durably increased hardship, then this left-truncation could skew understanding of the effects of the pandemic on hardship. Here, the research is not definitive. On the one hand, comparing estimates of hardship before the pandemic as measured in the CPS and National Health Interview Survey (NHIS) against measures of hardship early in the pandemic measured in the Census Pulse, Diane Schanzenbach and Abigail Pitts (2020a, 2020b) and James Ziliak (2020) find a dramatic increase in hardship through the first months of the pandemic. However, comparing material hardship in December 2019 and December 2020 using the Well-Being and Basic Needs Survey, Michael Karpman and Stephen Zuckerman (2021) find significant reductions in food insecurity, bill payment hardship, problems paying medical bills, and deferring medical care because of costs. The discrepancy between these estimates could be due to an acute period of hardship at the onset of the pandemic, before aid reached the neediest families (Bitler, Hoynes, and Schanzenbach 2020), or to artifacts introduced by the pronounced differences in data collection between the NHIS and CPS on the one hand and the Census Pulse Survey on the other.
Marianne Bitler, Hilary Hoynes, and Diane Whitmore Schanzenbach (2023, this issue) describe the high levels of food hardship in the Census Pulse data and the stark inequalities in hardship by race and ethnicity. They also provide a valuable comparison of rates of food hardship in the 2019 and 2020 December CPS, finding the overall rate of food hardship was unchanged (in contrast to comparisons of CPS and Pulse early in the pandemic) but that this stasis masked substantial heterogeneity: black and Hispanic respondents reported increases in food hardship and white respondents reported reductions.
Further, that the safety net response may have effectively buffered households on average almost certainly conceals the fact that some households disproportionately suffered. Rocío Calvo and Mary Waters (2023, this issue), draw on in-depth interviews with older Latinx immigrants to describe the difficulties this group faced when excluded or deterred from pandemic supports, the adaptive strategies they developed, and the hardships they nevertheless encountered. Marci Ybarra and Frania Mendoza Lua (2023, this issue) analyze longitudinal data from in-depth interviews with forty-two mothers who identified as immigrants or as black and Latina. These immigrant and racialized mothers faced precarious economic circumstances before the pandemic and often struggled to access the safety net, contending with exclusion, disrespect, stigma, and administrative burdens that bred institutional distrust. When the pandemic arrived, mothers who were undocumented faced exclusion from related supports; even some who were eligible struggled to navigate these programs and receive much-needed assistance.
Material hardship measures provide one view of household economic security during the pandemic. The supplemental poverty measure (SPM), which accounts for the kinds of taxes and transfers implemented during the pandemic to aid families that are missed by the official poverty measure, provides another view. Although early projections warned of a sharp increase in poverty (Parolin and Wimer 2020), other analysis indicates that the safety net response was remarkably successful (Bitler, Hoynes, and Schanzenbach 2023). Poverty declined between 2019 and 2020, showing sizable reductions in material poverty among households with children. Direct stimulus payments, expanded UI benefits, and participation in SNAP contributed greatly to these decreases.
In 2020, 9.1 percent of the population was impoverished by the SPM benchmark, a full 2.7 percentage points lower than in the 2019 pre-pandemic year and the lowest poverty rate recorded since 2009 when the SPM were first published (Fox and Burns 2021). The pandemic era programs of stimulus payments and UI benefits played a crucial role, lifting more than 17 million people out of poverty (Fox and Burns 2021; Han, Meyer, and Sullivan 2020). Although it is possible that differential nonresponse to the CPS ASEC, the core source of data for SPM calculations, during COVID-19 might have downwardly biased the estimate of the SPM, the Census Bureau work constructing weights from complete administrative data suggests this was not in fact the case (Rothbaum and Bee 2021). However, because the timing of stimulus payments and of UI expansions was uneven across 2020, the reductions in poverty also varied by month, and were largest from April through July and then smaller in August and September (Parolin and Wimer 2020). Such monthly disaggregation is valuable, but it also appears that households across the income distribution managed to save at least some pandemic support funds—checking account balances of low-income households were as much as 50 percent higher in December 2020 than in 2019 (Farrell et al. 2020), though the external validity of these estimates is likely limited as the data only pertain to JP Morgan Chase customers. This private savings of public support likely allowed many households to smooth essential consumption across the pandemic period.
Peter Hepburn and his colleagues (2023, this issue) and Vincent Reina and Yeonhwa Lee (2023, this issue) provide direct evidence of the effectiveness of federal efforts to stabilize renter households. Using an expansion of federal assistance targeted at supporting households behind in their rent payments as well as federal and state moratoria on evictions, Hepburn and his colleagues document sizable declines in eviction filings, the largest effects concentrated in majority-black neighborhoods. Reina and Lee investigate the effect of a housing assistance lottery in Philadelphia on the likelihood that recipient households are behind in their rent, the extent of their arrears, and the degree to which renter households experience anxiety associated with their housing insecurity. Although the limited assistance certainly did not eliminate the housing distress of households that received assistance, recipient households owed less in back rent, were less likely to have recently borrowed funds from others to pay rent, and were less anxious. Interestingly, the treatment effect on back rent was substantially less than the amount of the grant, suggesting that some of the assistance was allocated toward other households budget categories.
Overall, despite the sharp contraction associated with the onset of the pandemic, the federal government’s direct cash payments, expansion of UI, Child Tax Credits, as well as SNAP and Earned Income Tax Credit (EITC) benefits appears to have more than offset the lost earnings, and likely prevented considerable suffering.
VIOLENCE AND ITS IMPACTS ON CRIMINAL JUSTICE
2020 was a violent year. Most forms of property crimes and violent crime actually declined relative to 2019 (Abrams 2020; Bird, Nguyen, and Grattet 2021).3 Homicide rates, however, increased sharply from 5.1 to 6.5 per hundred thousand. This increase reflects a sharp departure in rates from recent years, as presented in figure 10. Following years of historically low rates, that for 2020 reflects a level of violence not seen since the late 1990s.
This increase in lethal violence was both simultaneously widespread and concentrated. Nearly all regions of the country experienced higher homicide rates (Lopez and Rosenfeld 2021). However, the increases were extremely concentrated demographically. Figure 11 presents homicides per 100,000 for each year from 1999 through 2020 for males (panel A) and for females (panel B) by race-ethnicity.4 Before the pandemic gender disparities in homicide rates were already large (men are much more likely to be murdered than women) and within gender by race-ethnicity (African Americans have the highest homicide rates followed by Native Americans, Hispanics, whites, and Asians). However, figure 11 reveals a notable widening in race-ethnicity disparities in homicide with the onset of the pandemic; increases in rates for African Americans, in particular, African American males, were especially large.
Homicide Rates by Gender
Source: Authors’ tabulation based on CDC wonder data system (CDC 2022a).
The increase in violence has generated much attention and is an active area of research. While the homicide increase roughly corresponds with the timing of the pandemic, a closer look at data on the change suggests that the timing of the pandemic does not align with the increase in violence. Several researchers have documented that city-level homicide counts initially were suppressed by stay-at-home orders and then began to rise during early summer 2020 (Bird, Nguyen, and Grattet 2021; Lopez and Rosenfeld 2021). Similarly, the UN Office of Drugs and Crime documents similar patterns in several countries, homicide rates staying below average for much of 2020, though for most countries not increasing as in the United States (UNODC 2020).
Among the possible explanations for this increase are higher unemployment, declining effectiveness of policing due to either withdrawal of effort or a decrease in cooperation after the murder of George Floyd in May 2020, efforts to defund local police departments, and high proportions of existing officers leaving police work, among other explanations.
Despite the tenuous link between the pandemic and homicide, the pandemic certainly at least temporarily and perhaps permanently changed criminal justice practices and had disparate impacts on correctional populations. Early in the pandemic, correctional institutions saw some of the largest and most quickly spreading COVID-19 outbreaks, leading the National Academies of Sciences to issue a quickly drafted consensus report on steps correctional departments could take to decarcerate to minimize the spread of COVID and facilitate the ability to implement disease control procedures within what are normally quite crowded institutional settings (Wang et al. 2020). Heather Harris (2023, this issue) documents the steps taken within California’s fifty-eight circuit courts to keep criminal case processing moving through remote proceedings, time extensions pertaining to key steps in the adjudication process, and the implementation of a zero-bail schedule for misdemeanors and in some instances less serious felonies. Interestingly, she finds that some of these innovations affected case outcomes (for example, the likelihood of conviction, sentence length) and may have had disparate impacts on criminal defendants from different racial and ethnic groups.
Samantha Plummer and her colleagues (2023, this issue) document how COVID affected New Yorkers incarcerated at Rikers Island. The analysis provides a vivid example of how unprepared the nation’s jails were to address the challenges created by a highly infectious virus and that poor conditions and crowding likely elevated the speed of transmission. The analysis also demonstrates the general poverty and co-occurring challenges that New Yorkers most likely to be incarcerated in the city’s jail system face.
THE GREAT RESIGNATION
By the second quarter of 2021, the labor-market dynamics of widespread unemployment that characterized the beginning of the pandemic had essentially inverted. In June 2021, a government report noted a record high 10.1 million job openings and 3.9 million quits (BLS 2021c). Prominent service-sector firms announced that they were unable to fill openings (Haddon 2021). In all, approximately 1.6 million workers who were in the labor force before the pandemic were not in the labor force as of October 2021 (BLS 2021a) The causes of this rather dramatic reversal are not yet well understood. It appears likely that reductions in immigration to the United States drove some of this shortfall (Smialek 2021) as might have retirements or a slowdown in transitions from retirement back to employment (Omeokwe 2021). As discussed earlier, ongoing challenges of care provision may also have contributed to the shortfall by reducing women’s labor-force participation (Hegewisch 2020).
Beyond these compositional effects, many scholars, policymakers, and worker advocates suggest that the COVID-19 pandemic represented an inflection point in both workers’ tolerance for precarious working conditions and their ability to exercise power to demand better jobs. The evidence for this argument is largely circumstantial. First, it appears that workers are quitting their jobs, especially in the service sector, at a significantly higher rate than before (BLS 2021b). Second, although the job vacancy rate is high and some employers report difficulty filling their positions, firms that offer “good” jobs are apparently facing far fewer problems in recruitment and retention. For example, although UPS has maintained a strong workforce with its higher-paid unionized drivers, FedEx’s low-pay contract worker model has led the company to significant losses due to labor shortages (Black 2021). Third, wages at the bottom of the distribution appear to be rising (Aeppli and Wilmers 2022) and several prominent firms have publicly announced wage increases in response to perceived labor shortages to great fanfare (Rosenburg 2021). Fourth, although still rare given the overall size of the labor market, organizing campaigns have been prominent recently, including at Amazon, which has had mixed results, and at Starbucks, where workers held a series of successful votes through May 2022 (Scheiber and Marcos 2021).
Is this activity consistent with changing worker standards and rising labor power? The former, changing standards, would not be surprising. Workers have long contended with job conditions that were highly precarious, characterized by low wages, limited fringe benefits, and unstable and unpredictable work schedules (Kalleberg 2013; Schneider and Harknett 2021). That the difficult and dangerous conditions of the COVID-19 pandemic, the social comparison to the working conditions of professionals during the pandemic, and the public rhetoric lauding frontline heroes might have sparked a new consciousness among precarious workers is plausible. Since the pandemic, significant anecdotal evidence pointing to workers’ increased standards for their working conditions has emerged (Goldberg 2021; Lipman 2021). Survey data on this phenomenon, however, are still limited. One Pew Research study finds that since the pandemic, two-thirds of unemployed Americans have seriously considered changing their occupation or field of work and one-third have already taken steps to reskill for new types of work (Parker et al. 2020). Although Pew does not have earlier data on this survey question, it did report in 2016 that 80 percent workers were either somewhat or very satisfied with their jobs, suggesting a change in worker expectations (Pew Research Center 2016).
Rising worker power is more speculative. The low-SES American workforce is economically insecure and, before the pandemic, had few economic resources to draw on in the event of prolonged unemployment (Schneider, Tufano, and Lusardi 2020). However, the combination of pandemic augmentations of UI, stimulus checks, and even Advance CTC payments could have buttressed household finances and in so doing provided low-SES workers with a degree of holdout power (Casselman 2021; Irwin 2021; Miller 2021). That checking account balances of households in the bottom quartile and the second quartile of income remained almost 40 percent above pre-pandemic levels through late 2021 is consistent with this idea (Greig et al. 2022). Studies that have exploited variation in the expiration of UI supplements, however, find little evidence that such reductions led to any increase in employment (Bunker 2016; Chodorow-Reich and Karabarbounis 2016; Marinescu and Skandalis 2021).
DISCUSSION
Nearly three years into the pandemic, we attempt to draw some conclusions about what it has revealed about inequality and the effectiveness of the response of U.S. public institutions. As is in all emergencies, those with the fewest resources stood to suffer the most from the health risks posed by the emergency and from the economic risks posed by the shutdown. Although we did not have a set of programs in place sufficient to automatically offset the earnings loss from the sharp contraction, federal policymakers exhibited and acted on the will to mobilize resources to limit human suffering to the extent possible. Although certainly poorer households likely suffered increased economic insecurity at the start of the pandemic, the worst was likely avoided through direct stimulus payments, enhancing the generosity of our existing social safety net programs, and targeted efforts to provide in-kind support. The articles in this issue point to the limits of these ambitious programs in reaching all of those in need (Calvo and Waters; Ybarra and Lua; Ravenelle and Knoble; Bell et al.), but also clearly show the broad efficacy of this response for reducing eviction (Hepburn et al.), poverty, and material hardship (Bitler, Hoynes, and Schanzenbach).
Nevertheless, the evidence assembled in this issue is clear that the pandemic reinscribed racial inequality in many respects. The shocks to unemployment were most severe for black and Hispanic workers and remote school operating procedures had the largest effects on reducing black mothers’ employment (Landivar et al). Although the safety net and stimulus response to the pandemic was broad based, undocumented Hispanic immigrants were excluded from these supports and even some Hispanic immigrants with lawful immigration status struggled to access many of these benefits (Calvo and Waters; Ybarra and Lua) as did racialized mothers who at times struggled to navigate the safety net (Ybarra and Lua). Moreover, Bell and his colleagues demonstrate how access to the unemployment insurance system (one of the principal avenues of support, especially during the first phase of the pandemic) seemed to correlate with access to technology, information, and geographically concentrated affluence. Despite broad-based reductions in poverty as measured by the SPM and even though food insecurity declined overall as a result of the enormously successful government response, black and Hispanic families experienced increases in food hardship over the first year of the pandemic (Bitler, Hoynes, and Schanzenbach). As we learned from both New York (Plummer et al.) and California (Harris), criminal justice systems scrambled to address the logistic challenges created by the pandemic as well as the consequences for those most likely to be involved with these systems.
Some aspects of the COVID-19 response, however, also appear to have effectively reduced racial inequality. As Peter Hepburn and his colleagues show, pandemic era supports, eviction moratoria, and emergency rental assistance significantly reduced evictions in large cities, with the largest effects in majority-black neighborhoods, with the result that racial inequalities in eviction were narrowed, even if by no means eliminated.
Will any of these temporary policy choices survive the pandemic? Some signs suggest that certain expansions of the social safety net, the Child Tax Credit, and the expanded EITC for childless adults in particular are popular. But, despite concerted efforts, they as of now have not been instituted permanently. The Child Tax Credit in particular affected a wide swath of children in the country and would go far to reduce material poverty among the poorest households with children if extended and made permanent. Moreover, the steady source of additional income created by the expanded Child Tax Credit, albeit an income transfer that on its own is not enough to raise a family, may improve the bargaining position of low-wage workers with children in the United States, perhaps leading to permanent wage gains at the bottom of the earnings distribution and a reduction in overall inequality. However, the politics of permanency have proven challenging and, as of this writing, the expanded Child Tax Credit was not extended beyond the temporary increase.
The pandemic has likely sped up the transition to greater remote work, especially for workers with more education. The implications of this development for housing markets, internal migration, residential choice, and the nature and health of some of the country’s largest cities are potentially profound. Changes have been dramatic in the labor-market conditions that workers (who generally have less education) face in retail sales, food service, hospitality, transportation, and health-care industries. These workers have seen pronounced wage growth over the course of the pandemic (Aeppli and Wilmers 2022) and evidence of their rising power is clear in the successful union vote at Amazon’s JFK-8 warehouse and in the string of successful organizing campaigns in Starbucks stores across the country throughout 2022. These improvements in working conditions and successful organizing campaigns appear to be the product of tight labor markets, perhaps resulting from reduced immigration (Grittayaphong and Bandyopadhyay 2022), exit or nonrentry of older workers (Quinby, Rutledge, and Wettstein 2021), and the debilitating effects of COVID-19 (Bach 2022). These campaigns, so far focused on a small subset of employers, could spark broader organizing (Biggs 2005) and some are likely to prove persistent. Other working conditions, however, such as the degree of exposure to work schedule instability and unpredictability have not changed over the course of the pandemic (Zundl et al. 2022) and it is not clear how durable this newfound worker power will prove to be given concerted Federal Reserve action in 2022 to reduce inflation.
The focus of this issue and this introduction is on the socioeconomic consequences of the pandemic. We have not addressed the obvious first-order effect on overall mortality. As noted, nearly one million people have died as a direct result of the pandemic, the number of excess deaths attributable to COVID infections and other sources likely to be much higher. The crude-death rate per hundred thousand as of April 2022 (slightly more than two years into the pandemic) stands at nearly three hundred. For comparison purposes, murders per hundred thousand in 2020 (a violent year relative to the low levels of the past decade) stood at 7.8. Hence cumulative documented COVID deaths over the past two years is thirty-eight times the annual murder rate in 2020. As noted, death rates differ considerably across demographic groups, racial and ethnic minorities having higher rates, especially after adjusting for differences in the age distribution, relative to white Americans. This wave mortality leaves behind many grieving relatives and orphaned children who will disproportionately be the children of frontline workers and other adults who were unable to shelter in place and continue with their lives with little interruption. The consequences and trauma of this high human toll will be felt for generations.
FOOTNOTES
↵1. The APM Research Labs’ Color of Covid (Gawthrop 2022) project documents cumulative death rates by race-ethnicity both in terms of crude deaths per hundred thousand as well as age-adjusted deaths per hundred thousand benchmarking against the 2020 age distribution. As of April 6, 2022, the crude death rate from highest to lowest is 447 for Indigenous persons, 344 for Pacific Islanders, 339 for African Americans, 322 for White Americans, 259 for Latinos, and 159 for Asians. The age-adjusted cumulative mortality rates from highest to lowest are 548 for Indigenous persons, 471 for Latinos, 460 for Pacific Islanders, 440 for African Americans, 263 for White Americans, and 195 for Asians. Interestingly, age-adjusting increases the death rate for all groups with the exception of White Americans. Age-adjusting causes the largest increase in death rates for Latinos.
↵2. We code respondents into the following categories: non-Hispanic white, non-Hispanic black, non-Hispanic Asian, non-Hispanic Other-multiracial, and Hispanic.
↵3. Although property crime overall declined sharply between 2019 and 2020, auto theft increased sharply across a wide geographic area (Abrams 2020; Bird, Nguyen, and Grattet 2021). Despite considerable speculation regarding cause, the source of this increase is not well understood.
↵4. Because Hispanics-Latinos may be of any race, the Hispanic-Latino time series is not independent of the other time series depicted in these graphs.
- © 2023 Russell Sage Foundation. Raphael, Steven, and Daniel Schneider. 2023. “Introduction: The Socioeconomic Impacts of COVID-19.” RSF: The Russell Sage Foundation Journal of the Social Sciences 9(3): 1–30. DOI: 10.7758/RSF.2023.9.3.01. We are grateful to Annette Gailliot, Evelyn Bellew, and Ismael Soto for excellent research assistance. Direct correspondence to: Steven Raphael, at stevenraphael{at}berkeley.edu, Goldman School of Public Policy, University of California, Berkeley, 1893 LeRoy, Room 343, Berkeley, CA 94720, United States; Daniel Schneider, at dschneider{at}hks.harvard.edu, Harvard Kennedy School & Department of Sociology, Harvard University, 79 JFK Street, Cambridge, MA 02138, United States.
Open Access Policy: RSF: The Russell Sage Foundation Journal of the Social Sciences is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
REFERENCES
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.