Abstract
The COVID-19 pandemic has been unprecedented in many ways, but perhaps no more so than in the sudden expansion of—and increase in—unemployment assistance benefits. We ask how precarious workers, many of whom were “hustling” for money or engaged in creative fields, feel about making more on unemployment. How are they using the funds? We draw on remote interviews and online surveys with 199 gig and precarious workers in New York City during the first wave of the pandemic. We find that workers are ambivalent about unemployment assistance and concerned that a financial influx today portends a shortage tomorrow. This “specter of the unknown” affected workers’ use of their benefits. As a result, even though the Coronavirus Aid, Relief, and Economic Security Act was intended to mitigate the social and economic impact of the pandemic, these programs—despite being helpful—may have also contributed to precarious workers becoming even more certain of their insecurity.
The COVID-19 pandemic has been unprecedented in many ways, but perhaps no more so than in the sudden expansion of—and increase in—unemployment assistance benefits (DeParle 2021; Edwards 2021; Liu 2020). Under the Coronavirus Aid, Relief, and Economic Security Act (CARES Act) Pandemic Unemployment Assistance (PUA), independent contractors, such as those doing freelance or gig work, were for the first time eligible for unemployment assistance (Brumberg 2020). Partnered with the $600 a week Pandemic Emergency Unemployment Compensation (PEUC), many low-wage workers found themselves in the unique position of making more on unemployment assistance than they had been earning while working (Ganong, Noel, and Vavra 2020). The goal of this support was to inject a “countercyclical stimulus into a declining economy” and encourage workers to stay at home (Burtless 2020).
The COVID recession began suddenly and ended almost as quickly, lasting just two months, becoming the shortest recession on record (Fazio et al. 2021). At the same time, the need to socially distance to “flatten the curve” (Chow and Abbruzzese 2020) and the continuing pandemic led to a service economy meltdown (Porter 2020). Many of the low-wage jobs that unemployed workers might turn to during an economic downturn—such as restaurant server, retail worker, or TaskRabbit assistant—have faced disproportionately high job losses or decreased demand for services (Bateman and Ross 2021; Iacurci 2021; Tomaskovic-Devey, Dominguez-Villegas, and Hoyt 2021). During most of 2020 and into early 2021, with restaurants and bars closed or operating at reduced capacity, entertainment venues shuttered, shops limited to “curbside pick-up,” increasing supply chain issues, and the pandemic ongoing; the opportunity to “grow the economy” was semi-limited (Babson 2021; Horowitz, Brown, and Minkin 2021; FRED 2021). The question arises, during these extraordinary times, have our ordinary expectations regarding unemployment assistance also been turned upside down?
In this article, we draw on a mixed methods panel study of 199 precarious and gig-based workers. Our data includes in-depth remote interviews and online surveys conducted mostly from April 2020 through June 2020, during the height of the first wave of the pandemic, and 168 follow-up interviews conducted between November 2020 and June 2021. We ask how precarious workers, many of whom were “hustling” for money or engaged in creative fields, feel about making more on unemployment. How is this income influx affecting worker perceptions of their jobs or careers and how are workers using the funds?
We find that although many workers were “grateful” for the increased unemployment insurance eligibility and $600 per week supplement, they also felt uncomfortable making more without working for pay. Additionally, respondents also reported fear that the PEUC supplement would not be renewed, a fear borne out by reality for most of the second half of 2020. As a result, the “specter of the unknown” regarding the potential length of the pandemic—and the potentially short-lived nature of the supplement—affected unemployed workers’ use of their benefits. Contrary to the expectation that unemployment assistance is quickly spent, many respondents reported a concerted effort to save a significant portion of their aid in order to survive the months ahead. We also find that some workers who faced extensive delays in receiving unemployment assistance reported long-term feelings of financial insecurity and a sense that they would never again feel financially secure. As a result, although the PUA and PEUC were intended to mitigate the impact of the pandemic on social and economic outcomes, these programs—despite being helpful—may have also had the inadvertent effect of making precarious workers feel even more certain about their financial insecurity. In the following section, we provide a brief history of unemployment assistance in the United States.
THEORETICAL BACKGROUND
The federal unemployment insurance program was created in 1935. One impetus for the creation of state unemployment programs was to reduce the pressure that local communities were experiencing when suddenly unemployed workers found themselves reliant on soup kitchens and community largess. Early literature, created by the Social Security Board (Friedman 1937), explained the program by noting that “unemployment is incalculably expensive. Its cost to workers, to business, to government, and society at large can hardly be exaggerated. . . [unemployment insurance] more equitably distributes a part of the unavoidable cost of unemployment.”
The structuring of state unemployment plans was a compromise between a subsidy plan, where the federal government would grant funds to states that passed laws that met federal standards, and a credit-off set plan, where the tax would be levied on all employers, with credits against the tax available (DeWitt 2003). Employers who maintained stable employment were rewarded with rebates on the theory that unemployment was largely the result of employer decisions (DeWitt 2003). Unemployment was intended to replace 50 percent of salaries, the exact level set by states (Price 1985). In 1955, the secretary of labor recommended a federal policy of covering individuals through twenty-six weeks of unemployment benefits (Price 1985), a six-month standard that remains today.
Unemployment assistance is generally viewed as an automatic economic stabilizer for its role in helping unemployed workers maintain some of their purchasing power and to prevent a more severe economic downturn (Chodorow-Reich and Coglianese 2019; Maggio and Kermani 2016; Vroman 2010). The Center for American Progress describes unemployment assistance as providing “the biggest bang for the buck of the various kinds of government spending.” During the Great Recession, every dollar spent on unemployment insurance benefits “grew the economy by $2 since recipients typically spend—not save—those dollars” (Boushe and Separa 2011).
Research on Unemployment
One of the earliest interview studies focused on the impact of unemployment was Mirra Komarovsky’s (1940) book The Unemployed Man and His Family. Based on Depression-era interviews with fifty-nine relief families in a large industrial city just outside New York City. Komarovsky found that unemployment “made explicit unsatisfactory sentiments” that had existed in families before the Depression and that in many cases affected the husband’s status, mental health, or authority within the family, and affected the couple’s social life.
More recent research focuses on the differing experiences of men and women in their decision-making around attempting to return to work after a lay-off and how they viewed unemployment assistance (Damaske 2020). Intriguingly, Sarah Damaske finds that unemployed white-collar male workers consider unemployment benefits as their due, treating early periods of unemployment as a respite, whereas lower-income workers were more likely to begin job searching immediately. Indeed, work by Katherine Newman (1999, 2008) and Kathryn Edin and Maria Kefalas (2005) makes clear that the vast majority of the urban poor want to work—even when their lives would be easier if they relied on public assistance. In her two-year study, Newman (1999) finds a strong commitment to the work ethic, even as so-called minimum wage “McJobs” offer few chances for advancement.
Although unemployment itself is not new, “repeated layoffs have become part of working life” since the 1980s (Lane 2011). Research suggests that somewhere between 65 to 70 percent of Americans will experience unemployment at least once during their careers, some experiencing numerous bouts (Damaske and Frech 2014). Although unemployment itself can generate and reproduce inequalities between the employed and unemployed, it can also affect inequality among the unemployed (Damaske 2021). As Damaske (2021, 10) writes, “differences in the path one takes to a job loss shape the job loss experience itself.” For instance, Ofer Sharone’s (2014) Flawed System/Flawed Self finds that American white-collar workers blamed themselves for their failed job searches, whereas Israeli white-collar and American blue-collar workers blamed the employment system. Meanwhile Allison Pugh’s (2015) Tumbleweed Society, finds that as part of the “one way honor system,” workers expect that they will remain positive, continue to put forth their best effort, and “move on” without a fuss to the next “opportunity.”
Although workers may be expected to take the “challenge” of unemployment in stride as they move onto the next “opportunity,” much as Komarovsky (1940) found, unemployment in the United States is linked to “deep and intractable hardship,” where unemployment assistance is a “poor substitute for work” and does not fully mitigate the impact of job loss (Young 2012). Research demonstrates that job losses have significant negative effects on incomes. Earnings are estimated to drop by 25 percent in the first year and to linger for years afterward (East and Kuka 2015; Stevens 1997). Job losses also lead to negative impacts on health and mortality and children’s academic achievement (Kuka 2020; Schaller and Stevens 2015; Stevens and Schaller 2011; Sullivan and von Wachter 2009). Part of the challenge is that many unemployment insurance recipients do not find new jobs before exhausting their benefits, even when those benefits are extended (Rothstein and Valletta 2017).
As Chloe East and Elira Kuka (2015) note, during the Great Recession, when more than ten million workers received unemployment assistance, some for as long as ninety-nine weeks, attention increased to the “moral hazard” of unemployment assistance and the optimal level, and length, of benefits (Schmieder, Von Wachter, and Bender 2012; Rothstein, Aaronson, and Kahn 2011; Landais, Michaillat, and Saez 2010). Peter Ganong and Pascal Noel (2019) suggest that extending weeks of benefits leads to more successful “consumption-smoothing” than increasing benefits, whereas Kuka (2020) finds that “higher unemployment insurance generosity” leads to improvements in self-reported health, especially when the unemployment rate is high. Likewise, work by Pinghui Wu and Michael Evangelist (2022) and Jonathan Cylus, Maria Glymour, and Mauricio Avendano (2014) suggests that increasing levels of unemployment assistance may decrease deaths of despair, such as from opioid overdoses or suicide, and “significantly alleviate the adverse health effects of unemployment among men” (Cylus, Gylmou, and Avendano 2015).
Yet, as Johannes Schmieder and Till von Wachter (2016) note, more generous unemployment tends to increase unemployment duration, perhaps by increasing the reserve wage of recipients (Marinescu 2017) or reducing job search efforts (Marinescu and Skandalis 2021). However, during the pandemic, job postings decreased even as applications remained relatively stable after the CARES Act was passed (Marinescu, Skandalis, and Zhao 2020). As a result, the number of applications per vacancy increased, leading to the observation that it may be “socially optimal” to increase benefits in order to decrease job applications (Landais, Michaillat, and Saez 2018). Or as Iona Marinescu, Daphné Skandalis, and Daniel Zhao (2020, 3) write, “when there are too many applicants per job, one person not applying makes no material difference to the job being filled.”
Finally, although research has examined the impact of unemployment on former tech workers impacted by the dot-com bubble (Lane 2011), low-income urban youth (Newman 2008, 1999), unemployed autoworkers (Chen 2015), and married couples (Rao 2020), little is known about how 1099 workers—a group that traditionally is not eligible for unemployment benefits—may be affected by such assistance.
Unemployment, Independent Contractors and the PUA
Independent contractors, gig workers, and freelancers are traditionally excluded from the safety net offered by unemployment insurance under the logic that they seek their own work and are less dependent on the vagaries of a single employer. Additionally, employers classify workers as independent contractors as a savings mechanism to avoid paying into unemployment or Social Security, a strategy that outsources risk to workers (Ravenelle 2019; Hacker 2006). It wasn’t until the March 2020 CARES Act, which established Pandemic Unemployment Assistance, that these workers were included. Independent workers were included in the CARES Act because it included broader language that allowed anyone, regardless of employment type, to receive benefits if they could prove their work was affected by the pandemic (Bellon 2020). The CARES Act, as mentioned earlier, also created the PEUC, which provided an additional $600 a week supplement until the end of July 2020 (Adamczyk 2020).
The weekly supplement of $600 a week was based on the premise that unemployed workers typically receive about 40 percent to 45 percent of their previous earnings. Given a median weekly wage of $1,000, the supplement was intended to ensure that workers did not experience a drastic income fluctuation (Evermore 2020; Long 2020). This relatively generous unemployment assistance had two purposes: to moderate the financial impact of the widespread business closures necessary under social distancing efforts and to reduce the impetus for workers to find a new job quickly (Long 2020). Given the original policy decision to structure unemployment systems as state-run programs, benefits are not consistent, from as low as $101 weekly in Oklahoma to as high as $531 a week in Massachusetts (Stone and Chen 2014; Long 2020). In an effort to account for this variation, the authors of the bill slightly overshot: approximately 76 percent of U.S. workers received an income that exceeded their lost wages due to the pandemic from April to July 2020 (Ganong, Noel, and Vavra 2020). In New York, where the maximum weekly unemployment benefit in New York is $504, the receipt of the PEUC funding more than doubled the unemployment incomes of the workers who received it. As Alex Bell and his colleagues (2023, this issue) note, states with higher levels of benefits—such as New York—had higher rates of unemployment insurance access during the pandemic.
PEUC, Earned Income Credit, Negative Income Tax, and Universal Basic Income
The PEUC benefit of an additional $600 a week, by seeking to approximate the median wage, sought to reduce the impact of widespread and considerable income volatility due to the pandemic. However, extensive delays in the processing of the benefits meant that for some unemployed workers, their initial deposits included thousands of dollars of back benefits, making the program—at first—similar to the Earned Income Credit (EIC). The EIC is a tax credit that was intended to “make work pay” (Edin and Shafer 2015) for low-income earners. Expanded in the 1990s as part of welfare reform, it provides a financial boost to low-wage workers. Research suggests that the credit, which may amount to as much as six weeks of take-home pay (Farrell, Greig, and Hamoudi 2019), is often used to purchase big-ticket items, pay for educational expenses, or to set aside savings (Parker et al. 2013; Beverly, Schneider, and Tufano 2006; Beverly 2002).
However, once payments were regularized, for recipients, the program was similar to the negative income tax (NIT) experiments of the 1970s, a “cousin” to universal basic income (UBI) programs (Marinescu 2018). Whereas a UBI offers universal, unconditional cash payments that are paid to individuals on a periodic basis (Hasdell 2020), NIT schemes include a maximum cash benefit that is reduced by earned income.1 Cash injections into households, such as an NIT, UBI, or even Earned Income Tax Credits, can increase household expenditures (Baird, McKenzie, and Özler 2018; Jones and Marinescu 2018; Marinescu 2018) and have been linked to increased food expenditures and significant impacts on health and educational outcomes, including a reduction in hospitalizations and a decrease in self-reported criminal activity (Hoynes and Rothstein 2019; Marinescu 2018; Forget 2011; Akee et al. 2010). UBI and NIT programs remain controversial, and a research gap remains in the establishment of an experimental, sustained UBI “guaranteed income” program aimed at addressing economic and racial injustices (Hasdell 2020).
Additionally, research suggests that access to a financial safety net can increase rates of entrepreneurship or small business creation. For instance, Gareth Olds (2016) finds that an awareness of the availability of food stamps increased entrepreneurship, suggesting that financial risk from leaving wage employment may be a primary barrier to small business creation. Additionally, revisions to the French unemployment system that allowed workers to remain eligible for benefits while starting a business also contributed to a growth in entrepreneurship (Hombert et al. 2020). The Alaska Permanent Fund Dividend program, established in 1982, also suggested a positive effect on entrepreneurship although the results appeared to dissipate over time (Feinberg and Kuehn 2020). The question arises, did the PUA and PEUC money—as a guaranteed, weekly income for unemployed workers—function like a universal basic income and provide the same benefits and opportunities? Or was it too short-lived, or delayed, to have the same possibilities?
METHODOLOGY
This article is based on the first two waves of data collected in a mixed methods panel study. The first phase was from April through August 2020, during the initial outbreak of the virus in New York City, when New York was regularly described as an “epicenter” of the outbreak in the United States (McKinley 2020; Thompson et al. 2020).2 A second wave was undertaken from November 2020 through June 2021, to coincide with the second surge of the virus and early stages of the coronavirus vaccine roll-out. A third wave of data collection, ongoing from November 2021 through 2022, focuses on the impact of the end of these unemployment supports.
Gig-based and precarious workers were recruited via Facebook groups for gig workers and unemployed workers; advertisements on Craigslist, OffStageJobs.com, and Dance/NYC; posts on New York City–focused Reddit, and snowball sampling. One worker was recruited when he was found busking in a local park. Workers in the New York metropolitan area (broadly defined as Westchester to Newark, New Jersey) who used gig platforms for work, or were in precarious jobs such as retail, restaurants and bars, or freelance work were eligible to participate. For the second wave, only respondents who had participated in the first wave were eligible to participate.
Similar to Pugh’s (2015) work on work insecurity, our interview sample deliberately draws on three groups: creative freelancers, who are generally well educated but still often paid as 1099 workers; low-wage W-2 employees in both essential (grocery store and drugstore) and nonessential (personal care and restaurant workers) fields; and gig economy workers. The goal of this three-group sample is to tease out whether the challenges experienced by precarious workers during this unprecedented time are due to their status as 1099 workers or their status as low-wage workers.
In the first round of data collection, 199 precarious workers participated: including sixty creative freelancers, thirty-three restaurant workers, and thirty-one nonmedical low-wage workers. Sixty gig-based workers were included from platforms such as TaskRabbit, Doordash, Instacart, and Uber, in addition to fifteen nonplatform-based gig workers who secured work via websites such as Craigslist. All workers had worked face-to-face before the pandemic; and more than three-quarters (76.5 percent) reported job or income losses due to the pandemic. In the second phase, respondents included fifty-five creative freelancers, twenty-six restaurant workers, twenty-eight nonmedical low-wage workers, forty-seven gig workers, and twelve non-platform-based gig workers. The attrition rate between the first and second phase was 15.6 percent.
In both waves of data collection, participants completed a short online survey and participated in a respondent-directed telephone interview (Weiss 1994). The first round of interviews sought to identify the immediate impact of the pandemic on workers: what they were doing before the coronavirus and their income source or sources before and during the pandemic; their experiences (if any) in applying to unemployment; how their daily routine had been affected; and their perceptions of how platforms, clients, or employers were handling the situation. The second round focused on changes or continuances in sources of income or jobs, how respondents thought political changes may have affected the pandemic, and their views on the COVID-19 vaccine.
All interviews were recorded before being transcribed and analyzed. Given the quantity of data, transcripts were analyzed using flexible coding (Deterding and Waters 2018), an iterative coding method well suited for collaborative analysis of in-depth interviews. In the first round of coding for both phases of data collection, an undergraduate research assistant “indexed” the interviews at a broad level by coding by interview question (Deterding and Waters 2018, 15). A second stage involved the development and application of conceptual codes. Respondents were assigned pseudonyms based on popular baby names for their generation, and given a $25 gift card incentive for phase 1 participation, and a $50 gift card after phase 2 participation.
In phase 1, 103 of the respondents (51.8 percent) were women, 46.2 percent were men, and 2 percent identified as gender-nonconforming, transgender, or nonbinary. Fewer than half (41.2 percent) identified as white, 15.6 percent as Black, 14.1 percent as Hispanic, 14.6 percent as Asian, and 14.5 percent as multiple races. The interviewees ranged in age from nineteen to sixty-four years old, averaging thirty-three. Equal numbers had an associate’s degree (9 percent) or a high school diploma (9.0 percent). Just over 25 percent had some college experience, 36.2 percent held a bachelor’s degree, and 19.6 percent either held a graduate degree or had some graduate school experience. Nearly two-thirds (60.7 percent) made less than $40,000 per year; 28.5 percent earned less than $20,000. Slightly less than a quarter (23.5 percent) made between $40,000 and $70,000 per year before the pandemic.3
Some differences across these three groups were evident: the low-wage workers were the most likely to report having some college or below (53 percent) relative to gig workers (34.7 percent) and creative freelancers (16.7 percent). The creative freelancers in the sample reported higher levels of education, 45 percent reporting a bachelor’s degree relative to gig workers (30.7 percent) and low-wage workers (34.4 percent). Although gig workers (18.6 percent) and creative freelancers (20 percent) had nearly equal percentages of graduate degrees, only gig-based respondents reported having PhDs or medical degrees (5.3 percent). However, contrary to the assumption that their higher levels of cultural capital increased their incomes, creative freelancer respondents generally reported slightly lower incomes than gig workers; 73.3 percent reported incomes under $50,000, versus 70.8 percent of gig workers and 76.6 percent of low-wage workers.
FINDINGS
Many workers expressed a sense of gratitude for the CARES Act, PUA, PEUC, and assorted stimulus payments. “Thank you, thank you, thank you, unemployment and food stamps. The thing is, I hope they can continue to the first of the year, because I’m making more money in unemployment,” said Bernice, age sixty-four, an unemployed creative director. “When unemployment gives you more money—and I’m a Buddhist, I chant that they keep it up for everybody—that plus the stimulus for everybody. That would be fabulous.” Most respondents were less effusive, but still expressed appreciation, such as Aretha, age fifty-two, an unemployed TaskRabbit worker, who explained, “I did get a stimulus check. I did. That didn’t last very long, but I’m thankful for that.”
Workers also compared what they were making through the CARES Act with what they would have made previously on unemployment, a comparison that contributed to a sense of thankfulness. As Kelsey, a twenty-eight-year-old unemployed dancer, choreographer, and production manager, explained, “The CARES Act, that is immensely helpful. Unemployment in general is not very much money at all. I would have lost my apartment if I was only on unemployment. With the CARES Act, I’m able to at least hold onto my apartment for a little while longer, but that is ending at the end of July,” she said. “And so that’s why I’m already preparing myself to lose my apartment and to leave New York because I won’t be able to keep it without that money. Unemployment alone is so little.”
Workers also expressed a sense of discomfort with making more money on unemployment than they were making working multiple jobs, as was true for Jennifer, twenty-eight, a furloughed events center worker and an unemployed grocery store worker. As she explained, “It’s infuriating. But I’ll take it because I’m making way more by being unemployed than I was working three jobs.” For workers who were in the arts or in creative fields that they often described as a passion (DePalma 2020; Ocejo 2017; Leidner 2016), being paid more to not perform or produce further highlighted their low incomes. “Obviously I didn’t go into the arts to necessarily make a bunch of money. . . . And it didn’t necessarily feel great to get money without working,” said Gabriel, a twenty-two-year-old unemployed dancer. “If I could make that much money while working, I would feel better, but I also just took it as sort of a weird ten weeks where I was able to save all this extra money, but not sort of think about it too much in the grand scheme of things, because that was just going to make me feel very sad.”
This sense of gratitude, which was especially prevalent among the creative freelancers who would not normally be eligible for any unemployment assistance, is especially striking relative to research on other unemployed workers. For instance, Damaske (2021) finds that even though one white male respondent referred to a “nice unemployment” benefit from the state and was also able to rely on severance and a spouse for support, most workers find unemployment benefits low, in some cases putting them below the poverty line.
However, as Kelsey, the unemployed dancer and production manager, noted, the CARES Act funds were short-term solutions. Three common issues often accompanied the sense of gratitude for the unemployment funds: first, workers faced a strong sense of uncertainty about how their future income would compare with their future needs; second, the funds were temporary; and, third, the money came after extensive delays and difficulties.
The Specter of the Unknown: “I Could Be Unemployed the Rest of the Year”
Although the CARES Act enhanced unemployment benefits of $600 a week was intended to end at the end of July, many experts believed that it would be extended (Bauer, Edleburg, and Parsons 2020; Stewart 2020), especially as other regions of the country faced their own summertime surges. New York provided several weeks of a $300 supplement during August from Federal Emergency Management Agency (FEMA) funds, but an additional $300 a week Federal Pandemic Unemployment Compensation (FPUC) benefit, lasting through August 2021, wasn’t authorized until December 2020 (FEMA 2021; Iacurci 2020).
As a result, the specter of the unknown regarding the potential length of the pandemic—and the short-lived nature of the supplement and potentially of unemployment assistance for workers classified as independent contractors—affected how unemployed workers viewed their benefits. Unemployment assistance is usually viewed as high-velocity money, which quickly transfers from buyers to sellers, but respondents reported spending little of it. Part of the issue was that with much of New York under so-called lockdown, opportunities to spend on shopping binges or travel were limited. As Theresa, age twenty-seven, an unemployed early childhood education teacher explained, “I’ve been able to save a lot of money, actually.”
But for many respondents, saving money was not simply a response to few spending opportunities, but part of a concerted effort to reserve funds for the future. Contrary to the expectation that unemployment assistance is quickly spent, or research that has found income volatility to be associated with a preference for more immediate spending (Peetz, Robson, and Xuereb 2021; West and DeVoe 2020), respondents often reported a concerted effort to save their unemployment benefits. For instance, Olivia, thirty-two, an unemployed voice actor, was excited when she began receiving her unemployment assistance, but soon realized the influx of cash was far from a carte blanche.
So I was like, “Oh, all this is extra money I’m going to have, I’m going to do fun things with it.” And then I think, what did I do? I had to get something from Sephora for like a face wash, I don’t know. Something that I needed to get anyway. And so I think that’s the only thing I bought because I realized, “Wait, what if this doesn’t end as soon as everyone thinks it does, and I have to start saving for the time when I might not have money or whatever?” So it felt really good for a couple of days. And then I was, “Oh no, now I have to be responsible.”
Similarly, Claudia, thirty-three, an unemployed actor, filmmaker, and telemarketer, received a small grant of $100 from an arts organization and some money from her parents while waiting for unemployment assistance. When she finally received her unemployment assistance, she felt “okay,” but it didn’t assuage her concerns about the future.
I’m making a little bit more money each week and I’m putting that money mostly in savings because I’m not sure. I could be unemployed the rest of the year, so I’m paying off a little bit of debt, a few credit cards that . . . And then just putting that money away because the potential to not go back to work for many more months could be there. So I’m trying to be smart about that extra money that I’m getting because I think the reason they’re giving me that extra money is because there’s a potential that businesses might close. If my telemarketing job—if they close down—I don’t have anything to go back to.
For Claudia, the unemployment assistance was seen not just as a way for her to stay afloat during the early months of the pandemic, but also as portending future financial calamity if her main source of income were to close permanently.
This desire to save the funds, not often found with unemployment benefits, is common with both EIC and UBI. Sondra Beverly, Daniel Schneider, and Peter Tufano (2006) reveal that EIC recipients often planned to save a portion of their refunds although they faced challenges with the lack of options for splitting refunds into both savings and checking accounts. Ruby Mendenhall and her colleagues (2012) also find that the majority of households receiving EIC planned to allocate a significant portion to their savings. Similarly, individuals receiving UBI payments, such as through the Magnolia Mothers Trust, also reported saving a portion of their benefits for future food and household expenses (Kaufman 2018).
Claudia was not alone in perceiving the additional unemployment funds not as an incentive to stay home, but as a hedge against the future. Katelynn, a twenty-four-year-old unemployed bartender who turned to essential work while awaiting unemployment funds, also spoke about saving her stimulus funds and, later, her unemployment assistance. As a bartender, Katelynn had previously stored her savings in an assortment of empty Jameson bottles hidden in her home. “Yeah, I mean now I’m definitely working on my savings account just because whatever I get now, I try to squirrel it away.” For Katelynn, who was used to stockpiling funds in whiskey bottles, squirreling away excess funds was not a new phenomenon, but the unknown future of the pandemic gave it new meaning.
The idea of saving the funds for an unknown future was a common theme among unemployed precarious workers. “It was nice, but it was also something that you realize that you have to save because there’s no idea of when it’s going to end. So it was nice,” said Rueben, twenty-seven, a college student and unemployed restaurant worker. “But it’s also like, ‘Hey, you’ve just got to squirrel it away because you have no idea when things are going to quote unquote return to business as usual.’ Although there’s no business as usual in this thing, when we’re all adapting, but you know what I mean.” Even in the follow-up interviews conducted during the second surge of the pandemic, and after the $600 a week supplement had long since ended, workers spoke about their efforts to conserve the funds. As Angelica, thirty-four, an unemployed actress and part-time nanny, explained in her phase 2 follow-up interview, “I mean I live a pretty simple life, but it’s still living in a city that’s quite expensive. And so I knew that [the supplement] was ending, and so I did a really good job of squirreling as much money as I could.”
The phrase squirreling away is particularly instructive. Much as a squirrel hides away reserves of acorns in preparation for a long winter, our unemployed precarious respondents also saw themselves as preparing for a long road ahead. Indeed, Ruben, the unemployed restaurant worker, took the squirrel metaphor to a new level. His first stimulus check—$1,200—was received during the gap between his last paycheck and before he began receiving unemployment assistance, as his money was running out. He explained, “The first thing I did when I got that was I just bought a bunch of groceries and I filled the freezer and everything. So that’s the first thing I did. And then the rest I just squirreled away.” For Ruben, his preparations for a long pandemic included not just reserves of funds, but also reserves of food that he stored in his freezer, itself a site for long-term storage. This squirreling away of funds and foods is similar to findings by Laura Tach and her colleagues (2019) that respondents used their income tax refunds to stockpile household staples and pay off debts, essentially leveraging their tax refunds in “multiple forms of self-insurance.”
The sense that the money needed to be saved and not spent is especially emphasized when workers discussed the purchases that they later viewed as frivolous and regretful. Ruben noted that he “probably spent $100 on random stupid things but other than that, I saved it.” Lauren, twenty-five, an unemployed waitress, also described her early unemployment expenditures in a negative light. “Yeah, so I got unemployment, which was really nice but I was kind of an idiot with it . . . especially in New York when you get that extra $600,” she said. “Most of us are like, ‘This is more money than we normally make when we’re actually working!’ . . . I’ve saved a good amount, but I’ve also spent way more than I ever should have. . . . I do have a fair amount saved, but it’s not what I would have liked.” For Lauren, being “kind of an idiot” with the funds meant that she did not have as much saved as she would have preferred.
Disparaging one’s purchases as “random stupid things” or describing one’s self as being “kind of an idiot” with the money suggests that respondents did not just support the saving of the funds, but also believed that any spending needed to be careful and only on items of importance. Research suggests that increased uncertainty about future income decreases consumption and increases savings (Fisher 2010; Sandmo 1970). The challenges recipients experienced in managing the funds are further supported by Jonathan Murdoch and Rachel Schneider’s (2017) work on income volatility, which finds that influxes of cash—although better than a sudden effluence—are also disruptive. A lack of funds can cause problems paying bills, but an influx can also cause budgeting issues as recipients attempt to determine how likely the funds are expected to last and whether an influx today might mean less in the future. Alessandra Guariglia (2001) finds that households save more if they expect their future financial situation to deteriorate. Meanwhile, the theory of precautionary saving posits that saving is seen as offering protection from income shocks, such as the loss of unemployment assistance, and households facing higher income risks are more likely to save (Lusardi 1998).
Although the temporary nature of the funds contributed to respondents’ perceived need to save the money, early delays in receiving the funds also contributed to a sense of unknown and challenges in how the assistance was perceived and used, as the next section demonstrates.
Delays: “It Was a Horrible, Horrible Time”
Bell and his colleagues (2023, this issue) highlight the geographic disparities that reduced access to unemployment insurance during the pandemic, noting that areas with more disadvantaged residents had less access to benefits. Alexandrea Ravenelle, Ken Kowalski, and Erica Janko (2021) find that workers experienced three main hurdles in regard to accessing unemployment assistance: knowledge, sociological, and temporal-financial barriers—or didn’t know, didn’t want, and can’t wait. Workers often did not know that they qualified for unemployment benefits, or did not understand how the benefit worked. As Adira, twenty-three, a food delivery worker on the app Caviar explained, “I don’t get it. You just sign up, say ‘I don’t have a job’ and the government gives you money? What is that about? If it was that easy, wouldn’t everybody do it? I don’t get it” (Ravenelle, Kowalski, and Janko 2021). For immigrant workers, even those with work visas or green cards, receiving unemployment assistance was seen as potentially risking a claim of public charge that could affect their ability to remain in the country or future citizenship applications (Selyukh 2020; Smith 2020).
Additionally, workers face a sociological barrier to receiving the benefits, professing a lack of desire to receive unemployment funds. One respondent, a private investigator turned delivery worker, described unemployment assistance as a benefit that “doesn’t really benefit people like me or people who work” (Ravenelle, Kowalski, and Janko 2021). Finally, much as Marci Ybarra and Frania Mendoza Lua (2023, this issue) found, disenfranchisement from multiple sources of government, including the safety net, was common. Some workers, experiencing weeks or months of delays, simply felt that they could not wait any longer for unemployment assistance and opted for the “side hustle safety net” of gig economy work on food delivery and grocery shopping apps.
This temporal-financial barrier of extensive delays also affected how respondents viewed their unemployment assistance when they did finally receive it. Many reported extensive delays in completing their application for unemployment assistance. The initial application could be completed online, but respondents often reported that the website regularly crashed or froze during periods of high demand, requiring them to enter their information repeatedly. In many cases, claims information had to be confirmed over the phone with a representative from the Department of Labor, leading some respondents to spend days continually calling the office. As William, forty-two, an unemployed TaskRabbit, explained in his April 2020 interview, “I’ve basically been calling them almost every day, multiple times, hundreds of times even, and can never get through. I’m still waiting to hear back from them actually.”
As a result of these delays, when the funds finally arrived, workers found themselves playing catch-up with their bills. “I struggled with the website for the first three weeks being unemployed. So that is when I had to just use the money that I had to pay basic bills and had to come to my grandparents’ house in Rhode Island to cut back on expenses,” said Kelsey, the unemployed dancer and production manager. Once she received the unemployment assistance, she was able to pay off debt accrued during the early days of the pandemic, but was not able to get ahead of the expenses and replenish her savings. She explained:
So in the end it’s not that I’m making more money. I would get to that point if the CARES Act continued for a little bit longer. And then I wouldn’t feel like this whole situation was such a financial blow. I could feel like I could at least hang on to my apartment and earn back the savings that I’ve had to dip into. And then I would feel better supported and better equipped to step back out into the world, into my career and not have to feel I am now living back with my grandparents because I don’t have the means to move forward.
Similarly, Amir, twenty-six, the unemployed owner of entertainment-focused small business turned food delivery worker, experienced extensive delays in receiving financial assistance during the pandemic.
Thankfully my wife and I, it was delayed for quite a bit, but we did get our [stimulus] share, which is $2,400. But at this point that $2,400 because it’s been so delayed, there have been so many bills that have piled up. There’s so many. . . . It was money that just disappeared. It wasn’t even . . . We couldn’t use it towards food, we couldn’t use it toward our general livelihood, because there’s so many credit cards that are being just maxed out, because people have to live, people have to eat, people have to get food.
At one point, Amir tried to obtain free food from local schools, part of the Community Meals program, but was discouraged by the number of people also seeking meals. Eventually he turned to the side hustle safety net of food delivery apps. “We’re like, ‘I might as well just do deliveries and make an extra couple of dollars, if I have to make . . . If I can just . . . Even if I make a three dollar order in one hour . . . with three dollars I can buy myself a White Castle meal or something.’”
The experiences of Kelsey and Amir, while poignant, are not unusual. Vance Larsen and his colleagues (2021, 2) find that income shocks (such as unemployment), rather than expense shocks (a sudden car repair) are harder to cope with, have more impact on daily life, and are “perceived as more of a loss than expense shocks of the same amount.”
In some cases, the delays that workers experienced in receiving assistance were so considerable that they reported long-term implications in terms of their sense of security. For instance, Marissa, twenty-five, an unemployed server and performer, applied for unemployment assistance soon after New York’s shut-down began.
The problem I had though was I didn’t receive my stimulus check or a single unemployment payment until I believe the second week of May. So I went from March 15th to about May. . . . So it was about a full two months without. . . . My stimulus was behind. And my whole family, who was still working, received their [stimulus] before me, who was unemployed. That was hurtful. No one’s fault, of course. But I’m like looking around, seeing all these people talk about their [stimulus] money, “I’m going to buy this and go on vacation.” I’m like, “I haven’t received a dime from the government in two months. I am struggling here.”
A writer friend, living out of state, sent her money every few weeks so she could purchase groceries.
It was a horrible, horrible time. I just remember every day thinking, “I don’t know how I’m going to make it another week. How am I going to get my cat’s food? How am I going to get my own food? How am I going to pay rent next month? Am I going to get evicted?” It was the most uncertain time I’ve experienced probably ever . . . And it was for about two straight months. I lost 20 pounds right away. Everyone kept commenting: “Oh, you look so good. You look so good.” I wasn’t eating. I wasn’t sleeping. I was afraid to eat cause I didn’t know if I could afford to eat. I was afraid to pay my rent because if I paid my rent, I wouldn’t afford to eat. You know, it was horrible, horrible, horrible.
When she finally received her unemployment assistance, it included several weeks of payments at once. “And then all of a sudden, I was kind of rich,” she said. She paid some medical bills, her taxes, and several months of back rent. This need to use a lump sum of unemployment assistance to pay bills that had piled up previously is similar to how EIC funds are often used. Shawn Cole, John Thompson, and Peter Tufano (2008) find that more credit-constrained individuals spend their funds more quickly and that a greater fraction of their refund is used to pay for necessities such as food and transportation expenses. Research examining a two-week delay in 2017 of more than $40 billion in tax refunds found that the delay led to a notable change in the timing of spending. Aditya Aladangady and colleagues (2018) find that the delay also affected spending on nondurable necessities, such as groceries.
For Marissa, the months of being destitute have taken their toll. She explained:
Ironically, now I have more money in my savings account than I’ve ever had in my whole life. And I still don’t feel secure. I feel every week, even with that unemployment coming in, it doesn’t ever feel like it would be enough. ‘Cause I think I’m a little bit traumatized from having the ground ripped out from under me like that. You could hand me a million dollars a day and I’d be like, “Oh my God, put it in savings. Don’t touch it.” Like it’s never going to be enough. I feel that trauma now of like, I could never have enough money. I’ll never feel secure because I still don’t even know if I’ll have a job in a few months. Like yeah, my savings account looks pretty right now, but I’m going to be living off of that probably for the rest of 2020. So, that’s scary.
Although Marissa’s delays in receiving unemployment were extensive, her concerns about the future were not unique. Research suggests that uncertainty regarding income, employment, or health affect savings behavior (Fisher 2010; Lusardi 1998; Carroll 1994; Deaton 1991; Sandmo 1970). As the next section demonstrates, for some workers, concerns about the future—partnered with the short-term nature of the CARES Act, were a deterrent to receiving unemployment assistance, or increased the desirability of employment, even if it paid less.
It’s Only Short Term: “It’s More About Longevity.”
A good deal of media attention and conservative handwringing was paid to the weekly $600 week unemployment enhancement, which enabled 76 percent of workers eligible for benefits to make more on unemployment than they had previously been making (Ganong, Noel, and Vavra 2020). Even among the unemployed workers interviewed, the concern was that their cash assistance was disproportionately high and would lead to mass layoffs or resignations as other workers attempted to secure the same benefits.
“I am making more than the grocery store workers with my unemployment. . . . I’ve talked to a lot of people that make minimum wage, or don’t make that much money. And they would make a lot more on unemployment,” said Cody, age thirty, an unemployed dog walker. “Has the government provided a lot of subsidies to businesses so they don’t furlough people? . . . They could easily furlough these people . . . and they would make more money on unemployment.”
Cody’s concerns—that receiving enhanced unemployment was more desirable than working, especially when making minimum wage—was also echoed by state governors in various states that voted to end their participation in the CARES Act early in an effort to reduce local labor shortages (Goger 2021; Mills 2021; Romm and Rosenberg 2021). Yet, among the respondents we interviewed, although many were aware that they would make more on unemployment, the time-limited nature of the assistance was a deterrent. For instance, Jessie, thirty-one, was working full time as a drug store bike courier during the early days of the pandemic. “So yeah, I’d be making more on unemployment,” he said, noting that he had chosen to continue working instead. “Which would be nice, but then I tell myself, it’s like, ‘oh, it’s only until July that that $600 a week is coming in, right?’”
Likewise, Vilma, a laid-off hotel cleaner, was receiving unemployment assistance, but for some reason not the additional $600 a week when she was interviewed in the spring of 2020. She noted that the additional funds would be helpful, but the short-term nature of the assistance affected her views on whether she should seek the help. “Well, it would be comfortable at the beginning but I would still be looking for something else to do to work. Because I can’t depend on that, that might be for a little bit and then all of a sudden, guess what? They’re not sending it anymore,” she said. “I’m still going to look for a job.”
Respondents saw work, even precarious work, as more stable and secure than the temporary uptick in income that the enhanced unemployment assistance provided. “For me, it’s more about longevity. I also have a child, so those things are important to me,” said Alonzo, thirty-seven, a TaskRabbit-turned-janitor. Alonzo also noted that he would be making almost $100 more weekly on unemployment, but returning to work was preferable because it offered him the sense of “still being able to have some security that I’ll have a job within the next two or three months.” Although low-wage work is generally seen as unstable, undependable, and offering few opportunities for advancement or benefits (Kalleberg 2011), even low-wage jobs provide a sense of daily order and stability, and a sense of dignity that is not often associated with governmental assistance, even temporary unemployment benefits (Newman 1999).
Interestingly, much as Newman (1999) finds in her research on low-income urban workers, the respondents who were currently or previously paid as W-2 workers were especially reluctant to be on unemployment and espoused a strong work ethic. Jennifer, twenty-eight, an unemployed events worker and former grocery store employee, is perhaps the most notable example of a worker reluctant to go on unemployment benefits instead of working. Early in the pandemic, Jennifer believed that she might have COVID-19, and then began experiencing panic attacks from the stress of public-facing essential work. “I feel this obligation to work. It’s an essential service and I felt pride in that,” she said. “The fact that I would be making twice as much and also not be putting myself in a situation that would either get me exposed or make me have panic attacks. I just figured it was the right thing to do, but I still feel really guilty and weird about it.”
Uncertainty Now, Entrepreneurship Later: “Maybe There’s Some Opportunity Here”
Although many workers remained focused on the temporary nature of the assistance and how quickly they could return to work, some also took advantage of the increase in their income and lack of work obligations to focus on artistic pursuits or pursue entrepreneurial ventures. Much as the EIC is seen as “springboard for upward mobility” (Sykes et al. 2015) or how NITs and UBIs lead to significant impacts on education outcomes (Hoynes and Rothstein 2019; Marinescu 2018), the PEUC was seen as providing recipients with an opportunity to pursue additional opportunities. For these workers, the enhanced unemployment assistance functioned as a guaranteed income that allowed them to begin creating a small business, or further invest in themselves.
For instance, thirty-six-year-old Josh, an unemployed artist who had spent the last few years working in a self-described “dead-end job” in food service, used his time on unemployment to take free classes in computer coding and website development. “My time is better spent kind of trying to work on bigger professional goals, so it feels like a big opportunity for me,” he said. “I feel like this is giving me an opportunity to build my skills back up without any capital output.”
After years of struggling on minimum wage or close to it, the influx of funds and his growing savings account also allowed Josh to begin envisioning new possibilities. “I’m like, ‘Oh, maybe there’s some opportunity here,’” he said. “People are going to be selling stuff to try to raise money, maybe this is an opportunity for me to kind of get some of the things that I’ve wanted for a long time and haven’t been able to.”
Similarly, Crystal, twenty-seven, a married, unemployed TaskRabbit cleaner with toddler twins, used her PUA and PEUC funds to invest in a microgreens business that she planned to run from her home, offering her a chance to save on childcare expenses while also earning money. “It’s pretty exciting. Like I said, if it wasn’t for the COVID, I probably wouldn’t have been able to do this,” she said. “We just got our LLC for it. We have to finish our good standing paperwork and stuff like that, but we’re pretty much legally ready to go.”
The one other respondent who started a business during the pandemic did not receive unemployment, but was in a similar situation in regard to income security when his boss refused to lay him off despite a lack of business. Colton, twenty-one, a spa manager, took advantage of his secure employment to purchase a programmable sewing machine and to start an Etsy business selling embroidered sweatshirts, making a few extra thousand dollars per month.4 “This is also a job, but I actually enjoy doing it. It’s cathartic and peaceful, it’s even satisfying to hear the needle going up and down really fast,” he said. “ I’m trying to find a place where I can have space to potentially run my own business out of and make that my full-time thing if it could work. So I’m trying to work that into my lifestyle.”
These business starts may not be unusual. Even though “the CARES Act did not directly support new business formation,” the distribution of relief “was followed by a relative increase in start-up formation rates, particularly in neighborhoods with higher median incomes and a higher proportion of Black residents” (Fazio et al. 2021, abstract). For all of the upheaval, nearly a million more new business applications were filed in 2020 than in 2019 (Anagnos 2021; Fox 2021). One possibility is that, much like in France where workers could continue to receive unemployment while starting a business, workers felt financially secure enough to take a risk on new business development (Hombert et al. 2020).
Although most of the respondents did not start businesses with their unemployment assistance, their focus on laying the framework for a career change or more successful entrepreneurial venture was not unique. Creative freelancers in particular often focused on taking their time on unemployment to “make the future” that workers wanted to have after the pandemic (Ravenelle and Kowalski 2022). For many workers this meant getting back in shape, or taking classes in preparation for future auditions or roles. In the most extreme cases, future planning involved reaching out to former contacts with letters of introduction, to ensure that they wouldn’t “forget” them, and pitching future travel stories along with definitive plans for 2021 (Ravenelle and Kowalski 2022).
CONCLUSION
To describe the COVID-19 pandemic as unprecedented has become a truism. Yet never before in American society have the majority of unemployed workers found themselves making more on unemployment than they had been making while working (Ganong, Noel, and Vavra 2020). Additionally, the pandemic was the first time when gig workers and the self-employed could receive unemployment assistance, albeit at half the level of their unemployed employee peers. At the same time, social distancing efforts and local lockdown orders also resulted in widespread business closures and a sense of anxiety about the future. As a result of these factors, we ask, how do precarious workers, many of whom were hustling for money or engaged in creative fields, feel about making more on unemployment? How are they using the funds?
We find that workers, though they expressed a sense of gratitude for the funds, explained that it “didn’t feel great” to be paid more for not working, especially when their previous work involved juggling multiple jobs or pursuing artistic passions. Additionally, the specter of the unknown affected workers’ use of their benefits, and generated concerns that a financial influx today was simply preceding a shortage tomorrow. As a result, workers emphasized the importance of squirreling away their unemployment funds—upending the expectation that unemployment assistance is high-velocity money quickly spent on necessities. Finally, delays in securing unemployment assistance—and the penury that workers sometimes endured first—also contributed to the sense that the financial scars from their experiences would linger.
In some cases, the additional funds were used to start businesses, lay the framework for a career change, or contribute to efforts to jumpstart the workers’ careers whenever the pandemic ended. In this way, and for some workers, the enhanced unemployment assistance served as a guaranteed income that allowed them to begin investing in a small business venture, or further invest in themselves, and may have allowed for a major career reset. Future follow-up interviews, funded by a Russell Sage Foundation Future of Work grant, will allow us to see the long-term implications of this assistance—and its end in September 2021—on worker careers.
Within a sample that comprised three distinct groups of workers—creative freelancers, low-wage W-2 employees, and gig economy workers—two differences were notable: creative freelancers were more likely to express a sense of gratitude for the unemployment benefits that they would not typically be eligible for, and low-wage W-2 workers were more likely to express a reluctance to be on unemployment and a desire to be working. We theorize that part of this issue is that low-wage workers were cognizant of the challenges of job searching (Newman 2008) and that the typically higher wage creative freelancers often used unemployment funds to enable an investment in personal growth (Ravenelle and Kowalski 2022).
We caution here that even with these challenges, we do not suggest—nor does the research support—that the PUA, PEUC, or FPUC were problematic policy decisions in terms of the experience of precarious workers. Our longitudinal interviews with respondents highlight the challenges that workers experienced early in the pandemic, including sudden job loss, a sense of anxiety and uncertainty regarding the coronavirus, and challenges in accessing unemployment assistance (Ravenelle, Kowalski, and Janko 2021). Indeed, the experience of Marissa, for whom delays in receiving unemployment funds and anxiety about feeding herself had resulted in a sense that she’ll “never feel secure,” suggests that, if anything, the problem with the PUA was that it simply wasn’t distributed fast enough, or long enough.
However, these findings suggest that although enhanced and expanded unemployment assistance is crucial in a sudden and significant economic downturn, it is not enough to simply give workers more unemployment funds and assume that it will immediately be spent “stimulating” the economy. Instead, we find that worker uncertainty about the future of the pandemic supports, partnered with delays and administrative burdens, led to a savings response. The implications for public policy are clear: easy-to-understand guidance—such as publicizing when, and how, unemployment assistance limits are extended, and the likelihood of continuing PUA funds—would strengthen the economic security of workers and their ability to move forward in novel times. Otherwise, additional funds without transparency could simply make workers ever more certain of their insecurity.
FOOTNOTES
↵1. Although unemployment assistance is generally limited to those without work, the CARES Act allowed anyone who could prove that their work was affected by the pandemic to receive benefits, even if they continued working or earning money up to a state-set limit.
↵2. More than 95 percent of the interviews and surveys were conducted between April and June 2020.
↵3. Although $60,000 annually might sound like a considerable income, the 2015–2019 median individual income in New York City was $63,998 (U.S. Census Bureau 2021). Additionally, illustrative of the high cost of living in New York, single residents are considered eligible for subsidized or public housing if they make $66,850 or less annually (NYCHA n.d.).
↵4. The title of spa manager suggests a relatively secure middle-management job but is a bit of a misnomer. A recent college graduate, Colton’s pay was never particularly high, but during the pandemic it was reduced to New York’s minimum wage of $15 an hour, garnering him an annual salary of $31,200. Additionally, during the pandemic, any cash bonuses he had previously received were discontinued.
- © 2023 Russell Sage Foundation. Ravenelle, Alexandrea J., and Savannah Knoble 2023. “‘I Could Be Unemployed the Rest of the Year’: Unprecedented Times and the Challenges of ‘Making More.’” RSF: The Russell Sage Foundation Journal of the Social Sciences 9(3): 110–31. DOI: 10.7758/RSF.2023.9.3.05. We thank Dawn Culton for her work in index coding this data, and Erica Janko and Ken Cai Kowalski for their work in respondent recruitment and interviewing. This material is based on work supported by the National Science Foundation under Grant no. 2029924. Direct correspondence to: Alexandrea Ravenelle, at aravenelle{at}unc.edu, University of North Carolina at Chapel Hill, Sociology Department, Hamilton Hall, Campus Box 3210, Chapel Hill, NC 27599, 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.