Introduction

Around 2:00 PM, a black middle-aged customer in her forties abruptly cut through my register line, leaning in close as I mechanically scanned barcodes across the red laser. Her eyes widened with concern as she explained, “A bottle of laundry detergent spilled all over the aisle. Someone’s gonna slip and fall on it.” She repeated the last sentence a few times as I continued ringing up merchandise for a long line of customers. I nodded my head slowly, shifting my eyes forward to the register display. “Okay. Okay,” I repeated to her until she gave up and disappeared. Twenty minutes later, I finally handed the last customer in line his receipt. Bolting to the back of the store, I shoved the stockroom doors open and dodged around carts stacked high with cardboard boxes. In the very back corner was the slop sink. Restricted within the tiny square of space, I grabbed hold of the yellow plastic mop bucket and glanced inside. It was filthy, coated in a film of gray grit with thick sediment at the bottom. The mop head was also caked with dirt, crunching between my fingers as I broke up the fiber strands with my bare hand before running it under the tap. As I vigorously squeezed the coarse yarn, dark water continued to flow out. After a few moments of rinsing I decided it would just have to do. Filling up the trough halfway, I steered it quickly towards the detergent aisle. An empty Tide bottle with a cracked plastic cap lay on its side in the middle of the floor. Seventy-five fluid ounces of the slick blue liquid had oozed its way down the aisle and crept underneath the shelving. I slopped it up the best I could using swift broad strokes, smearing suds across the dingy linoleum while thinking, “This must be the cleanest spot in the store now.”

“We are awakening to a dollar-store economy,” the New York Times declared in 2011, a culture of fear-induced bargain hunting spurred by the 2008 market meltdown and jobless recovery (Hitt 2011). More than ever before, author Jack Hitt states, anxious consumers are looking to stretch a buck due to the shrinking middle class and a widening gap of economic inequality. Yet dollar stores thrive in climates of economic uncertainty. In fact, “their success is built on the death of the American middle class” (Taylor 2017, 1). Fortune 200 corporation and industry leader Dollar General (DG) has attributed its financial success to the expanse of their target demographic—households earning $40,000 or less a year. DG CEO, Todd Vasos, has stated, “The economy is continuing to create more of our core customer…We are putting stores today [in areas] that perhaps five years ago were just on the cusp of probably not being our demographic, and it has now turned to being our demographic” (Taylor 2017, 2). VSN Strategies, a technology firm DG has relied on to help track, analyze and understand ongoing trends in shopper behavior, told Forbes Magazine that while DG’s profit strategy “may be a tad cold-hearted… there’s no denying that they’ve been responsive to economic reality… In mass retail, it’s smart to target the demographic bulge. If that happens to consist of folks who can just afford the four-pack of toilet paper before payday, then don’t fault the retailer” (McClain 2017, 2).

The data I present in this paper was collected during six months of fieldwork working as a low-wage sales associate at Downtown DG, as I refer to my primary fieldsite. I also conducted participant observation at the eighteen other DG stores in my fieldsite’s district and completed fifty in-depth interviews with DG coworkers and employees. My findings revealed significant disparities in store quality and customer service between the DG locations I examined. Of the nineteen stores in the district, three standout stores emerged as the very worst. Conditions there were dirtier and more hazardous than the rest, with under-stocked shelves, slow customer service, and a contentious atmosphere for those who worked and shopped amidst the squalor. According to DG’s Customer Satisfaction Survey statistics posted in the back of my fieldsite’s stockroom, these three stores were rated the lowest in the district. Census tract data confirmed that these DG locations also had the highest concentration of black residents in the district. Therefore, I identified these three stores as “consumer redlined,” a concept introduced by Columbia sociologist Adam Reich in his 2016 analysis of online Wal-Mart Yelp reviews. Reich found that Wal-Mart stores located in lower-income communities of color consistently received lower ratings than those in wealthier, whiter locations. Poor black neighborhoods were host to the lowest-scoring Wal-Mart centers, places that were frequently described as the “worst,” “nasty,” “unorganized,” “ghetto,” and having slow customer service (Reich and Bearman 2020). Reich was the first, and until now, the only sociologist to explicitly identify and name consumer redlining as a phenomenon. Yet, the extent to which it may plague other major retailers is unclear, as are the mechanisms that give rise to it.

My ethnography captured the egregious conditions and perpetual frustrations that workers and customers navigated daily at a DG consumer-redlined store. I found that DG’s algorithmic labor management system standardized employer-driven “flexible” scheduling by carefully monitoring staffing levels across its 16,000+ stores. This automated system freed upper management to make last-minute scheduling changes while subjecting its workers to precarious scheduling. The result was an unequal distribution of payroll hours amongst DG store locations. Frontline managers relied on a matrix scheduling program called Atlas, which calculated the number of weekly labor hours each store was permitted and outputting an optimal schedule each week. Atlas’ custom-built schedule for each DG location was generated based on an algorithm that factored in various metrics (some obvious and some obscured) such as local store sales from the previous year, projected customer flow patterns, and weekly projects and work tasks. Not even frontline managers fully understood how their store’s hours were calculated. In general, systemic understaffing functioned to ensure maximum productivity and profit by carefully restricting labor expenditures at each DG location. However, in the district I examined, Atlas appeared to critically understaff stores located in the poorest neighborhoods, areas that were also home to more black residents. This practice, I argue, encouraged the reproduction of racial, socioeconomic, and spatial inequality for workers and shoppers by generating consumer redlining.

My research extends our knowledge of consumer redlining by providing empirical evidence documenting its occurrence at DG stores as well as an explanation as to why it occurs in the first place. While automated scheduling systems function to minimize labor cost and maximize profit, they also risk exacerbating the degradation of labor and instigating the reproduction of inequalities—in this case, by generating consumer redlining. My findings also illustrate how the automated management of DG’s labor supply in each store undermined frontline managers’ authority and workers’ ability to resist precarious scheduling. In fact, if store managers attempted to schedule in more paid labor hours than Atlas instructed, they faced possible sanctions from upper management. Therefore, as the only salaried employees in the store, frontline managers were expected to fill any scheduling gaps that Atlas created. Therefore, my case study illustrates how computer technology can intentionally undermine the power of frontline managers, making it difficult to remedy the source of complaint--systemic understaffing. Chronic understaffing detrimentally impacted workers as well as consumers, which should give local officials and policy makers pause before willingly embracing DG (and other low-end retailers) as “job-creators” and boosters of economic development.

Contexts

Precarious Labor Managed by Technological Control

Decades of scholarship document increasing employment precarity with waves of neo-liberal policymaking influencing the spread of employer-driven “flexibility” (Bernhardt et al. 2008; Finnigan and Hunter 2018; Fligstein 2001; Golden 2016; Hacker 2006; Halpin 2015; Henly et al. 2006; Kalleberg 2011; Pedulla 2013; Schneider and Harknett 2019; Snyder 2016). While workplace practices vary greatly across industries, changes in the organization of retail, the largest employment sector in the United States, magnify a larger shift that has taken place within the US labor market. In order to boost short-term profits and maximize shareholder values, corporate strategists developed practices transferring business risks from employers to employees (Boushey and Ansel 2016; Hatton 2011; Lambert 2008; Lambert et al. 2019; Pedulla 2013; Vargas 2017). Scholarship on retail has evidenced both the economic and temporal dimensions of workplace precarity from its low wages and meager benefits to its enlarged involuntary part-time workforce reorganized around “flexible” scheduling (Carré and Tilly 2008; Ikeler 2016; Lambert and Henly 2012; Reich and Bearman 2020; Van Oort 2018). A disproportionate number of retail employees are women and racial minorities, and recent research has found that workers of color receive more unpredictable schedules than their white counterparts, adding to the reproduction of racial inequality at work (Finnigan and Hunter 2018; Lambert et al. 2014; Ruetschlin and Asante-Muhammad 2015; Storer et al. 2019).

In Working for Respect: Community and Conflict at Walmart, Reich and Bearman (2020) explain how Wal-Mart paved the way in risk-transferring retail strategies such as withholding schedules until the last feasible hour and making last-minute adjustments to staffing levels. Positioning workers as a buffer insulated the company from financial risks such as oscillations in the local and broader economy, seasonal fluctuations, and changes in customer flow. These payroll allocation systems rely on algorithmic formulas to facilitate automated staffing based on a variety of metrics and conditional calculations (Boushey and Ansel 2016; Lambert and Henly 2012). They are not so much difficult to develop as they are complex due to the wide range of variables, conditions, and objectives that can be considered in order to effectively estimate “optimal” staffing. Because the logic driving these systems is focused on profit-generation, outcomes have been far less than “optimal” in terms of workers’ needs and consumers’ experiences.

In 2016, Wal-Mart began implementing its “Customer First Scheduling,” developed by JDA Software Group, Inc. (Howland 2016; JDA Software Group, Inc 2018). JDA is presently “the leading supply chain provider powering today’s digital transformation,” with a mission to “help companies optimize delivery to customers by enabling them to predict and shape demand, fulfill, faster and more intelligently, and improve customer experiences and loyalty.” Headquartered in Scottsdale, Arizona since 1985, JDA has developed technologies for over 4000 of the largest retailers, manufacturers, logistics companies, and distributors around the world. In addition to Wal-Mart, some of their prominent retail clients have included Barnes and Noble, Best Buy, CVS, Dollar General, Dollar Tree, Home Depot, Hallmark, JC Penny, Lowes, and Macy’s. In 2002 DG implemented several of JDA’s applications, including its “Open DataBase Merchandising System host transaction system, Arthur Enterprise Suite of advanced planning applications, Intactix space management and assortment planning applications, and JDA Intellect data-mining applications” (Phoenix Business Journal 2002, 1). These systems were designed to reduce expenses, increase profitability, improve visibility, and fulfill promises to DG customers. JDA’s CEO at the time, Jim Armstrong, described their partnership with DG as:

A pocket of strength in this tough economy… [that] once again proves that Tier One retailers have rapidly embraced our value proposition… By offering the broadest suite of integrated best-of-breed applications, combined with our unique implementation engineering approach, it's simply the lowest risk and most cost-effective solution on the market today (Business Wire 2002).

While these maneuvers have helped to maintain the competitively low prices that discount retail customers’ demand, workers have consequently lost even more autonomy over their work schedules, while frontline managers have faced heightened levers of accountability. Today, “firms often hold frontline managers accountable for maintaining a tight link between variations in consumer demand and outlays for wages, reprimanding them for exceeding weekly or monthly labor hours provided for scheduling purposes” (Lambert and Henly 2012, 145). Importantly, although Wal-Mart helped normalize retail management practices that minimize employee costs in order to boost short-term profit, they are not the only model. Work schedule stability is a primary indicator of job quality in the retail sector, with those holding full-time status generally faring better than part-time workers (Carré and Tilly 2008; Ikeler 2016; Luce et al. 2014). Some retailers have attempted to operate according to the philosophy that investing in labor (through higher wages, full-time employment, stable scheduling, quality training, etc.) can improve the bottom line (Boushey and Ansel 2016).

The development of algorithmic systems to facilitate workplace decision-making extends far beyond the retail sector. Research has questioned the ways in which technology-driven managerial systems have been used to structure variations of digitized risk assessments and rapid decision-making regarding sentencing and bail, the distribution of public benefits, healthcare, financial credit, education, and more (Angwin et al. 2016; Corbett-Davis et al. 2017; Eubanks 2018; Monahan 2017; Van Oort 2018). Critical surveillance studies scholars also question the role of these technologies in the sorting of the economically and socially marginalized, suggesting that an over reliance on these systems may exacerbate inequalities (Browne 2015; Haggerty and Samatas 2010; Monahan and Torres 2010). In 2018, the American Civil Liberties Union voiced its concerns over the power of unchecked technological decision-making, stating:

It remains incredibly difficult to assess and measure the nature and impact of these systems, even as research has shown their potential for biased and inaccurate decisions that harm the most vulnerable. These systems often function in oblique, invisible ways that are not subject to the accountability or oversight the public expects (Reisman et al. 2018, 1).

Retail Precarity Triggers Consumer Redlining

While lean labor strategies and workplace optimization systems have sought to align retail staffing with real-time consumer demand, just-in-time scheduling software has not been a foolproof system of organization. Retailers looking to improve customer service tend to focus on two complementary areas: increasing the variety of products offered and increasing interactions with customers (Carré and Tilly 2008). Yet when staffing is lean fewer workers face increased demands combined with work speed-up, undermining customer service (Vargas 2017). Research has shown that cutting corners on staffing affects profits and productivity, making unpredictable schedules and its resulting high turnover counterproductive, driving down sales due to poor customer service. “In the short-term, this could mean customers leaving without buying anything, or losing regular consumers to competitors. In the long term, social media websites such as Yelp, Facebook, and Twitter have made businesses more worried than ever about how word-of-mouth may damage their reputation” (Boushey and Ansel 2016, 8). Understaffing at Wal-Mart, for example, has led to extensive complaints against stores by customers and employees alike who report resultant problems such as long lines, poor inventory, unkempt facilities, and missed sales opportunities (Reich 2016; Reich and Bearman 2020). For frontline managers, the pressure to keep within the allotted hours makes it difficult to fulfill everyday work tasks and performance requirements.

While ‘retail redlining’ has been documented for decades by researchers across several academic disciplines, ‘consumer redlining’ is a distinctly different practice (Benston 1979; D’Rosario and Williams 2005; Kwate et al. 2012; Myers et al. 2011; Zhang and Debarchana 2016). The term retail redlining has been used when referring to retailers that avoid opening stores in communities with more black and/or Latino residents. It has also been used to describe the significantly higher price differentials that may exist between goods and services in poor urban neighborhoods compared to other locations. By contrast, while retail chains guilty of consumer redlining may locate their stores in racially diverse communities and offer competitively low pricing, stores are noted for their substandard customer service in comparison to sister stores in white(er) communities (Reich 2016).

In 2015, Wal-Mart expanded its attempt to reach more of the urban populace by opening stores within the city limits of locations such as Washington D.C., Chicago, and Atlanta (Berliner 2015). The company is notably the largest private employer of black workers in the US, who make up nearly 46% of their 1.5 million employees (Making Change at Walmart 2018). Yet Reich’s (2016) findings provided a warning as to the hidden consequences of Wal-Mart’s expansion into more racially diverse communities. His extensive analysis of 35,000 online Wal-Mart Yelp reviews covering 2,840 stores revealed that those located in lower-income communities of color consistently received lower customer satisfaction scores than ones in wealthier, whiter communities. In fact, the racial composition of the surrounding community was more strongly associated with negative customer reviews than income level alone. Poor black neighborhoods were host to the lowest ranked Wal-Mart centers on Yelp, places that were frequently described by customers as the “worst,” “unorganized,” and “nasty,” with the most highly correlated word description being “ghetto” (Reich and Bearman 2020). By contrast, stores located in white communities were commonly described using language such as “typical,” “friendly,” and “smaller.”

Reich’s (2016) research remains the only sociological study, until now, to address the phenomena of consumer redlining head-on. In that respect, it is ground-breaking. However, he did not endeavor to explain why the racialized gap exists in the first place. He hypothesized that it was perhaps due to Wal-Mart establishing local monopolies in poor urban communities of color—and hence there being few competitive alternatives for local residents. Yet this hypothesis does little to explain why those stores were characterized by relative negligence. By contrast, in this article, I draw upon an in-depth ethnographic study of Dollar General stores to forge connections between precarious retail work, algorithmic management systems, and consumer redlining. I argue that consumer redlining—evidenced by disparities in payroll, customer service, maintenance, and more—are grounded in cost-reducing strategies implemented by corporate management’s embrace of algorithmic systems of control. In addition, I document how consumers’ constant grievances impacted workers, shedding light on employee-customer relations at redlined stores from the inside.

Methodology

The sociology of work has been steeped in a rich history of ethnographic research that has provided deep and contextualized understandings of work, workplaces, and occupations. In the service work literature ethnography has been well-established as the gold standard for investigating the organization of work, customer service, and class relations (Bolton and Houlihan 2010; Hanser 2008; Leidner1993; Sherman 2007; Smith 2001; Williams 2006). It has also been the preferred method for researching the lived experiences of poverty, offering new perspectives upon, and correctives to, previous sociological understandings (Anderson 1999; Duneier 1999; Newman 1999; Small and Newman 2001). Ethnographies of work have had a profound influence on our understanding of the social-relational dynamics and lived experiences of inequality (Smith 2001). By humbling one’s authorial posture as a researcher, “Even through self-reflective passages, the subjects remain at center stage” (Katz 2004, 281).

The ethnographic data I draw from in this paper was collected for a project animated by a series of interwoven questions about low-wage retail work and intra-class service relations between workers and customers. I considered the dollar store to be an interesting and under-researched economic and social space and set out to obtain employment at any of the three major dollar store retail chains in the US—Dollar Tree, Family Dollar, or DG. While I was limited to submitting paper applications at individual Dollar Tree locations, jobs at Family Dollar and DG were accessible by applying online. Therefore, I filled out application profiles through their digital employment portals and submitted them to numerous individual job postings listed on their websites. This equated to well over fifty different applications submitted to dollar store locations in multiple states. I was selective about the personal information I included on these applications, listing my highest level of educational attainment as a BA in Sociology, despite the fact that I was currently enrolled in a PhD program. Thanks to my working-class background, I was also able to be selectively honest about my previous employment experience, listing Burger King team member, Sears sales associate, and Maintenance Cleaner at my undergraduate university as my previous work experience.

Obscuring my role as a researcher was crucial in order to gain full access to a dollar store fieldsite and maintain my employment there. In fact, merely acquiring the job took me nearly a year of applying and I feel quite certain that I would never have been placed in my Downtown DG fieldsite store had I fully disclosed. In the case of workplace ethnography, full disclosure can result in being denied access by management altogether or pressured to heavily sanitize one’s findings to protect the business under investigation (Galliher 1980; Hilbert 1980; Israel 2014; Lofland et al. 2005; Nader 1972; Orzechowicz 2016). Furthermore, the mere presence of a researcher can dramatically alter fieldsite dynamics and research findings. For these reasons, there has been an extensive and long-standing tradition within workplace ethnography to obscure one’s role as a researcher (Braverman 1974; Diamond 1992; Ehrenreich 2001; Hanser 2008; Lamont 2000; Leidner 1993; McDermott 2006; Newman 1999; Orzechowicz 2016; Purser 2012; Williams 2006). DG also enforced several policies that frontline managers and workers interpreted as nondisclosure agreements, even requiring employees to sign a document stating they would not discuss the details of their job with the media or on social media. Employees believed they would be written up or possibly terminated for providing information about DG operations to inquisitive outsiders. Therefore, I found that concealing my role as a researcher was a necessary step in order to achieve my research objectives while providing employees with the highest degree of protection from retaliation. Obscuring my role as a researcher until I exited the field also granted me a more authentic work experience than I could have possibly achieved otherwise. I had access to behind-the-scenes information, conversations, and everyday interactions that would likely have been concealed had I somehow been permitted to conduct my ethnographic fieldwork under the terms of managerial approval.

My DG interview was conducted by Hiring Manager Jackie, a white woman in her late thirties, at a store located in a suburban strip mall. It featured a huge, freshly paved parking lot that welcomed droves of predominately white shoppers. When I entered the automatic sliding doors around 10:00 AM that morning, I was struck by the store’s immaculate linoleum floor, which had small flecks that sparkled under the bright LEDs. I tracked down Jackie in one of the food aisles. She was teetering on a stepstool stocking merchandise and sprinkled head-to-toe with flakes of cardboard. This, in combination with her faded black work clothes and windblown hair, gave a frazzled appearance. I complimented her on the cleanliness of her store as she nodded, leading me to her office in the back of the stockroom. Jackie informed me that she was not actually hiring help for her store, but rather for the Downtown DG store. At that time, I was not sure which DG location she was speaking of. She explained that while the two DG stores were just four miles apart, they were like two different worlds. My fieldnotes recollect this portion of my job interview:

Jackie told me that she likes to paint “the worst possible picture” of the Downtown store and its location to give new hires “a clear vision of how bad the store is.” She explained that she did this so that when a newly hired employee arrives to work they are not “shell-shocked.” She relied on the term “shell-shocked” a lot during our interview, claiming that the last few sales associates she’d hired were all “shell-shocked” by the store and quit… There was a methadone clinic located in the office space above the Downtown store, so “there are drug addicts all around” who come into the dollar store. District Court and a handful of halfway houses surround the store as well, she warned, so the area is riddled with “convicts” and “offenders.” Also, the store is located across from Central Park, so “pretty much any type of person you’d imagine that hangs out in parks is there [in the dollar store].” A development of public housing called “The Towers” loomed a few blocks away. “A lot of people [living there] don’t have jobs,” she informed me. She firmly believed that many of its residents also had “mental issues,” warning that former DG employees voiced a fear of “getting stabbed by customers.” She assured me that nothing like that had ever occurred, though workers were sometimes threatened. “…But you get that anywhere,” she added.

The thirty minutes I spent listening to Jackie in her office was not really a job interview, but rather a stern lecture on employee honesty interwoven with grave warnings about the neighborhood surrounding Downtown DG. I sensed that she had decided to hire me the moment she’d met me, or perhaps even before that moment. She also appeared evidently fearful that I would immediately quit once arriving at Downtown DG. I suspect my identity as a young white female was a major factor in her concern. Rather than asking about my background or previous job experience, she elaborated on topics such as employee theft and her perceptions of the “social derelicts” plaguing Downtown until our time was officially up. She appeared to be studying my body language and gauging my reactions as she spoke, but I remained calm, matter-of-fact, and inquisitive rather than displaying signs of alarm or trepidation about the job. Therefore, I never needed to engage in further deception to obscure my role as a researcher because next to nothing was asked of me during my interview. Later that week, I received a call from Stewart, the Downtown DG store manager, who informed me that I was hired to work at his store. After completing and passing my background check and urine sample drug test I was cleared for my first day of work.

Downtown DG was located in the hollowed out urban core of a deindustrialized Northeastern city, a place classified as “distressed” by the state’s Department of Community and Economic Development. The city’s overall poverty rate hovered around 38%, with a medium household income of about $23,600 (City-Data 2016). However, due to residential spatial segregation the city’s black population was concentrated in neighborhoods where the median household income was under $10,000. For six months I conducted ethnographic fieldwork while working as a part-time sales associate earning $7.75 an hour. Overall, I averaged twenty-four and a half hours per week, but my schedule was incredibly erratic and ranged anywhere from eighteen to forty hours a week. I also relied on Medicaid health insurance during this time since DG’s healthcare option came at an exorbitant cost in relation to what I was paid. My participant observation located me within an intense and continuous atmosphere of negotiated relations where I “learned by doing” (Gerson and Horowitz 2002). This learning by doing provided me with an embodied, visceral understanding of dollar store workers’ and customers’ experiences like no other method possibly could. Due to DG’s high turnover rate, especially at my Downtown fieldsite, by the end of my fieldwork I was already the senior cashier and had turned down a promotion for keyholderFootnote 1 twice. Therefore, I found that my six months of employment was a sufficient amount of time to reach the point of saturation in my fieldnotes.

To expand the scope of my fieldwork, I also conducted periodic observations at the eighteen other DG store locations within my fieldsite’s district, visiting them as a customer. I obtained the list of store locations at my workplace and visited each store (at least twice and some over ten times) to evaluate store conditions. I would walk the aisles jotting down notes, take photos with my phone, and have casual conversations with employees and customers. I would also ask permission to use the restroom if it appeared to be located in the stockroom so I could gauge their inventory and glance over their store data charts posted on the wall. I paid careful attention to the volume and variety of African American hair care products each DG store stocked because it was a good indicator, besides visual observation and census tract data, of the number of non-white customers the store serviced. I would also make small purchases to gauge the register wait time and ask other customers in line about typical store conditions. During the course of my research I also lived approximately four blocks from one of the DG stores I identified as consumer redlined.

To gauge the validity of my fieldwork and collect a broader account of DG employees’ experiences, I conducted fifty in-depth interviews according to Yin’s (2014) “case study logic.” This interviewing method proceeds sequentially so that each interview “provides an increasingly accurate understanding of the question[s] at hand… The number of units (cases) is unknown until the study is completed” (Small 2009, 25). By design, this interview sample collection technique aims not for equal representation of all worker experiences, but a deep inquiry of the research question(s) and developing themes to the point of saturation. “If the study is conducted properly, the very last case examined will provide very little new or surprising information” (Small 2009, 25). Therefore, speaking more candidly with DG employees about their work experiences looked to solidify my fieldwork data, identify any new information that differed from my own observations, and cement the core themes that I had honed in on during my fieldwork.

I first began by conducting in-person interviews with my fellow coworkers, those whom after I revealed myself as a researcher were willing to participate.Footnote 2 In the end, five out of seven coworkers at my fieldsite, including both the store manager and assistant manager, agreed to be interviewed. Those who declined were newer lower-level associates with whom I did not work much. One other coworker who declined an interview was Serena, a keyholder who told me that she considered DG her career and, as much as she “complained about the company…[she] liked her job and wanted to keep it.” Therefore, it was understandable that Serena valued her job more than participating in my research. These initial interviews led me to snowball sample interviews with five additional DG sales associates from the district.

Next, I attempted to recruit additional DG employees within a seventy-mile radius of my fieldsite by using posters and having informal conversations with workers in the stores I visited. However, the innumerable attempts I made to recruit in-person interviews within and near DG stores were unsuccessful. I believe this was at least partly due to the company’s culture of suspicion, secrecy, and discipline. After developing a second consent document and recruitment script I was granted an amendment to my initial IRB in the spring of 2015 to conduct phone and online interviews. I posted my online recruitment script on a variety of social media sites including Facebook, Instagram, Twitter, and LinkedIn. I also shared this call for participants by creating two Facebook pages to connect with dollar store workers. An unanticipated benefit to recruiting online was that my participants could easily keep in touch with me via text or social media after our initial interview. Some workers continued to update me on their employment progress/status at DG and would send me digital photos and video of their stores on their cell phones, connect me with their coworkers, or simply check in to ask how my research was progressing.

As for the practice of conducting the interviews, my objective was to mimic Bourdieu’s strategy of “active and methodical listening,” where the researcher initiates a relationship with their participant by providing their undivided attention, submitting to the subject’s singular worldview, and adopting/adapting to their own language, views, feelings, and thoughts (1996, 17). I conducted these interviews between 2015 and 2016, designing a script of questions to generally direct the interview. Each interview lasted between forty-five minutes to two and a half hours. I alternated between asking scripted questions and mining for more detailed information on specific topics of interest pertaining to my research themes. Ultimately, my data analysis and writing for this paper was an iterative process. While my discussion is deductively grounded in Reich’s (2016) conceptualization of consumer redlining, it was an inductive process of thematic coding using HyperResearch software that led me to seek out his research.

Slashing Labor Using Algorithmic Control

Payroll Formula

One of the defining features of DG’s organization of work was a “flexible” labor budget that factored in store sales. A key register statistic was the “average basket price,” a metric that expressed the average dollar amount a typical customer spent during one shopping trip within a particular store. Customer transactions were aggregated into a weekly average basket price and average store sales, data which was then used to predict staffing needs the following year during the same time period. Therefore, DG locations where customers could afford to spend more were provided more paid labor hours each week, which translated into more workers operating the store, more work being accomplished, and better customer service. Paula, a thirty-six-year-old employee earning $9.50 an hour, explained that, at her DG store in Florida, most customers were “low-income African American or Haitian.” Because these customers were “very poor,” they could not afford to spend much within the dollar store, which translated into lower payroll and understaffing. Their entire DG staff consisted of just seven employees, similar to my own Downtown DG fieldsite. “We’re always rushed,” she protested, “And I feel we’re understaffed. Corporate, however, said we have plenty of people based on sales.” Corporate-level manager Brad, a thirty-eight-year-old white male who had worked for DG seven years, explained the critical link between a store’s payroll budget and its sales:

Payroll is based on sales. When smaller volume stores don't produce the sales of their higher volume cousins, the store manager has to work to make up the difference in labor because the smaller volume stores don't get the payroll… It makes it nearly impossible to manage the store. This lack of payroll then translates into poor store conditions, worse customer service, and higher merchandise shrink.

In this way, each DG store’s labor budget was a direct reflection of the surrounding community’s purchasing power.

While store managers were not entirely sure how payroll was calculated, it was generally agreed that store sales were the largest component determining labor hours. This demonstrated how DG’s labor allocation algorithm was enshrouded in mystery, even to frontline managers. As one store manager tried to explain, “It’s a complex algorithm that generates a dollar figure for payroll, which is a percentage of sales expected.” A district manager had a slightly different answer, “It’s decided based on several factors: Sales trends typically five to six weeks rolling, sales from last year, funded projects…also certain store will have their hours adjusted based on the environment or location. Most importantly and what’s not shared is [that] a store can get their hours bumped up with an exception typically approved by a regional director. There are usually conversations had between the district manager and regional director.”

The regular shoppers that frequented my Downtown DG fieldsite were residents who lived nearby and commuters who worked in the city. Both relied on the dollar store for resources since there were almost no other businesses in town that sold groceries, household items, and other necessities. A significant portion of customers received public assistance and were easily identifiable by their Electronic Benefit Transfer (EBT) cards. The neighborhood had an extremely high number of residents in deep poverty, with a whopping 89.7% earning less than $10,000 per year. In employee interviews, I asked each participant to estimate the percentage of shoppers at their respective stores who used EBT and the most common response (mode) was 70%. When I shared this statistic with corporate manager Brad, he affirmed its legitimacy, stating, “Workers that mentioned 50% to 80% of customers using EBT, they’re right...Go to a [DG] store on the first and fifteenth of the month [when benefits are distributed] in the South and you’ll see 100% of customers using EBT!” His comment about stores in Southern states, where black poverty is widespread across rural and urban regions, was another signpost linking poverty, race, and the quality of customer service at DG stores. It also rang true at my fieldsite, where, like clockwork, the Downtown DG would be swamped with customers during the first week of the month when cash benefits and food stamps were disbursed. Waves of shoppers utilizing EBT cards wiped our store shelves nearly bare, leaving a disorganized trail behind. Our store manager, Stewart, would go on rants, complaining how one cashier (often me) took in $7000 to $8000 in a shift, yet upper management repeatedly denied his requests to staff a second cashier. This was important because, unless a DG store could significantly increase its sales (average basket price), it was nearly impossible to increase the store’s labor budget.

As a ploy to receive more paid labor hours, Stewart instructed me, and other employees, to entice customers into spending more money by pushing various deals and impulse buys at the register. I found this task incredibly difficult, however, since most of our shoppers were by no means frivolous spenders. Customers in general were on limited budgets, so once the money was gone it was gone. These tactics also pressured workers into extracting additional profit from customers through the beguiling “promise” of more labor hours. Yet selling more merchandise to cash-strapped shoppers never seemed to translate into additional labor hours, certainly not to the point where two cashiers were ever permitted to be staffed at the same time at my fieldsite. An additional obstacle to perpetual understaffing was that the Downtown DG did not have a parking lot. Only a handful of metered spaces lined the block, which were usually occupied. This repelled potentially more well-off customers with vehicles and attracted those traveling by foot or bus that purchased only what they could carry by hand or transport in small metal utility carts. This additional constraint ensured that Downtown DG’s payroll budget, based on the “average basket price,” remained low.

Scheduling Matrix

DG’s meager labor budget frustrated employees, who struggled to get by on a patchwork schedule of low-wage part-time hours. The corporation required that store managers use a just-in-time scheduling program called Atlas, which automatically distributed the weekly labor budget amongst employees. Yet, it was difficult for Atlas to produce a reliable schedule beyond one week in advance. These findings mirror erratic “flexible” scheduling at the forefront of workplace grievances and employment policy-making nationwide (Golden 2016; Halpin 2015; Williams et al. 2018). At my fieldsite, the work week began on Saturday and ended on Friday, with the new schedule typically posted two days before the following work week began. This meant that I would learn on Thursday if I’d be working Saturday. Worse yet, the schedule Atlas produced was not held static during the week, but continued to be tweaked and altered. Company policy italicized below the schedule read, “Scheduled hours are subject to change based upon business needs.” Ginger, a fifty-year-old store manager from Missouri, reported that upper management often dictated these last-minute schedule changes via phone or email, explaining, “A lot of times, after the schedule for the week was completed, the district manager would call and say we needed to cut hours given.” On other occasions, a last-minute email from corporate would instruct all stores in the district to reduce their payroll by X amount of dollars over the next few weeks. Atlas often forced double or even single coverage, where one or two employee operated the entire dollar store. Since the store manager was the only salaried employee, they had little choice but to fill the scheduling gaps that Atlas produced. Understaffing regularly resulted in the manager on duty working their store alone because they were not provided enough paid labor hours for the week. Several times during my fieldwork I, or one of my coworkers, arrived at the Downtown store for our shift only to leave minutes later because the schedule had suddenly changed, yet no one bothered to alert us. It was deemed to be the employee’s responsibility to phone the store or drop by on our days off to check the schedule for updates. The worst was six AM stocking shifts in the dead of winter when I’d brave the subzero wind chill and snow to reach work and find the lights off and doors locked. These “flexible” scheduling tactics exemplify Brian Halpin’s (2015) concept of the “mock calendar,” a micro-level management strategy that obscures shifts, changes, and daily manipulations within precarious employment. The mock calendar “mystifies the very conditions that it creates” while leaving “workers vulnerable to managements’ requests” (Halpin 2015, 435). It also functions to minimize workers’ paid labor hours on the clock as it maximizes employees’ productive capacity.

Erratic work schedules took a physical toll on DG workers, while the instability in week-to-week earnings ensured a perilous monetary existence. As a result, DG workers (mirroring the poor customers they serviced) often relied on food stamps and other forms of Public Assistance to get by. DG did not even provide an employee discount, despite workers frequently purchasing items within the store. The hourly wage of a lower-level sales associate within my sample ranged from $7.15 to $8.50, and keyholder “managers” did not earn considerably more at $8.85 to $10.10 per hour. To my knowledge, at least tenFootnote 3 workers received food stamps during my fieldwork. I learned this by servicing them at the register while they were on break or shopping in the store after their shift.

Understaffing Resulted in Second-Rate Customer Service

According to DG’s Customer SurveyFootnote 4 there were four levels of customer satisfaction: Level 1 was the best, followed by Levels 2 and 3, and finally the “Did Not Meet” customer satisfaction category. At the end of each month, our store manager would post the Customer Satisfaction Report results on the wall of our stockroom. Out of the nineteen DG stores in our district, my fieldsite Downtown DG was consistently ranked last. A bar graph illustrated the difference in customer satisfaction between stores within the district. In the month of December 2014, the end-of-the-year statistics showed that the store’s customer satisfaction score was an abysmal 22.2%, compared to the District average of 54.2% and a Regional average of 56.8%. Therefore, my fieldsite DG remained in the “Did Not Meet” customer satisfaction category each month. The tasks listed in the “Performed Well” column remained blank. In the “Needs Improvement” column, the lowest satisfaction score was neatness (16.3%), followed by cleanliness (19.4%), overall satisfaction (25.8), and fresh products (27.3%).

DG provided some incentive to solicit customer feedback, with each receipt advertising a phone survey that entered participants in a monthly drawing for a $1,000 cash prize. Plenty of poor customers at my fieldsite were willing to trade several minutes of their time for the chance to win money. A few even commented to me that they took the phone survey every single time they received a receipt, hoping to win. Store manager Stewart instructed me to circle the survey code advertisement on each receipt in red pen and persuade customers to call in to “say good things about the store.” However, this plan often backfired, because customers would call the number to complain. A few shoppers even dialed the corporate number on their cell phones right in front of me in protest of our store’s deplorable conditions. For example, one afternoon a regular shopper, a black woman in her forties, got on her cell phone while standing in front of my register and announced she was reporting me. “I’m in the dollar store,” she shouted into her cell phone. “These prices are never right and I bitch about it.” A second customer in line laughed out loud and said sarcastically, “I’m sure you’ll remember her.”

The Southside DG store had the second lowest overall Customer Satisfaction Survey score in the district at 33% and the Westend DG had the third lowest at 35.6%. Together, these three DG locations were only about a mile away from each other and had similar neighborhood demographics. According to census tract data, not only were the majority of their inhabitants in severe poverty (68–90% earning less than $10,000 per year), but they had the highest proportion of black residents (10–26%) in the DG district. Regular customers assured me that Downtown DG’s bare shelves, filthy aisles, and never-ending register line were not a recent occurrence, but had been an ongoing problem for as long as they could remember. Serena, a black coworker in her twenties, also affirmed this fact, telling me that over the past six years she had been employed at the store conditions had never improved. What she had witnessed during this time period, however, was a lot of turnover, including two district managers, four store managers, and an innumerable number of sales associates come and go. This confirmed that the troubles our “metro” store suffered were not simply attributed to one “bad” manager or specific team of employees, but were reflective of a deeper issue.

Store manager Stewart would leave the emails he received from district manager Courtney up by the register for employees to read, initial, and date. These emails often contained direct quotes from the customer satisfaction survey. One morning, after scanning the complaints on the list, I pointed out a particular quote to him, which described the manager and sales associates as “rude to the customers.” “It’s not necessarily about me,” he said, “But if it is, I just give back to the customers what they deserve.” Other complaints on the list included “The store was a mess,” shelves with “nothing stocked,” being “told the wrong information” about voids and returns, “The cashier did not speak to me,” and “The manager was on their phone the whole time.” Many excerpts from my fieldnotes documented how frustrated and angry local shoppers were with Downtown DG’s conditions, especially the unceasingly long register line:

Within the past two days our store received twelve phone complaints from customers about the long register line. As a result, the district manager sent an email to Stewart instructing all employees to operate the registers during peak business hours, 11 AM to 3 PM. In addition, we are forbidden from taking our meal break during these hours. The problem is our store only has two functioning cash registers to begin with. Also, the store is usually staffed by only two employees—a lower-level sales associate (me) and a manager on duty. When Assistant Manager Russell read Courtney’s email this morning he reluctantly sighed, saying to me, “Well, I was going to try to put a dent in our inventory and do some stocking this morning, but I guess I have to stay up here with you.”

Our district manager’s attempt to implement this new informal work rule to address our store’s customer service complaints lasted a little over one week. Last-minute, piecemeal rules never lasted very long at our store because the manager on duty had too many other responsibilities to take care of within the store. It was physically impossible to stay up front operating the register during their shift and accomplish the list of basic work tasks in the store. Given that corporate was unwilling to divvy out more labor hours to staff two cashiers per shift, the problem of a long register line persisted throughout my fieldwork. In fact, nothing was ever done to directly address or permanently remedy any of the service issues our store dealt with. Whatever consumer “power” shoppers believed that they had over DG’s organization of work when they called the corporate number was fundamentally superficial.

As a sales associate, I typically operated the cash register alone during my shift while the manager on duty accomplished other work tasks in the store, stockroom, and back office. The only time I interacted with the manager on duty was when a customer needed a price change, void, or return because these actions required register keys. Trying to improve store conditions in any capacity was always an uphill battle with only two people operating the entire dollar store. It was extremely difficult to accomplish the necessary daily tasks upper management required—stocking, price changes, cleaning, let alone adequately servicing our customers and watching for shoplifters. I regularly spent the entirety of my shift jogging between the checkout counter ringing up customers, stocking and/or recoveringFootnote 5 shelves, cleaning up disasters in the aisles, and assisting shoppers with findings products. While the Downtown store’s vicious short-staffing ran me ragged, for the first few months of research I held on to a naïve optimism that conditions would improve. Over time, however, my sunny outlook morphed into frustration and cynicism as I watched all of my hard work implode into chaos. For example, after spending several hours arranging the shelves while simultaneously ringing up customers at the register, the aisles would be decimated again by a new wave of customers flooding the store. Other times, I’d be several boxes away from stocking last week’s merchandise before a new distribution truck would arrive and pile 1,000 more boxes of product in the stockroom. The tools I was provided to clean with were laughably meager –a roll of paper towel, a spray bottle of cleaner, and a filthy mop bucket. I was even rebuked by Stewart early on in my research when I used the contents of the spray bottle up “too fast” because our budget for cleaning supplies was practically non-existent. Under these circumstances, my fieldnotes became an outlet for venting my frustrations. By the end of my six month stint, I actually fit in rather well with my Downtown DG coworkers. I had learned a lesson they already seemed to have known—that in order to survive the job, you had to lower your expectations, pick your battles, and numb yourself to the constant failures and criticism.

Corporate’s “Model” Stores and “Metro” Stores

After six months of fieldwork, talking informally with hundreds DG employees and customers, visiting dozens more store locations, and conducting fifty employee interviews, I concluded that my Downtown fieldsite had suffered the all-too-common fate of “metro” stores. “Metro area,” “metro market,” or just “metro” stores were common terms used by DG employees who held upper-level managerial positions, including district managers, regional managers, and corporate-level managers. The dichotomous labels signified awareness of significant differences between DG’s rural and urban stores. As one corporate-level manager explained during our interview, “The best metro store does less [profit] than the best rural store, for certain…The new strategy of pursuing DGxFootnote 6 format stores in dense urban areas yield stores that do lower sales than DG stores in rural areas. Profitable stores get remodeled first and profitable areas get pursued harder.”

District manager Jordan had spent five rigorous years working eighty hours a week overseeing a cluster of rural and suburban Dollar General stores. However, in his sixth year the company decided to redraw the district lines for stores within his region and Jordan found himself unexpectedly reassigned to a new district, which included several “metro” DG locations. “Metro area” DG stores, Jordan insisted, were much more challenging to manage than rural and suburban stores. A key reason why was their insufficient labor budget. During our interview, he explained his interpretation of DG’s labor model and his apprehension towards managing “metro area” stores:

DG pays the bottom. That’s their business model… And the most, the biggest expense that’s controllable, that we can control, is labor… That’s why you see disgruntled workers, because they pay the bottom and they watch their labor and, you know, it is what it is. If the store gets backed up, they [corporate] say that the store manager is not on the processes and they’re not good at staffing and they’re not good at training and they’re not good at building the team. Those urban metro areas are very challenging. Very very challenging. In fact, as a district manager, when I switched to take over the city, I was very very nervous. I almost didn’t take the position…because I was going to be a district manager of a metro area. I have all the metro problems that I did not have. I had one metro area before. It was a small city, but I didn’t have the big metro problems like what I have now... That’s where you hear the nightmares from district managers. There’s no quality of life. They’re working seven days a week. They have stores that are calling constantly because they have troubles with staffing, people walking out. They’re struggling to keep the doors open. And as DG grows, I have friends right now that have like twenty-two stores in metro markets. They have twenty-two nightmares and they hate their job.

For DG “metro” dollar stores, understaffing combined with heavy foot traffic and ongoing customer demands resulted in serious problems, which made DG stores located in rural and suburban communities comparably appear to be “model” stores. Within DG’s corporate culture, it was common knowledge that cities hosted the most blighted and stigmatized of store locations, as Jordan’s tone and language during our interview expressed. His interview excerpt also demonstrated a central theme that emerged within my research findings: DG’s “model” stores were regularly juxtaposed to “metro” (“problem”) stores. In this way, the geographic, economic, and racial composition of the surrounding community was disturbingly indicative of a DG store’s quality and customer service. “Model” stores were located in more financially stable, predominantly white suburban and rural communities, while “metro stores,” or “problem stores,” were located in more economically challenged, high-traffic areas with more racial minorities.

One of the only participants in my research who came close to verbalizing this discrepancy in racialized terms was a black twenty-nine-year-old single mother named Tonya. Tonya had been employed at my Downtown fieldsite for several years, but now supervised a cluster of self-checkout kiosks at the local Wal-Mart. However, she continued to frequent the Downtown DG because it was much closer to her apartment and the store offered certain products cheaper. She also enjoyed chatting with employees, so we formed a friendly association. While talking with Tonya one day, she compared two DG stores in the local area that served a higher proportion of people of color (my Downtown fieldsite and the Southside DG) to the Spring Street DG, which served primarily white shoppers. “I think they definitely get put on the back-burner Downtown. ‘Cause Spring Street, I remember Spring looks the bomb. I’ve been in there a few times and that thing is huge. It’s like really nice in there… Southside and Downtown get neglected.”

My fieldwork affirmed that all DG stores were not equally maintained. “Model” stores had more employees visible on the store floor, abundantly stocked shelves, cleaner facilitates, multiple cash registers open, shorter waits for service, and newer technology (cash registers, heating and cooling systems, security, etc.). However, “metro” stores appeared much like those described in Reich’s (2016) research findings on Wal-Mart—understaffed, understocked, dirty, dilapidated, messy, and slower service. In my conversation with Corporate Manager Brad, he explained how DG “has been toying with more and more urban areas in the last several years, but approximately 3,000 of those stores aren’t performing well against their rural counterparts.” He described the “corporate store” on Conference Drive in Madison, Tennessee as “gorgeous.” “It’s where the company films any commercials or videos to show how awesome a DG store is. The CEO does his interviews and such there. It is the definition of a ‘model” store.” However, what most people did know, he claimed, was that the store operated at a loss “because of all the payroll and expense it takes to keep it ‘model’ status. Then the company goes back and tells every other store they should look like the corporate store, but on a fraction of the payroll.”

The inconsistencies between DG’s “model” stores and “metro” stores were also apparent in how upper management responded to repair and maintenance requests. While some store locations had efficient heating and cooling systems, others, including my Downtown fieldsite, endured hot summers and freezing winters. On eighty degree days, the air conditioning was non-existent and the only relief came from a mini fan that store manager Stewart haphazardly duct-taped above the registers. Chocolate candy bars on the slanted grab-and-go shelves melted in their wrappers, dripping down to form a lumpy mass at one end of the package. In winter, employees dressed in layers because the checkout station was directly in front of the doors. A fresh blast of frigid air would blast me every single time a customer entered or exited the store, so it felt like I was working outdoors. One particular evening, I encountered a male customer who worked at the Southside DG across town, the same store Tonya had likened to my fieldsite during our conversation. The man bitterly complained that all of the Southside employees there were wearing coats in the store during their shift because the heat had been broken for two years. Although his store manager kept submitting tickets for a repair, no one ever came to fix it. “You’re lucky,” he lamented. “At least it’s nice and warm in here.” I was baffled by his comment and could hardly imagine how cold his store must have been.

Heating and cooling issues were a surprising theme in my data. However, with a bit of online research it became clear that these problems were common and recurring ones for DG’s urban stores, especially stores in the Southern US Over the past decade, various online message boards and local news outlets across the country had reported the same conditions again and again (ComplaintsBoard 2018; Glassdoor 2018; Pratt 2014; Summers 2018; WNCT-TV 9 2013). Training Manager Jerry provided his own personal insight into this phenomenon. After working at numerous DG stores across South Carolina, he was convinced that upper management purposefully ignored heating and air conditioning repairs in poor urban communities, which angered him greatly:

At one store, the air went out in March. Once it got into May and June, it became unbearably hot. This is South Carolina, after all. Well, the store manager would call it in over and over again, and then check back on the status of his request. As this dragged out over the summer, it became obvious that nothing was going to be done about it. In fact, the store manager, who was an elderly man, passed out in the store at one point… I called the company, called my district manager, and suddenly everything made sense. DG was not going to spend any money on air conditioning. So to save a few hundred bucks, they lost a store manager, no telling how much product, and probably more than a few customers. It was at this point that I truly realized that the employees out in the stores meant absolutely zero to the people at corporate. The air situation was ridiculous in all the stores.

While poor temperature regulation was a noticeable problem for DG workers and customers at my Downtown fieldsite, the most pressing safety issue I encountered was the store’s floor. If I had documented the floor using time-lapse photography it would have captured the disturbing deterioration taking place over the course of six months. My coworker Willis, a black sixty-seven-year-old sales associate also earning $7.75 an hour, lamented over the condition of our store floor:

There is no maintenance really to enhance the up-keeping of the store. I’m gonna say that 70% or 80% of it is contributed to the fact that the building is old and there is no upkeep on it… Me personally, I think there are some sections of the store that are very very unsafe. Alright? Because you don’t know whether or not if the floor is gonna collapse, how sturdy it is. You know? So that is an issue as far as I am concerned.

Willis’ concerns were not overstated. On truck day, employees struggled to push hundreds of pounds of merchandise stacked seven feet high across the shop floor. I would hear the crack of old floor tiles busting beneath the wheels, leaving a trail of plastic shards and wood splinters behind me. At one point during my fieldwork, store manager Stewart actually slipped on a loose floor tile and injured his knee. The tiles fragmented away in numerous sections of the store to reveal old, warped wooden floorboards underneath. There were large gaps between these floor boards, which Stewart tried to “seal” with duct-tape.

figure b

Other “metro” store employees I spoke with and interviewed reported many more types of unfulfilled maintenance requests and ongoing problems than I can adequately report in this paper. I witnessed caution tape draped around leaning shelves, buckets collecting dirty rain water from dripping ceiling tiles, and broken light fixtures that left whole sections of the store unlit. Workers also sent me photos and videos of maggot-infested food, rat feces covering boxes in the stockroom, and bugs invading the aisles. Importantly, several female workers also discussed feeling unsafe at work because so few employees were staffed each shift. Keisha, a twenty-two-year-old black sales associate from Chicago, told me she had called corporate anonymously to report injuries, dangerous conditions, and mistreatment by her store manager:

I want to help out this research because I want things to change at DG. Someone needs to speak up and say, ‘This is not right.’ I’m not afraid [to call], because workers need to stand up and fight back. If you speak up to the district manager, they just turn around and tell the store manager and then that employee will start to get less hours [as punishment]. Not enough [hours] to live decently, so it’s tiring and stressful.

Keisha described the issues she encountered as common to other DG stores in her surrounding metropolis. She felt strongly that upper management “used struggling and poor employees” by making them work harder and treating them badly. She also felt that workers were unsafe due to “metro” stores being “robbed daily because we’re in one of the worst sections… I had a friend who worked there [a neighboring DG store] who got a gun pulled on them. It was rough.” Her friend who was robbed at gun point was Maya, a black twenty-five-year-old sales associate who worked a second DG in Chicago. During my interview with Maya, she told me she always felt uncomfortable at DG because “there is no safety at all. All we have is a phone to call security incase anything happens, and by then it already happened… We need in-person security at the store because we’re more vulnerable to thieves and robbers.” Maya was working unaccompanied at the time of the robbery, yet even after the robbery occurred, single coverage on the schedule was common. Nothing changed. Cynthia, a Hispanic thirty-four-year-old keyholder employed at a metro DG in New York was also a victim of an armed robbery. The cashier on duty had been “roughed up and taken away in an ambulance” around 6 PM, the police had left a little after 7 PM, and her district manager instructed Cynthia to reopen the store until 10 PM by herself. “I refused and got written up for it. That describes DG. I’ve had a gun in my face three times, a knife twice; it’s a dangerous place.”Footnote 7

Conclusion

This article advances our sociological understanding of low-wage retail by demonstrating how technological systems have modified managerial practices of control through the use of algorithms and computer technology. My findings captured how the automated scheduling of DG’s labor supply undermined frontline managers’ authority and workers’ ability to resist “flexible” scheduling, contributing to the degradation of retail work. While DG’s labor management system standardized employer-driven precarious scheduling across their entire network of stores, the system also resulted in an unequal distribution of payroll hours amongst DG locations. This initiated critical understaffing for some stores, particularly those located in the poorest neighborhoods, locations that also serviced a higher portion of black customers.

My research findings challenge Robin Leider’s (Leidner 1993) conceptualization of “the service triangle” by complicating the usual picture of management, workers, and customers vying to maximize their interests. Ethnographic data demonstrated that workers and customers held common interests, such as increasing staffing and improving store conditions. In fact, store-level and even district-level management also shared these interests. However, it was corporate-level DG and its managerial logic that governed stores. Resources were spent on remodeling its more profitable stores and opening a record number of new stores each year rather than providing even minimal standards within some of its current stores—particularly those located in poor black neighborhoods. Therefore, my findings illustrate how algorithms may solidify the practice of relative disinvestment while stripping frontline managers and workers of agency to do something about it. These actions ultimately encouraged the reproduction of racial, socioeconomic, and spatial inequality for workers and shoppers by generating consumer redlining.

Within the district I examined, DG stores servicing a higher proportion of poor, non-white, residents had inferior customer service compared to those in more affluent, predominantly white communities. However, it was dollar store employees who were blamed for poor store conditions, viewed as “disposable,” subjected to working within squalor, and left as sitting ducks in areas highly susceptible to criminal activity. This is alarming, given that DG’s financial growth model has been reliant on rapid store openings to keep revenue rising and investors happy. Like Wal-Mart, the DG Corporation’s growth model includes locating more stores in poor, densely-populated neighborhoods. For example, in 2017 the company acquired 322 store site locations “in Brooklyn, NY, Chicago, and other cities” to serve as testing grounds for their new urban growth model (Nassauer 2017, 3). Branded DGx these urban stores are designed to be about half the size of a regular 9000-square-foot DG store. DG CEO Todd Vasos aims to “win in cities to achieve his goal of opening another 13,000 US stores… About a third of the company’s locations are currently in metro markets, and they are moving toward 40%” (Boyle 2018). Yet, my empirical observations show that the presence or absence of a particular type of amenity within a neighborhood is an incomplete indicator of neighborhood investment/divestment. These implications extend Small and McDermott (2006) by demonstrating that the same organizations in different neighborhoods can actually vary quite dramatically. A focus on the presence or absence of amenities alone thus understates the extent of inequality. This more nuanced realization has implications for the literature on neighborhoods, raising concerns that businesses, organizations, and institutions such as bank branches, grocery stores, community organizations, while appearing similar on the outside, may look quite different across neighborhoods.

Future scholars should be careful when drawing conclusions about neighborhoods based merely on the organizational composition of these neighborhoods. Given that the presence or absence of amenities alone is not a sufficient indicator of inequality, my research provides a cautionary warning to local officials and policy makers who may be eager to embrace DG, or other low-end retailers, as “job-creators” and boosters of economic development. More generally, the growing “dollar store economy” may be cause for concern, being that industries and corporations that profit off the poor risk reproducing the very economic and social conditions in which they claim to alleviate, exacerbating rather than alleviating class, race, and spatial inequalities. Further research is needed in order to better understand how consumer-redlining practices operate, especially larger-scale and systematic analyses investigating the use and consequences of algorithms in the workplace. Scholarship on the impact of consumer redlining within the service triangle is also needed in order to explicate the nuances of retail relations amongst and between workers and consumers. On the one hand, my research indicated that the egregious conditions of consumer redlining reinforced feelings of shame and stigma amongst workers. However, on the other, consumers often blamed workers for these very conditions, generating social division and antagonism rather than shared understanding and solidarity.

As a final anecdotal piece of evidence, my research findings also revealed that the DG Corporation has been well-aware of their blighted “metro” stores for some time, yet appear to reluctantly accept the ongoing inequality their “flexible” labor model has generated. This was demonstrated by the fact that, at times, underperforming employees were reassigned to a “metro” store as a form of disciplinary punishment. One such example involved my Downtown DG Assistant Manager, Russell. Russell had previously been the store manager of the nearby Westbridge DG, a suburban “model” store. Russell had stepped down from his frontline management position after district manager Courtney gave him, as he put it, “a mixed review.” He confessed that he had not handled her criticisms or her suggestions well—that he works harder, faster, and longer to improve the store’s profitability. Russell claimed he was already averaging eighty hours a week and was not prepared to invest any more time or energy into the Westbridge DG. He put in a formal request to step down and become an assistant manager. District manager Courtney agreed, but under one condition—that he leave the Westbridge DG and transfer to the Downtown DG. The way Russell saw it, being forced to leave his suburban store to “toil in the trenches” Downtown was a punishment. Just as Courtney had done to Russell, I learned that some district managers had a habit of transferring unruly or underachieving employees to a “metro area” store. Sales associate Emily, for example, described having a strained relationship with her district manager after she was “thrown out” of her old store and placed in a “metro” DG location. “Nothing works right. Register freezes, safe won’t open, locks don’t lock, credit machines don’t work… I tried calling ERC,Footnote 8 but had an hour wait time. I tried calling the district manager. No answer. I called my store manager and all she said was ‘Keep trying to get the district manager.’ I’m about ready to literally have a nervous breakdown, right at the register.”

The profit-driven strategy of major discount retailers like DG appear to contain labor (i.e. keep it at a minimum) in order to maximize worker output and productivity. Yet DG has not addressed concerns from workers and customers about the quality of their brand. While consumers, especially in consumer-redlined areas, have no competing retail options, workers and even frontline managers have little voice.