The Unequal Geography of the Gig Economy

Solving rural poverty in Arkansas will take more than odd jobs from smartphone apps.

A view of Main Street Dumas.
A view of Main Street Dumas.  (Andrea Morales)
Editor’s Note: This article is the second in a series about how the gig economy is shaping the future of labor and what that means for workers.

When Terrence Davenport first heard about the so-called gig economy, he was working at a free-meal program in his hometown of Dumas, Arkansas, a tiny village surrounded by cotton fields. Around 40 percent of Dumas’s roughly 5,000 residents lived in poverty. Most young people who left for college, as Davenport had done, never came back, and both the town’s population and its median salary—about $23,000 a year—were shrinking. “What did you eat today?” Davenport would ask kids he passed on the street. Often it wasn’t much, and he invited them to have a free meal. But what he really wanted to do was solve the deeper problems that made them hungry.

It was 2014, still the early days of Uber and Airbnb, and Silicon Valley was promoting the idea that its app-infused “gig economy”—which used digital technology to connect workers with projects—could solve the United States economy’s problems. “In many ways, we look at Uber as the safety net for a city,” then Uber-CEO Travis Kalanick said on a conference stage in 2016. He asked the audience to imagine that a factory had closed down. What would happen to those workers? “They can push a button and get to work.”

A San Francisco–non-profit called Samasource wanted to test the idea in Dumas. It already hired extremely poor people in East Africa and India to complete online projects for tech companies like Google. While the difference in living wages made that model impossible in the United States—the tech projects paid too little—Samasource hoped the gig economy could create similar opportunities for the unemployed here. It called the idea “Samaschool” and had chosen Dumas, along with Merced, California, to test the program; in Dumas, the local school district’s outgoing superintendent, who had sat in on Sunday-school classes taught by Davenport, had recommended him to run it.

As a child, Davenport had lived in Dumas with his grandmother and mother in an old sharecropper’s house, before attending some high school in nearby Pine Bluff, where his father lived. When he started classes at the University of Arkansas, he never thought he’d return to Dumas. But his little brother’s death, in 2012, brought him back. Now he was living with his grandmother, cooking her dinner.

Davenport had a habit of letting his phone go uncharged for days, but a hand-delivered message finally reached him at the food program, and he ran home, changed his shirt, and arrived in a sprint at the community-technology center on Dumas’s main street, still having only a vague idea of what the job entailed. As his interviewer explained Samaschool’s intentions—to teach residents how to get jobs in the gig economy—Davenport immediately saw it as a way to make a bigger difference. A few weeks later, Davenport got the job. Finally, he thought, he’d found a conduit to new work opportunities that had so far left his small town, and others like it, behind.

Terrence Davenport at the Arkansas Adult Learning Resource Center (Andrea Morales)

For a small town, attracting businesses, nurturing start-ups, and retraining workers for local jobs can be slow processes. In Dumas, Samaschool hoped it could sidestep these workforce-development challenges by connecting residents, through the internet, to gigs that had been created elsewhere. “The average student we work with is unemployed for 16 months prior to the program,” Samaschool’s then-managing director told me in 2015. “For them to take training that is months or years, they don’t have that time.”

Research, data entry, and customer service—all work that was plentiful on online freelancing websites—didn’t require college degrees or trade skills. All Dumas residents needed, the thinking went, was some instruction about self-promotion and digital literacy, and an internet connection. An 80-hour program, to teach all this, would span 10 weeks. Students would learn how to build an online portfolio of work, search for jobs, write proposals, and use the freelancing websites Elance and Odesk (which would later be combined into Upwork). Around 70 Dumas residents applied. Davenport chose 30 of them, including farm workers, home-care aides, and a few people who were chronically unemployed. The Dumas program quickly attracted attention both locally and nationally. But as he started teaching, Davenport knew almost instantly that the gig economy wasn’t going to provide the easy solution to unemployment that had been promised—at least, not in Dumas.

Before any of Davenport’s students could work in the gig economy, they had to apply and be selected for projects—which proved more challenging than expected. According to Samasource, 60 percent of the students who took Davenport’s class did not own a computer, and 44 percent did not have access to the internet, even on their phones. Some Dumas residents were so unfamiliar with computers that they didn’t know where to type the URL on a search browser, or how to send an email.

But familiarity with online work—and the inadequate internet speeds in Dumas—were far from the only hurdles to winning job contracts. Although 90 percent of Davenport’s students had high-school degrees, many couldn’t read or write at a high-school level. Davenport was trying to teach some students how to craft an online resume profile when they could barely write a grammatical sentence, and trying to coach them in deciphering online job postings when they struggled to read. Gigs on Upwork, Davenport learned, tended to fall into one of two categories. They either didn’t require any formal qualifications, which meant they could often be completed by workers in other countries who were willing to work for as little as $3 or $5 per hour, or they were specialized and well-remunerated, which meant they often went to workers with undergraduate or even master’s degrees.

Davenport’s students had another, deeper disadvantage. Most of his students faced bias against African-Americans. In a 2013 experiment, researchers tried to sell an iPod on online marketplaces, and found they received 13 percent fewer responses and 17 percent fewer offers when the photo showed the iPod being held by a black hand than when it was held by a white hand. In a more recent experiment, Airbnb users with distinctively African-American names were 16 percent less likely to be accepted as guests than identical users with white-sounding names. Most online work platforms show employers profile photos of applicants.

Beyond that, Davenport had begun to notice the many ways in which slavery had left a lingering mark on Dumas. That trauma had led to a feeling of hopelessness, which had been passed down in the culture through generations. On top of that, there was the added trauma of being systematically cheated out of land and voting rights, and of present-day racism. Davenport had trouble keeping students believing in themselves as they were rejected from gigs to which they’d applied—which was their primary experience.

A railroad runs right through Main Street Dumas. (Andrea Morales)

One student landed a customer service job at $7 per hour. A few tried working for $3 or $5 per hour. Someone found a $50 project, but it took so many hours to complete that the hourly pay became meaningless. Outside of Upwork, Davenport found some of his students jobs as contract workers at call centers. But in the end, just five of the 30 students he’d admitted to the first class found digital work. Gary Foster, a Dumas Samaschool graduate, said that during his several weeks of mandatory unpaid training for the gig, he had trouble paying his bills. “I had disconnection notices everywhere,” he told me. He was one of the five students to find work, making $9 per hour as a contracted customer-service representative (in a nesting doll of contingent employment that made him a contractor to a contractor to a contractor to Sears), but that lasted less than a year before Foster, finding that he often could not rely on getting enough hours to make ends meet, instead applied for a license to drive a semitruck. Later, he moved to Milwaukee, Wisconsin (where he had lived once before).

Hazel Jackson, who enrolled in Samaschool after having trouble finding a traditional job, enjoyed working on new skills, but never found a gig. Previously, she’d worked as a cashier and as an operator of a machine used for testing the grade of cotton. “The most important thing is that I learned something I would never learn,” she said. “Hopefully someday I’ll be able to use it.”

Not everyone’s experience in the gig economy was so fraught. Two workers who graduated from Samaschool pilot programs in California told me they’d used the gig economy to find sustainable work. And throughout the years I spent reporting on gig-economy workers, I met plenty whose work matched the optimistic Silicon Valley vision. Most often, these workers had specialized skills, savings accounts that helped them weather slow work times and take vacations, and resources to pay for expensive health-insurance plans. (In the gig economy, none of the labor protections or benefits reserved for full-time employees apply.) Davenport’s students typically did not have these luxuries. Almost half of them said they did not have reliable transportation, and around 15 percent of them experienced homelessness while taking the Samaschool course. As Davenport put it: “If you have a water bill that you can’t pay, you’re not going to focus on a class.” It wasn’t that the gig economy didn’t solve problems; it just didn’t solve Dumas’s problems.

At the end of Davenport’s first course, the feedback was consistent: This is a nice platform and all, but we’re spending a lot of time working on it without getting paid. Where are the jobs? So Davenport and Samaschool pivoted. They thought students would win more gigs if they had been trained for specific ones. So, for the second class cohort, Davenport taught students not only how to use Upwork and promote themselves, but also how to do work in areas like customer service, social-media marketing, and virtual assistance (helping clients manage their email, calendars, and errands, from afar).

Still, only two people found digital work. For the third cohort, Samaschool narrowed the curriculum to one skill: social-media marketing. It also tried partnering with businesses to offer paid social-media-marketing internships to Dumas residents, outside of the gig-economy sites. The field paid fairly well and favored native English speakers. But it also required constant creative thinking and perfect grammar. “We’ve been doing physical labor,” Davenport told me, “where you get a job and obey what your boss tells you.” Marketing didn’t jibe with most of his students’ past experiences, and the training wasn’t enough to overcome that; only one student found gig-economy work.

A car leaves a trail of dust behind it while driving in the fields near Dumas. (Andrea Morales)

Attendance dwindled, along with hope. One evening in 2015, I visited one of Davenport’s classes. It was supposed to start at 6 p.m., but when I arrived, 10 minutes late, the big classroom was empty, and I found Davenport leaning on a railing outside. “I don’t understand what is happening,” he told me. He had just hung up his cell phone after speaking with one of his students, who had a headache and wouldn’t be coming to class. Another student was watching her son’s football game. “This has never happened before,” he continued, his eyes wettening with frustration.

In the months that followed, Samaschool tried abandoning Upwork altogether and instead hiring Dumas residents itself to complete projects, similarly to how the original Samasource program hires workers in Kenya, Uganda, and India for virtual work. It tried recruiting students who had some college experience. But the problem didn’t seem to be a matter of iteration. After three years, only nine of the course’s 76 participants found gigs.

Late last year, I met Davenport again, at a McDonald’s in Pine Bluff, a city of about 44,000 people where he now lives. After the failure in Dumas, Samaschool had changed its strategy. Rather than providing stand-alone classes, it had begun partnering with existing training programs that prepare students for work in industries like IT, hospitality, and food service. The organization hoped students would use the gig economy to build their resumes and experience while they looked for more stable jobs. After all, while gig work alone might not be a quick fix to unemployment, it is growing faster than other segments of the workforce.

Hazel Jackson, a student in the program that was helping local people work in the gig economy, stands with her son Braylon. (Andrea Morales)

Samaschool had added a 10-hour training module to other organizations’ skills-focused courses. In it, the non-profit discusses how to use gig-economy platforms and how to plan for some of the shortcomings of the gig economy—that students should plan to save for a larger tax bill, plan for inconsistent work, look out for their own safety, and buy their own insurance.  Several months before my visit, Samasource had informed Davenport that it would be ending its program in Dumas; It wasn’t having the impact it intended. Davenport wasn’t all that surprised. “You can say, let’s go to Arkansas and get people jobs,” he said. “That’s a great ambition. It’s nice to talk about and write about. But then you start to think about what it takes.” As part of his training, he’d traveled to San Francisco, which he found to be a “place filled with the leaders of our world who are uninformed about it.” If Silicon Valley really wanted to make a difference in Dumas, he now felt, it should have studied the people first.

When it did, it would realize that the presence of the digital gig economy couldn’t alone halt a cycle of poverty that had begun with slavery. “You can’t change someone’s culture in two or three months,” he said. Davenport thought chronically unemployed people should have caseworkers who could help address the immediate needs, like transportation and childcare, that often prevented them from completing job-training courses or applying for work. It would require mentorship, caring for a whole person rather than only job skills. All of this, and other programs, would require at least ten times the financial resources that Samaschool had allocated to Dumas.

Davenport had also begun to consider how residents might make better use of the Arkansas Delta’s existing resources. He often thought about all of the land, and how “children grow up with food growing all around them, and they’re hungry.” Sometimes he thought about small things: At a meeting in which the city council decided to buy dumpsters, he thought about how, when he was looking for work in Dumas after his mother’s death, he’d dug up scrap metal to sell by the pound. Why not teach Dumas residents to weld and have them build the dumpsters themselves? Residents would learn a skill that is in demand nationwide and get paid to practice it, and the city might wind up with cheaper dumpsters as a result. Create enough of these types of opportunities by using resources differently, Davenport reasoned, and Dumas could create its own gig economy of sorts.

In Pine Bluff, Davenport had been putting together a final report on Samaschool’s Arkansas program, for its funders. As I listened to his ideas for alleviating poverty in his own community, they sounded nothing like the ones that had once been proposed by gig-economy entrepreneurs. Davenport’s ideas were multi-dimensional where Silicon Valley’s solutions were simple; they were individual rather than scalable; and, they would definitely not come on-demand.

As Davenport and I spoke, a boy in a white tank top and gym shorts ran, shoeless, past our booth. The kid, who looked about 10 years old, was one of Davenport’s neighbors. “Hey, how are you doing?” Davenport called. As the boy told us he was doing okay and bolted into the bathroom, a woman at another table yelled after him, “Where are your shoes?” The boy appeared again a few minutes later. “What did you eat today?” Davenport asked him. The child muttered that he could use a burger, and Davenport got up to order one.

This post is adapted from Kessler’s upcoming book, Gigged: The End of the Job and the Future of Work.