Muncie, Indiana, known since the 1920s as the prototypical American “Middletown,” recently found itself at ground zero of an experiment in high-tech governance. It would be easy to see Muncie—after a rash of plant closings, a foreclosure crisis, and skyrocketing poverty throughout the first decade of the 21st century—as a ghost of America’s industrial past. In fact, it is a canary in the coal mine of our future.

In November 2006, Indiana Governor Mitch Daniels announced a 10-year, $1.16 billion contract with a consortium of tech companies, led by IBM and Affiliated Computer Services (ACS), to modernize and privatize eligibility procedures for the state’s Medicaid, food-stamp, and cash-assistance programs. Central to the plan was “remote eligibility”: Rather than applying in person at county welfare offices, Hoosiers would submit online applications, fax in supporting documents, and then be interviewed by private call-center employees.

Daniels, who was considered at the time a likely presidential contender, pitched remote eligibility as more than just a technical fix. Throughout the process, he repeatedly called Indiana’s welfare system “broken, wasteful, and often fraudulent” and the “worst in America.” It was true that Indiana lagged behind other states in reducing its welfare rolls. In the 10 years after so-called welfare reform passed in 1996, Indiana’s caseloads dropped only 6 percent, compared with 58 percent in the country overall. IBM’s system promised to accomplish what legislation had not yet achieved.

The modernization project arrived in Muncie, the largest city in the first pilot area, a year after its launch. System failures were immediate and widespread. Applicants waited 20 or 30 minutes on hold, only to be denied benefits for “failure to cooperate in establishing eligibility” if they were unable to receive a callback after having burned through their limited cellphone minutes. They faxed millions of pages of photocopied driver’s licenses, Social Security cards, and other supporting documents to a processing center in Marion, 40 miles away; so many of the documents disappeared that advocates started calling it “the black hole in Marion.” If one page among dozens was too dark to read, damaged in transmission, or incorrectly indexed to a case, the application was denied.

Kay Walker, a Center Township trustee, realized that these changes were devastating Muncie families. “They were getting kicked off assistance,” she said in March 2015. “They were confused, and they didn’t know where to turn. There was no case management, no personal connection, no communication among agencies. It was just the biggest mess.”

By February 2008, the number of households receiving food stamps in Delaware County, which includes Muncie, dropped more than 7 percent, though requests for food assistance had climbed 4 percent in Indiana overall. Calls to the LifeStream 211 telephone hotline requesting information about food pantries doubled. The Second Harvest Food Bank faced severe shortages. Hospitals, nursing homes, and pharmacies failed to receive Medicaid reimbursements. The municipal graveyard complained that it had not been paid thousands of dollars for the funerals of poor and indigent people.

* * *

In response, Walker and her staff organized a People’s Town Hall Meeting in May 2008. Dozens of public-assistance recipients testified before a packed room of 500 people. Melinda Jones of Muncie, the mother of a 10-month-old with cancer, was fighting to keep her Medicaid and food stamps. “I have to beg and borrow from my family to give my daughter her food,” she said, “and I think it’s utterly ridiculous that we do our children like this.”

Christina King, a diabetic and working mother of three, lost her Medicaid during the modernization. She was unable to afford insulin for seven months, and her blood sugar was out of control, putting her at risk of stroke or coma. “What good does it do when my 7-year-old walks in and I physically cannot get out of bed?” she said. “I spent two days in the ICU because I have no medicine. My kidneys are now at risk. My eyes are at risk. But I get up every day and I go to work, because I think it’s important for me to show my kids, ‘Don’t be dependent on the system. Don’t be like me—do better.’”

Deaf, blind, disabled, and mentally ill clients were particularly hard-hit. Modernization materials were not translated into Braille. Applicants in wheelchairs struggled to answer incoming calls from Marion, only to have the caller disconnect before they could reach the phone. “I’m deaf. How can I do a telephone interview?” asked Dionna McGairk through a sign-language interpreter at the town-hall meeting. “I tell [call-center operators] to use my relay service. They don’t understand what relay service is.” She says the operators told her she needed to get help. “I am saying, ‘No—I can answer my questions myself. You are discriminating against the deaf.’”

Ignoring these initial troubles, Indiana rolled the pilot program out to 47 additional counties over the course of that year. By 2009, more than 1 million state residents had lost or been denied access to food stamps, Medicaid, and cash benefits—a 54 percent increase in denials compared with the previous three years—despite the worsening recession, relaxed eligibility rules for federal food stamps, and catastrophic flooding in the region.

Then Governor Daniels did something unexpected: He admitted that the experiment was a failure and canceled the state’s contract with IBM, calling the modernization plan a “flawed concept that simply did not work out in practice.” IBM asserted that unexpected economic, political, and natural events had doomed its efforts to streamline the eligibility process, and sued the state for breach of contract.

The governor and IBM both blamed forces out of their control for the plan’s collapse. But it was easy to predict that Indiana’s modernization effort would fail. An almost identical system had been tried twice before—in Florida in 2004 and Texas in 2005—with similar results.

The Florida project replaced caseworkers with online questionnaires, touting the resulting freedom from in-person eligibility interviews. Because many low-income residents lacked an Internet connection in their homes, Florida asked a network of more than 3,000 community partners to provide computer access in their offices; as a result, the state reduced its welfare staff by 40 percent. But according to a 2008 report in Clearinghouse Review, a journal covering poverty law and policy, these organizations received little to no funding, training, or oversight in taking over the traditional casework responsibilities of state employees. There were no provisions for dealing with barriers to access due to disability, limited literacy, or limited English proficiency. Many applications were subsequently denied for failure to cooperate, usually as a result of problems with scheduling telephone interviews and faxing documents.

Likewise, the Texas Health and Human Services Commission signed an $899 million contract with Accenture to build a system that computerized eligibility decisions for the state’s Children’s Health Insurance Program (CHIP), among other programs. Accenture incorrectly terminated 6,000 kids’ Medicaid cases, including a 14-year-old Houston boy who died of kidney cancer as a result. After quickly spending $244 million in taxpayer dollars, Texas canceled the two-county pilot program in 2007 and terminated Accenture’s contract.

Texas Comptroller Carole Keeton Strayhorn called her state’s remote-eligibility efforts a “perfect storm of wasted dollars, reduced access to services…and profiteering at the expense of our Texas taxpayers.” Judge David Dreyer of the Marion Superior Court used strikingly similar language in his decision in State of Indiana v. IBM. “Neither party deserves to win this case,” he wrote. “This story represents a ‘perfect storm’ of misguided government policy and overzealous corporate ambition. Overall, both parties are to blame and Indiana’s taxpayers are left as apparent losers.” Even so, he found the state in breach of contract and awarded $52 million in damages to IBM.

* * *

In the wake of the stunning collapse of these three programs, a bigger question remains unanswered: Can the failures of remote eligibility really be blamed on a “perfect storm” of bureaucratic incompetence, poor contractor performance, and unexpected technical glitches? Or was “welfare modernization” a high-tech Trojan horse concealing a very old-fashioned effort to cut public benefits and lay off public workers? Storms are unpredictable—random acts of nature. Remote eligibility failed for entirely predictable human reasons. Here are three big ones:

1. Casework is crucial. Remote eligibility profoundly changes the relationship between public workers and poor and working people. No longer are caseworkers responsible for families; instead, they’re responsible for tasks assigned through a computerized queue. Any available worker can complete the next task, but no single worker ever sees the whole picture of an applicant’s life.

“It’s all piecework now. We became slaves to the task system,” says Fred Gilbert, a retired caseworker who served and supported refugee families in Fort Wayne, Indiana, for more than 30 years. “Like any other private call center, it’s ‘Just the facts.’ But the welfare system is very complicated. That’s the job of caseworkers: to help people wade through the mess.”

Casework has never been simple. It’s a complex human endeavor that relies on relationships, requires a difficult mix of canniness and compassion, and is vulnerable to all the biases about race, class, and gender that are woven throughout our society. And caseworker discretion can cause real trouble: Until the national welfare-rights movement of the 1960s won equal access to welfare programs, for example, white caseworkers used subjective “man in the house” and “suitable home” rules to exclude African-American families from public services. But after the modernization in Indiana, eligibility workers were no longer authorized to take any independent action to help clients get verified for benefits.

2. Volunteers cannot replace government employees. Remote eligibility assumes that low-income families have ready access to the Internet. But even today, according to the American Community Survey, only half of those households making less than $25,000 a year have Internet service at home. The majority of applicants in Indiana relied on a community partner, such as a local library, nursing home, or health clinic, to access electronic applications and learn how to use remote-eligibility programs.

“We had lines of desperate people waiting for help,” says Muncie Public Library director Ginny Nilles. As in Florida, community partners in Indiana received little to no compensation, training, or oversight to do what amounted to volunteer casework. Library staffers worked hard but were quickly overwhelmed, and the situation worsened when the state passed a budget that cut library funding by $1 million. Librarians were being laid off, so they trained community volunteers to help patrons submit applications, which created serious challenges.

“The forms ask very personal questions,” Nilles explains. “If they couldn’t use the computer, it was incumbent on us to read the questions out loud and get the answers: Social Security numbers, mental and physical health. Volunteers are great, but if you pay someone to do a job, it’s their responsibility. It’s about accountability.”

3. The real problem with welfare is lack of access. Just as too much voting isn’t the key problem with American politics, too much access isn’t the key problem with our welfare state. Only three-quarters of eligible people across the country claim Supplemental Nutrition Assistance Program (SNAP) benefits, even under today’s relaxed eligibility requirements. A 2010 Robert Wood Johnson Foundation study showed that, before the mandates of the Affordable Care Act, only 70 percent of Americans eligible for Medicaid were signed up. The problem, in short, is not that too many people are fraudulently claiming welfare; it’s that so few people who qualify for public benefits manage to get them, keep them, and use them.

Before its modernization experiment, Indiana ironically fared quite well in getting benefits to those who were eligible. It was 16th among states in food-stamp participation rates, for example, providing benefits to 74 percent of the people eligible. Indiana also had relatively low fraud and error rates: Only 5.3 percent of food-stamp dollars were spent incorrectly in 2005, according to the state’s own statistics.

Florida, Texas, and Indiana all described their modernized systems as paperless, convenient, and cost-effective. Underlying the rhetoric about flexible-access options and administrative efficiencies, however, was a strong focus on eliminating fraud by taking a zero-tolerance approach to application errors. In the long run, the modernized system shifted the administrative burden of documenting welfare eligibility—as well as the blame for errors—from caseworkers to the applicants themselves, the majority of whom are the disabled, the elderly, and children.

* * *

The State ofIindiana and IBM went into mediation over their contract dispute in February. But the legal case only seeks to apportion blame and levy penalties. We still haven’t answered that bigger question: Was this a system failure? Or was the widespread loss of benefits a predictable—even intentional—outcome?

“The system doesn’t seem to be set up to help people; it seems to be set up to play gotcha,” said Christopher Holly, a Medicaid attorney in Bloomington, in a December 2014 interview. “In our legal system, it is better that 10 guilty men go free than one innocent man go to jail. The modernization flipped that on its head.” In other words, the new system was based on the assumption that it is better for 10 eligible applicants to be denied public benefits than for one ineligible person to receive them.

The design of electronic-governance systems affects our material well-being, the health of our democracy, and equity in our communities. But somehow, when we talk about data-driven government, we conveniently omit the often terrible impacts that these systems have on the poor and working-class people in places like Muncie. But despite their dismal record of failure, new remote-eligibility projects continue to crop up. Xerox acquired Affiliated Computer Services in 2010; rebranded as Xerox Services, the company developed an “integrated eligibility solution” that controls access to benefits and services for tens of millions of people in 37 states.

These systems touch all of our lives, even if they remain invisible to many. According to Mark R. Rank’s One Nation, Underprivileged: Why American Poverty Affects Us All, the great majority of Americans—65 percent—will use means-tested public assistance at some point. Even for those lucky enough never to need help, the cost of these systems to taxpayers far outstrips the fraud they supposedly prevent. Indiana has already lost more than $500 million—and counting—on its remote-eligibility gamble.

But there’s also a larger toll. “There’s a cost to people,” says Jamie Andree of Indiana Legal Services. “A case might ultimately be resolved in the applicant’s favor—after a year! But the cost of just waiting around without Medicaid benefits is enormous; it’s really hard to make somebody whole. Most people will stop getting medical care while eligibility is being determined. There’s no way to compensate them for that.” Also in This Issue

Leah Hunt-Hendrix and Astra Taylor, “‘Tech’ Is Political—How We Respond to It Needs to Be Just as Political”

Tim Shorrock, “How Private Contractors Have Created a Shadow NSA ”

Eleanor Saitta, “The Key to Ending Mass Surveillance? Math.”

Ingrid Burrington, “What Amazon Taught the Cops”

Astra Taylor and Jathan Sadowski, “How Companies Turn Your Facebook Activity Into a Credit Score ”

Jessica Bruder, “These Workers Have a New Demand: Stop Watching Us ”