For Poor Consumers, Poor Technology

This week, we’ve seen a slew of pieces documenting technological barriers to accessing the safety net—and keeping specific benefits secure. As these pieces illustrate, technological glitches are widespread and have been a fixture of the public assistance system for decades.  Why do we see such pervasive dysfunction – and is there anything we can do about it?

A quick survey of the coverage:

• Over the weekend, an op-ed in the New York Times by Georgetown Law professor David Super described the “litany of automation and contracting meltdowns” that have plagued states’ SNAP and Medicaid systems, automatically terminating benefits for hundreds of thousands of families – well before the HealthCare.gov debacle got the media’s attention. Ezra Klein followed up with a nice interview in which Super described how the ACA website’s recovery has been “astonishingly fast” compared to broken state-level systems serving poor families.
• Similarly, in Tuesday’s New York Times, Frances Roble documented how efforts to modernize state systems administering unemployment compensation have met with technical problems across the country – resulting in tens of thousands going without their checks.
• Last but not least, last Friday, a CBS affiliate in Virginia exposed how the Target data breach also impacted nearly 4000 users of the state’s EPPICard, a MasterCard-branded prepaid debit card used to distribute Virginia’s TANF assistance and child support payments. Cards were blocked as of January 1^st, leaving families with no access to these funds – though a press release to notify affected families didn’t emerge until two days later.  

Each of these examples reflects how the digitization of our safety net can have unintended negative consequences—and how these systems themselves are often a reflection of the values and resources behind them. As we explored last month in an event with our colleagues from the Open Technology Institute, while shifting to a safety net that is computerized and automated can help streamline access to needed services, without appropriate safeguards, it can simultaneously pose significant risks to privacy and trigger mass eligibility errors and inequitable outcomes.

What’s more, policymakers’ tolerance of these systems’ deficiencies more broadly reflects a lower standard of care and diminished responsiveness to families who rely on public assistance, and its incumbent technology, to purchase basic goods and services. This may be understood as yet another manifestation of “two-tier welfare.” The poorer you have to be to access a particular program, the more you’re expected to submit to humiliating application rituals, ongoing suspicion, and potentially unreliable access to the assistance you qualify for.

For example, since we’re talking about the EPPICard: as we’ve noted before, state-issued cards used to distribute means-tested assistance like TANF, including EBT cards, are explicitly exempt from the federal statute that protects consumers against loss or theft and guarantees provisional use of stolen funds pending resolution of their claim. Why? In a word, distrust. When Congress was debating this issue in the late 1990s, states argued that extending these basic consumer protections to EBT users would encourage fraud, despite a pilot project sponsored by the USDA that proved otherwise. The default assumption was that low-income people were likely to try to “game” the system, notwithstanding evidence to the contrary—and thus the policy remained in place. Today, states that want to extend these same protections to families accessing the safety net have to take proactive steps to do so, as California recently did. The consequences of this so-called “EBT exception” are that families who lose their (already meager) benefits through no fault of their own may face devastating delays in getting those funds replaced – if they are replaced at all.

For a less wonky example, look to the EBT outage that affected families in 17 states in October. Though those states’ EBT contractor, Xerox, clearly acknowledged its responsibility for the outage, the media coverage instead fixated on isolated examples of shoppers who “took advantage” of the situation, when two Louisiana Wal-Mart stores enabled unlimited purchases as a result of the glitch. Following the episode, Louisiana even announced it would seek to prosecute shoppers who left a store with groceries exceeding their EBT balance. Rather than holding the system accountable for failing to ensure millions of families could put food on the table, the prevailing response was to blame and demonize the individuals harmed by the failure, and frame the whole incident as a confirmation of stereotypes.

Moreover, a more routine version of the October outage occurs on a regular basis when a given store’s EBT system is down. A friend of mine, who has accessed SNAP for the first time in the wake of a lay-off, encountered this precise situation last week at a grocery store in California. In an email, he described for me how his cashier “stated rather loudly, ‘We don't have EBT,’” before he even reached the register, and proceeded to loudly inform the confused customer behind him in line that “EBT is down, so I can't ring him up." With no other options, my friend left, humiliated and empty-handed.

Blaming “technology” for these widespread and persistent failings of our public assistance system is a cop-out. The technology behind these systems’ administration is a reflection of the resources and priorities of the governments that create them.  If the people accessing these programs had more political power, we might see things fixed a little faster, or even see more stringent data security safeguards built into platforms and greater legal recourse available in instances of data breaches. In the meantime, it’s refreshing to see some high-profile media attention to what we need to do better.