Abundance Has a Last-Mile Problem

The Supplemental Nutrition Assistance Program serves about 41 million people. Roughly five million more are eligible for it and don't receive it. Most of them are eligible last year, eligible this year, eligible next year — they qualify the way the building inspector qualifies for a paycheck, because the rules of the program say so. The benefit fails to reach them anyway. The number USDA reports is an 88% take-up rate, which sounds like a passing grade until you do the multiplication and notice that the missing 12% is larger than the population of Connecticut.

The Earned Income Tax Credit is the country's largest anti-poverty program. About one in five eligible filers leaves it on the table. The unclaimed amount runs around $7 billion a year, according to the Treasury Inspector General for Tax Administration. Property tax exemptions for seniors and disabled veterans exist in at least 39 states plus DC per the Lincoln Institute of Land Policy and almost never apply automatically — the household has to find the form, file it on time, and refile it when something changes, which is to say file it forever. Indigency affidavits in civil court let people on means-tested benefits skip filing fees they can't pay; the eligibility check is essentially "are you already receiving SNAP," which the court already knows how to verify, and which it requires you to prove on a separate form anyway.

These are the programs working as designed and that's the problem.

I bring this up because the discourse about whether America builds enough housing, enough transit, enough clean energy has been absorbing most of the oxygen in the room labeled "what's wrong with government." Ezra Klein and Derek Thompson's Abundance, published last year, is the tentpole text. The Roosevelt Institute's response report from Kate Andrias and Alexander Hertel-Fernandez, Democratic Abundance, came out this week and argues that Klein and Thompson underweight the role of organized labor in delivering the housing and infrastructure they want to see built. Jennifer Pahlka's Recoding America and the broader state-capacity wing she anchors sit adjacent to both, focused on internal agency reform. There is genuine intellectual work happening about why permitting takes nine years and why a subway station costs eight times what it costs in Stockholm. None of that work touches the question of why the woman who's been on SNAP for eleven years loses her benefits in February because the recertification letter went to her old address.

The abundance debate is about the production layer. The construction site, the procurement office, the zoning board, the labor agreement. The argument is whether deregulation or democratic participation produces more housing, more energy, more infrastructure. Both sides agree the central problem is the gap between what gets proposed and what gets built. Both sides have a theory of why that gap exists.

Neither side has a theory of the layer where the program already exists, the entitlement is already statutory, the building is already built, the bus is already running — and the household still doesn't get the thing.

This is where most of the lived experience of government actually happens. People interact with the production layer once a decade, when a new transit line opens or doesn't, when a housing development goes up or doesn't. They interact with the delivery layer every month: the renewal envelope, the eligibility screener, the document upload that times out, the call to a number where nobody picks up, the form that asks for information the agency already has on file in a different building. The abundance debate has produced approximately zero pages on this. The Roosevelt response has produced approximately zero pages on this. The state-capacity wing gets closest, and most of its work is still about the inside of the agency — the procurement and personnel and software — rather than the experience of the household receiving the service.

Procedural complexity at the delivery layer is itself a redistributive mechanism. It taxes time, attention, literacy, transportation, technology access, English fluency, work schedule flexibility, and the ability to take a Tuesday afternoon off to sit in a county office. The households that pay this tax most heavily are the households the program was designed to serve. The households that pay it least heavily are the ones who never needed the program. A SNAP recertification gap of five million people functions as a regressive transfer from poor households to the state's general fund, mediated by the cost of paperwork.

You can describe this in the language of administrative burden, which is what Pamela Herd and Don Moynihan have spent fifteen years doing in their book Administrative Burden and the work that followed it. You can describe it in the language of behavioral economics, which produced the SNAP recertification text-message studies showing a single reminder bumps renewal rates by five percentage points. You can describe it in the language of civil rights, which is the frame the indigency-affidavit advocates have used for forty years. You can describe it as a power problem, in which case the Roosevelt authors should be very interested, because the household sitting on hold with the benefits office has approximately the same bargaining power against the state as a non-union worker has against an employer, which is to say none.

You can also describe it the way I do, which is that the entitlement layer is where the architecture of a public service either honors its own promises or doesn't, and most of the time it doesn't, and this is fixable.

The fix lives below the level of policy reform. The agency already has the data, already has the rules, already has the authority, already has the obligation. What's missing is the routine procedural step that uses what's already there to produce the outcome the law already says should happen. The senior who qualifies for the property tax exemption based on age, income, and length of ownership — the assessor's office has all three of those data points in its own database. The SNAP household that qualifies for renewal based on the same income, household composition, and address as last year — the state's eligibility system already has all three. The court fee waiver for someone receiving Medicaid — the court can verify Medicaid status against the state Medicaid file in about four seconds.

These are the procedural equivalent of a building that has plumbing on every floor and a faucet in every unit, where the water company declines to turn the valve unless the tenant fills out a form requesting water, refiled annually, in person, during business hours.

Call this whatever you want. Administrative burden, second-chance government, latent service delivery, the paperwork tax. The label matters less than the recognition. There is a real and large category of public service failure that sits below the production debate, and most of it can be fixed with a cron job, an eligibility rules engine, and a mail merge.

Whether the abundance frame stretches to include the entitlement layer is a question for the people who own the abundance frame. The five million people will continue to not receive the benefits they qualify for while the discourse decides what to call this.