The Benefits Apocalypse

The Emergency

Most emergencies announce themselves. A siren, a headline, a number that spikes overnight. As of today, June 30th, 2026, the sirens are blaring, the headlines are catching up, the leading indicators are spiking — we are in an emergency. I see an American century that will be plagued by the unintended and intended consequences of A.I. integration and automation.

A few months ago, new college graduates voiced their frustrations as tech CEOs gushed over the virtues of the coming artificial intelligence revolution. They were not booing because they disagreed with the thesis of the respective speaker; rather, their dissatisfaction stemmed from the reality that this revolution has already upended their entire life — every theory they had learned, the practical value of the degree they had spent two decades earning, and their real employment prospects for the next century have been changed by this tool. And like a canary, these students booed, screaming from under a lid being tightly closed on their future.

The layoffs and automation that have always existed in the tech industry have been swiftly making their way through the entire workforce ecosystem. Employers are cutting back core and fringe benefits. The unemployment rate of 4.5% holds steady, yet fewer workers are seeking full-time employment, more people are working two or more jobs, fewer men are in the workforce as a whole, real wages are under the rate of inflation, and the energy crisis caused by the conflict in the Middle East is just now working its way through our economy. This says nothing of the impact on city and state benefits systems, on nonprofit local supports, and nothing of the real calamity that will face a large majority of American households.

Here, I want to describe this emergency honestly and be specific about the policy machinery that is making it worse. And then I want to tell you why, despite all of it, I am genuinely hopeful — because the way out is something we already know how to do. We just have to decide to do it together.

A System Not Built for a Growing Majority — ALICE

At Zion we spend a great deal of time modeling what the next ten years could look like for the families most exposed to the A.I. revolution. Researchers have a name for the households most exposed. They call them ALICE — Asset-Limited, Income-Constrained, Employed (United Way) — what we traditionally called the middle class. Unlike a topline U-3 unemployment rate or the general poverty level, ALICE includes those who work, go to school, have a social life, and are now more likely to need two incomes to get by. Even still, an ALICE household typically earns too much to receive many of the social benefits and government supports (with an exception for very high cost-of-living regions) but still cannot reliably cover the basics each month where they actually live.

By the most careful counts, this is not a small group at the margins. It is something close to two in five households (42%) in the country. It is not that earnings fall short in a few struggling regions. It is that, in essentially every region measured, the typical wage no longer covers the real local cost of a stable life. The official poverty line stopped describing the actual edge a long time ago. ALICE families live in the wide, under-counted space between that outdated line and the real cost of getting by.

ALICE families face two real emergencies. The first is that need is rising. Prices for the essentials — housing, food, childcare — have been climbing faster than wages for long enough that the gap is now structural, not a blip. At the same time, new technology keeps quietly removing the kind of steady, entry-level work that ALICE families have always used as a first rung or a fallback. The second is the one that does not make the news, so let me be concrete. The system meant to meet that rising need is becoming harder to use, not easier — and several specific policy shifts are driving that. The pipes that were built to connect a household to benefits or services are increasingly narrow, leaky, and disconnected from one another. The water is there. It just does not reliably reach the people under the tap.

These are the families the safety net’s plumbing fails most often, and for a reason worth saying plainly: the system was not built around them. It was built for a sharper, simpler picture of need — poor or not poor, in or out. ALICE families are neither. They move in and out of eligibility as hours change and seasons change. They qualify for one program and not the next. They sit precisely on the seams of the patchwork. And the seams are exactly where it tears.

Patchwork Systems: Benefits, Food, Housing, and Well-being

Work requirements are expanding. To keep basic help like food assistance, more people now must document, on a schedule, that they worked eighty hours every month, with the right paperwork, through every disruption life reliably produces. A requirement like that is easy to satisfy if your work is salaried and stable. It is a recurring hazard if your hours are set by a manager’s discretion, if your income is inconsistent because of gig work or creator work, if you face periods of disability, or in the many other scenarios that may prevent a person from meeting the requirement.

At the same time, recent federal law has begun shifting more of the cost of programs like SNAP onto individual states, with penalties tied to administrative error rates. This sounds like accounting. It is not. When a state suddenly owns a larger share of the cost and is penalized for mistakes, it does the rational thing: it tightens. It adds verification steps. It shortens deadlines. It asks for more proof, more often. Every one of those tightenings is sensible in a budget office, but when placed in action is a trap.

Which brings me to the single most important word — and one most people outside these systems have never had to learn: recertification. Benefits are not granted once. They must be re-proven, on a clock — every six or twelve months, depending on the program and the state. Recertification is where the largest share of families lose their coverage. Not because their lives improved. Because a renewal interview landed during a work shift, or a verification document expired, or the letter went to an address they had already left. As windows get shorter and the required steps multiply, recertification stops being a formality and becomes a recurring trap.

Another particular cruelty woven into the patchwork is called the benefits cliff. Earn a little more — pick up extra hours, accept a small raise — and you can lose more in benefits than you gained in pay. ALICE families familiar with the cliffs navigate them with a care that would impress any accountant; those entering this patchwork for the first time, however — students, the long-term unemployed, forced early retirees — may come to learn the perils of recertification, and that the system can punish the exact upward effort it claims to support.

Put all of that together. More families need help. The help is delivered through a patchwork that is becoming more complex, more conditional, and more failure-prone at exactly the moment more people depend on it. The brittle plumbing is not being repaired; it is being asked to carry more water through narrower pipes.

The Worst Case: When the Plumbing Clogs the Patchwork

At Zion we model what would happen to these ALICE families if nothing were done to solve this emergency. This worst-case scenario was not modeled to frighten anyone or to establish some hyperbolic call to action. Rather, looking at the trends already present in our systems, we wanted to see what would actually happen when the patchwork of systems failed.

The worst case is what happens when several pressures land in the same few years instead of politely taking turns. A faster wave of job automation. An energy shock that pushes the cost of nearly everything upward. A stretch where credit tightens and public budgets are squeezed hard at the very moment need is surging. Each of those is survivable on its own. Arriving together, they change the picture for ALICE families.

In the worst case, the administrative complexity that merely slows families down becomes a hard clog on benefits deployment. States under real fiscal stress do what stressed institutions do: they cut administrative capacity — fewer caseworkers, fewer open offices, longer queues — worsening at precisely the moment demand peaks. Wait times go from weeks to seasons. And the people who most need help are exactly the ones who can least afford to wait. The recertification backlog would no longer be a delay but a wall, preventing families from accessing available federal, state, and local resources. A safety net built for a steady trickle of need, suddenly asked to absorb a flood through pipes that were already narrowing, does not stretch gracefully. It backs up. It overflows.

When people reach for help and the help cannot reach back, something deeper than a benefit is lost. Trust in our systems and institutions is lost. People stop believing the system will be there, and they stop trying. A safety net that people no longer trust enough to use is not really a safety net at all — more akin to a restaurant with no customers. Doomed to the same fate.

Unclogging the Pipes

You can pass the most generous benefits ever written, and if a family cannot prove who they are, cannot carry their own information from one office to the next, cannot make it through recertification without starting from zero — the generosity will never reach them. The plumbing is prone to clogging and ripe with benefits cliffs not because no one cares, but because no one built the connective tissue.

There is, today, no data layer that belongs to the household. No record that travels with a family instead of being scattered in fragments across a dozen institutions that do not talk to each other. No simple, trustworthy way for a person to hold their own information and decide who gets to see it, and for how long.

We have built astonishing digital infrastructure in this country. We can move money in milliseconds. We stream entertainment to any screen. We have even made targeted advertising ubiquitous, with eerie precision. Yet we have not built a nationwide digital infrastructure for the plain dignity of proving you are who you say you are, and that you need what you say you need — once, in a way that travels with you. That gap is the problem.

Solving this problem requires starting with what it actually takes for a working family to hold onto the benefits and social services available to them. Food assistance. Health coverage. Childcare support. Local assistance with keeping the heat and electricity on in winter. These are not luxuries. For tens of millions of households, they are the thin margin between getting by and not.

However, the data, the recertification, and the touchpoints that trail with and determine eligibility for each of these supports all live somewhere different. A different office, a different form, a different website and login, a different caseworker, a different definition of who you are and what counts as proof. A family does not experience something called “the safety net.” They experience nine front desks, nine waiting rooms, nine separate chances for a single piece of paper to go missing.

And here is the part that should bother all of us. Most families who lose a benefit do not lose it because they stopped being eligible. They lose it because a renewal notice went to a previous apartment. Because a pay stub would not upload. Because a deadline closed while they were working a double shift and caring for a parent. The system does not really ask, “Does this family still need help?” It asks, “Did this family complete the paperwork, correctly, on time?” Those are not the same question. The space between them is exactly where people fall through.

Why I Am Hopeful Anyway

If we — and I mean all of us, not one company — do not build that missing layer, we may very well see that worst case come to fruition. The missing household-controlled data layer is not a moonshot. It does not wait on some breakthrough that has not happened yet. The technical standards exist. The technology exists. What has been missing is the will to build the connective tissue — and the trust to hold it once it is built. Those are human problems, not impossible ones.

This is where I become genuinely hopeful, because the real answer to this problem is not a product. It is a community. For that reason, I do not believe the worst case is the most likely case. At Zion, we understand that the safety net, for all its broken plumbing, already runs through the most trusted institutions we have. It runs through congregations and public libraries. Through food pantries and community health centers. Through the recovery group that meets on Tuesday nights and the workforce office down the street. When a family is in trouble, they rarely call a state agency first. They walk into a place, and they talk to a person, and that person already knows their name. That web of trust is not broken. Quietly, it is the strongest thing we have.

The opportunity in front of us is to give that web the infrastructure it deserves — to build the missing layer not as something done to communities, but as something built with them. Something truly anchored in the libraries and congregations and clinics that people already trust, and owned in a real and lasting way by the people whose lives it describes. The problem is shared. The answer can be shared too.

We started Zion to build that layer. But I want to end on the truest thing I know about this work: Zion cannot do it alone, and it should not. A records layer for the people the existing systems were never designed to serve only works if those people — and the institutions that serve them — help build it, help govern it, and can trust it because they hold a real piece of it. The infrastructure is technical. The trust is communal. You cannot ship the second one in a software release. You can only earn it, in a place, with community buy-in and with that community holding us accountable.

So consider this less a warning than an invitation. The worst case is certainly not a destiny. It is simply what happens if we keep looking away from an emergency because we have decided not to hear the alarm.

We do not have to look away. We can build the missing layer, together, starting now — in each of our own communities, with the institutions already doing the work, for the families who have waited long enough. That is the real choice in front of us. I think we can make the right one.

— Dimitri, Founder & CEO, Zion Datum Public Benefit Corporation

In the next post, I write about what we are actually building, and how it is designed to answer exactly this.