Most American communities aren't program-poor. They're connection-poor.

Consider what happens when someone walks into a community-services intake today.

The worker takes the person's name, their household size, their address. They learn that the family has been without reliable housing for three months. They learn there's a child with an untreated asthma condition. They learn the primary earner lost a job when a bus line to their employer was rerouted.

The organization does what it does — food, or housing counseling, or workforce support. Well, and with dignity. The worker hands the person a paper list of other organizations for the rest of what they need.

The family leaves. They may visit two of the five places on the list. Or three. Or one. They will tell their story again at each one. Some of those conversations may lead to help. Others will end in "we don't have capacity this month," or "you need a different referral first," or silence.

None of this is a failure of effort. Every organization involved is doing meaningful work. But the system, taken as a whole, is not a system. It's a collection of programs that happen to be in the same city.

Section IProgram-rich, outcomes-poor

A mid-size American city typically hosts dozens to hundreds of nonprofits, public agencies, clinics, and faith-based organizations providing services across the social determinants of health. Durham, North Carolina has seventy-plus organizations in the ReCity Network alone. Monroe County, New York has been coordinating a large community-services network for years. Anchorage, Alaska; Detroit, Michigan; Richmond, Virginia — the numbers look similar in every mid-size city where we've worked.

What also looks similar: outcome disparities that aren't closing at the rate the program investment would predict. Food insecurity, eviction filings, maternal health gaps, post-incarceration recidivism, workforce retention. Per-capita investment up; measurable outcomes flat or down.

We call this being program-rich and outcomes-poor: the condition of having substantial service capacity that produces far fewer outcomes than its inputs would predict, because the capacity is not coordinated.

This is not hypothetical. It's visible in the financial reports of human-services nonprofits nationwide: per-participant cost climbing year over year, outcome metrics flat or declining, staff burnout rising, and the administrative overhead of every organization separately screening, intaking, assessing, and tracking its own participants growing roughly linearly with the number of organizations in the community — while outcomes grow more slowly.

Section IIWhy more programs don't solve it

There's a standing temptation, when outcomes fall short, to fund another program. The instinct comes from a real place: programs are visible, fundable, measurable at the program level, and they deliver immediate value to the people they touch.

But adding programs to a coordination-starved system produces diminishing returns. A new food pantry in a city with ten food pantries doesn't triple the number of families fed; it often just rearranges which families go where. A new workforce training program without connection to transportation, childcare, and basic-needs support produces completion rates that disappoint its funders. And the person navigating the system — a parent trying to hold their family together, a person returning from incarceration, a young worker transitioning between jobs — still has to tell their story five times and hope something connects.

That's what a Community Information Exchange is. Not another program. Not a replacement for what communities already have. Infrastructure — connective infrastructure — that lets the programs a community already funds function as a coordinated network rather than as disconnected points of service.

Section IIIWhat a Community Information Exchange is

A Community Information Exchange, or CIE, is software that enables a network of community-serving organizations to coordinate care for the people they share in common. At its simplest, a CIE does four things.

It maintains a shared directory of services — accurate, consent-based, up-to-date. More than most communities have today.

It tracks referrals as closed loops, not open-ended handoffs. When a food pantry refers a family to a housing program, the CIE registers that the referral was made, that the program received it, that the family was contacted, that services were (or weren't) delivered, and what the outcome was. "Referral sent" is the beginning of the loop. "Outcome documented" is the close.

It holds a person-centered record that, under the person's explicit consent, travels with them across participating organizations. The family doesn't tell its story five times. It tells it once, to the providers it chooses, and the people in its care team can see enough of that story to do their part well.

It measures outcomes at the network level. Not "referrals sent." Not "services attempted." Outcomes — did the family achieve housing stability, did the child's condition resolve, did the earner return to employment. These measurements are what funders, agencies, and the community itself need to see.

Critically, a CIE isn't a replacement for any of the existing systems. The food pantry keeps its client management software. The hospital keeps its EHR. The workforce nonprofit keeps its Salesforce. The CIE sits above and between them, as the coordination layer. It's also not a directory. Directories tell you what exists. A CIE helps something actually get delivered.

Section IVThe evidence

Community Information Exchanges aren't new. California, Arizona, New York, Michigan, and others have deployed them with varying degrees of success. What we know from that body of experience is that two factors predict whether a CIE will produce outcomes: the strength of community governance, and the experience of the implementation team. Technology, surprisingly, is rarely the limiting factor — modern platforms are capable enough that the software is often the least of it.

The clearest existing evidence comes from Monroe County, New York, where TogetherNow has operated one of the most established Community Information Exchanges in the country, built on IBM's Connect360 platform, for years.

Approximately seventy percent of referrals reach a documented outcome for the referred person. Referral sent, referral accepted, person contacted, service delivered, outcome recorded. For context: the industry baseline for closed-loop referral completion in community-services networks is typically cited at three to five percent.

This is not a modest improvement. It's an order-of-magnitude-plus difference, sustained at scale — across dozens of participating organizations, thousands of referral events per year, and a diverse set of service types from food access to housing support to chronic disease management to workforce re-entry.

What explains the gap? Not the platform alone. Other communities have used equivalent or similar platforms and achieved numbers much closer to the industry baseline. What makes the difference is operating discipline — the playbook for provider onboarding, the cadence of outcome review, the governance structures that keep participating organizations engaged, and the implementation choices that minimize friction for front-line staff.

Section VWhat coordinated infrastructure makes possible

The Monroe deployment was initially shaped around classic social-determinants-of-health use cases. But the same platform — the same closed-loop mechanics, the same governance structures — extends to the hardest coordination problems any community faces.

Re-entry from incarceration. A returning person's success depends on coordinating housing, ID and documents, behavioral health, healthcare, and workforce simultaneously. The standard failure mode is that each service is its own fresh first impression. A CIE changes that pattern. A release-planning conversation can generate coordinated referrals across the whole care team. The first weeks become connected, not consecutive.

Workforce transitions. Job training is necessary but rarely sufficient — transportation, childcare, health care, and stable housing are typically the determining factors for whether a placement holds. A CIE is the only tool that actually coordinates those basic-needs services with the training program itself.

Rural health transformation. Where provider density is lower and individual organizations are smaller, the coordination layer matters even more. A CIE lets a rural clinic, a community action agency, a school-based health program, and a county HHS office function together at a scale none of them could individually.

Section VIWhat we ask

This is a deliberately general case. It's written to be legible to a state legislator, a health plan medical director, a county public health officer, a foundation program officer, and the leadership of a community convening body — because any of them can be the actor that makes the next CIE in their community real.

What we ask depends on who's asking. Funders: consider funding the connective tissue alongside funding the programs. Agencies: look at CIE as a layer across the programs you already administer, not an addition to them. Health plans and systems: this is where value-based care meets the social determinants in a way that's operational, not aspirational. Community convening bodies: you already hold the relationships; a CIE is how you operationalize them.

And to anyone reading this who has been in this work long enough to be rightly skeptical of the next new platform, new idea, new proposal: the case for coordinated infrastructure doesn't rest on a prediction. It rests on an existing deployment — currently in production, operating at scale, producing outcomes no fragmented system produces. The question isn't whether it works. It's whether your community is ready to build one.