Performance Variation Between CoCs Is a Build Problem, Not a Willpower Problem

By Common Ladder · May 25, 2026 · 9 min read

Two CoCs in the same census region

Two CoCs serve mid-sized metros in the same census region. Similar PIT counts, similar climates, similar funding mixes. One has cut chronic homelessness by roughly a third over five years. The other's count is flat. The conventional explanation reaches for leadership, culture, or political will. The evidence reaches for something more boring than that.

The high-performing CoC has a coordinated entry system that triages. The average one has a coordinated entry system that produces a waitlist. The first has rapid rehousing capacity sized to its inflow; the second has rapid rehousing capacity sized to last year's grant cycle. Both run an HMIS. Only one uses it to make weekly decisions about who to house next. The difference between these two CoCs is not effort. It is the maturity of three systems that take years to build and that, once built, do most of the work.

Thesis

What separates high-performing CoCs from average ones is the maturity of three structural systems — coordinated entry, rapid rehousing capacity relative to inflow, and operational HMIS use. Treating CoC variation as a performance problem rather than a build problem is the single biggest analytic mistake the field is currently making.

Coordinated entry maturity is not a binary

Kim & Sullivan (2023) used HUD AHAR data across 380+ CoCs to model the structural correlates of permanent housing exit rates. The strongest single predictor was not total per-capita system spending. It was coordinated entry maturity — measured through time-to-assessment, prioritization-logic transparency, and inflow-to-housed reconciliation cadence [F-21].² The finding worth pausing on: spending without the structural systems to deploy it produces less throughput than less spending with the systems in place. Coordinated entry, in other words, is the conversion layer that turns funding into exits. Without it, dollars sit.

Most CoCs check the coordinated entry box because HUD requires it. The literature is telling us that the box-checked version and the operationally mature version are different objects. The box-checked version logs people. The mature version triages them, reconciles inflow against capacity weekly, and updates prioritization in response to system-level data. This is not a matter of intent. It is a matter of build.

Rapid rehousing capacity is a ratio, not a headcount

Jenisa & Jang (2025) extends the analysis with newer data and a sharper focus on rapid rehousing as the proximate exit lever. Their finding: RRH capacity matters less in absolute terms than as a ratio to monthly inflow [F-21, PROVISIONAL]. CoCs with RRH capacity below 30% of monthly inflow rarely sustain reductions. CoCs above 60% almost always do. The middle band is mostly noise.³

This is the kind of finding that should change how funders and CoC leaders talk to each other. The standard conversation is about RRH unit counts. The conversation the evidence is asking for is about RRH-to-inflow ratios. A 200-unit RRH program in a CoC with 800-person monthly inflow is structurally different from a 200-unit program in a CoC with 250-person monthly inflow, even though the unit count is the same. Below the threshold, capacity simply absorbs into wait time. Above the threshold, it converts inflow to exits.

This is also why "we added 50 RRH units last year" sometimes produces measurable system change and sometimes doesn't. Whether 50 units matters depends on what 50 units is a fraction of.

HMIS used for operations is not the same as HMIS used for reporting

Nisar et al. (2019) — older but durable — established that the data infrastructure variable is real and not a proxy for funding. CoCs with HMIS used for operational decisions outperformed CoCs with the same funding but compliance-only HMIS use, with the largest gap on permanent housing destination outcomes and the smallest gap on emergency shelter throughput [F-21].⁴

The distinction matters because HMIS investment looks identical from the outside. Two CoCs both report on time, both pass HUD data quality checks, both publish their AHAR submissions. Only one is running weekly case conferences off the data, identifying inflow surges in time to redirect capacity, and using the data to make actual prioritization decisions instead of post-hoc justifications. The funding lines look the same. The systems are not the same.

This is the cheapest of the three structural systems to fix and culturally the hardest. The technology is mostly there. The practice of using it for operations rather than reporting is what's missing.

What this means for benchmarking

The standard practice in field benchmarking — ranking CoCs by exit rate, return rate, or length of stay — treats variation as performance. The literature suggests this is wrong on two counts. First, it punishes CoCs that have not been resourced to build out the three structural systems even when their per-capita effort is high. Second, it rewards CoCs whose structural systems are mature without distinguishing whether maturity was earned through investment or inherited from earlier funding cycles.

A better benchmarking practice would control for the three structural variables and then rank CoCs on residual performance. This is harder than what funders currently do. It is the comparison that distinguishes "well-built and well-run" from "well-built and ordinarily run." The first deserves replication study. The second deserves operational support.

What this means for CoC leadership

If your CoC sits in the average band, the literature is telling you the path to high performance does not run through harder work. It runs through deliberate, sustained investment in the three structural systems. Coordinated entry maturity is usually the highest-leverage starting point because it changes the unit of work — from intake to triage — without requiring proportional funding increases. RRH-to-inflow ratio is the most expensive to fix and usually requires a multi-year capacity build, but it is also the structural variable that most directly converts effort into exits. Operational HMIS use is the cheapest to fix and the most culturally hard, because it requires shifting the data team from reporters to operators.

Name the weakest of your three structural systems out loud. Then build a three-year plan to fix it. The literature is consistent: there is no shortcut, and there is also no mystery.

The PROVISIONAL caveat

The evidence base on CoC-level variation is thinner than the literature on individual interventions, and most of the cited analyses use HUD AHAR data, which has known limitations around veteran double-counting, racial categorization, and chronic homelessness flags. The three structural predictors recur across studies and across methodologies, which is the basis for treating directionality as firm. The specific magnitudes — the 30%/60% RRH-to-inflow thresholds, the size of the coordinated-entry effect — should be read as informed estimates rather than fixed coefficients [F-21].

Implication

For funders: when you next assess a CoC for additional investment, ask which of the three structural systems they are building, and what their five-year trajectory on each looks like. A CoC at 25% RRH-to-inflow with a credible three-year plan to reach 60% is a different investment than one at 25% with no plan. The first is a build problem. The second is a different conversation.

For CoC leaders: the field has trained you to talk about performance in terms of effort, leadership, and outcomes. The evidence is telling you to talk about it in terms of structure. The frame change matters because it tells you where to put your next three years of attention.

Build the systems. The outcomes follow.