Community college retention: what to do with a 63% median

Median full-time retention at public two-year institutions is 63%, computed across 793 institutions from the fall 2022 cohort returning in fall 2023 (IPEDS, Fall 2023, EF2023D). The all-sector median is 71%, and private nonprofit four-year institutions sit at 77%.

The 14-point gap between community colleges and private four-year institutions is the most important number in the sector, and it is routinely misread in both directions. Read as a performance gap, it produces demoralizing goals. Read as a structural gap, it produces a workable list.

Read the gap correctly first

A 68% retention rate is a crisis at a selective private college and an achievement at an open-access community college. Comparing your institution to the 71% all-sector median tells you almost nothing, because the composition of the two student bodies differs on every variable that predicts persistence.

What drives the sector gap is structural, and each element is visible in your own IPEDS record:

• Open access. No admissions filter means no pre-selection for the characteristics that predict persistence.
• Part-time enrollment. A large share of students carry loads that put them below the threshold for nearly every cohort structure institutions build.
• Commuting and employment. Almost no students accumulate the passive attachment that residential life produces.
• Age and dependents. A higher share of students over 25 with caregiving and work obligations that compete directly with programming.
• Transfer intent. A student who transfers and graduates elsewhere counts against retention while representing a success. Persistence, not retention, is the honest measure for that student.

That last point deserves attention in any board conversation. If a meaningful share of your students enroll intending to transfer, your retention figure is measuring something other than what your institution is for. Report both, and say which is which.

What the structure implies about program design

Every item on that list points the same direction: interventions that assume presence, unstructured evenings, or student-initiated contact will underperform at a community college regardless of quality.

Four design constraints follow.

Schedule inside the academic day. The only hour a commuting, working student reliably has is the one between the classes they came for. Evening programming is a screening mechanism that selects against your entire student body.

Offer an asynchronous pathway that was designed as one. A recorded version of a live program is not the same thing. A pathway completable across weeks, on the student’s own schedule, is what a working parent taking nine credits can actually finish.

Recruit by nomination. Campus-wide announcements reach students who read campus email, which correlates with students who are already attached. A faculty member naming five students reaches a different population, and at an open-access institution the course instructor is frequently the only staff member a student ever really knows.

Build peer groups, not office hours. Staff-led contact does not scale to an open-access enrollment on an open-access budget. A fixed group of six students with a standing commitment and a shared goal notices an absence faster than any early-alert dashboard and reaches the missing student more credibly than a dean’s email.

Cost is the constraint, and it is measurable

Community colleges have the least capacity to add staff and the largest population to reach, which means anything priced per staff hour is unaffordable at the scale required.

In our State of Higher Ed 2026 research, only 32.9% of administrators say their budget is adequate to serve student needs, while 77.6% wish for greater resources to connect with students. That is not a motivation problem. It is an arithmetic problem, and it means the only structures worth evaluating are ones where delivery does not consume staff hours linearly.

Externally supplied structure is worth pricing against in-house build for exactly this reason. What matters in the comparison is the total: whatever the institution is charged, plus staff hours at loaded salary, plus any student-facing fee stated openly rather than buried. For an NSLS chapter the institutional figure is zero. Our page on what an NSLS chapter costs separates those components, and the advisor time commitment covers the hours side.

The proof point that comes from your own sector

Named outcomes from private four-year institutions are of limited use in a community college board packet. Two published outcomes come from open-access institutions. At Delgado Community College, students who earned their NSLS induction reached an 81% semester-to-semester retention rate, nearly 20 percent higher than that institution’s overall rate. St. Petersburg College, which IPEDS classifies as a public four-year because it awards bachelor’s degrees but which admits on an open-access basis and runs mostly associate-level programs, ran a one-campus pilot in which 24 invited NSLS members were retained at a rate 37 percent above the college-wide rate, according to a 2020 NSLS white paper.

Read both carefully. Delgado’s compares inducted students with the institution’s overall rate; St. Petersburg’s compares 24 students invited to a one-campus pilot with the college-wide rate, so it describes a small, selected group. Students chose to join in both cases, so selection is part of each result: a structured program attracts the students already inclined to engage as well as changing students. The claim they support is that structured involvement is associated with persistence at open-access institutions specifically, not a size of effect to expect on your campus.

What they usefully establish is that the mechanism transfers. The design question, which students the structure actually reaches, remains yours. Our community college segment page sets both outcomes against the sector benchmarks.

How to present the number to a board

Three moves make a community college retention presentation defensible.

1. Lead with the sector median, not the national one. 63% across 793 public two-year institutions is the comparison. Anything else invites a conclusion the data does not support.
2. Report persistence alongside retention and name the transfer-intent share explicitly. A retention figure presented without it understates your institution.
3. Use concrete peers rather than rankings. Comparable institutions of similar size and Pell share sustaining higher rates is a useful, actionable frame. Percentile and quartile language produces defensiveness and stops the conversation.

Do this next

Pull your institution’s own IPEDS record and put three figures side by side: your full-time retention rate, the 63% public two-year median, and your part-time share. The third number explains most of the distance between the first two, and it tells you immediately whether your engagement programming is built for the students you actually enroll.

For sector benchmarks, the mechanisms behind attrition, and how to construct a comparison group leadership will accept, see our guide to student retention and persistence.