The State of Merit-Aid Transparency, 2026

Finding 1 — Opacity

Roughly 3 in 10 published policies never state the rule

The single question that decides whether a $5,000 outside scholarship actually lowers a family's bill is: what does the college do with it? We read the published outside-scholarship policy of 588 colleges — every verified school in our data with a policy page we could classify — and on 135 of them (23.0%), the page never says how an outside award is treated. Restricting to the 543 human-curated pages (excluding 45 auto-compiled pages that default to "unclear" by construction), the rate is still 118 of 543 — 21.7%, roughly 3 in 10.

n = 588 verified pages with a typed outside-scholarship policy (543 human-curated + 45 auto-compiled). Schools with no located policy page are excluded, not counted as unclear.

Finding 2 — The stated rules

What the 453 schools that do publish a rule actually say

Among the 453 schools that state their policy, only 19 (4%) publish the most harmful rule — cutting your institutional grant first. That 19 is a documented floor, not a ceiling: another 135 schools publish no rule and could behave the same way. The clearest dollar-for-dollar examples are Princeton and Williams, whose aid is all-grant with no loans — there is no self-help cushion for an outside award to absorb first. The caveat matters: a grant-first rule only bites students who actually receive an institutional grant. Full-pay families are unaffected.

At the other end, only 14 of 588 schools (2.4%) publish a clean stacks-on-top policy, where the school states in writing that an outside award will not reduce institutional aid. The verified-clean examples we can name: the University of Oregon's Apex award, the University of North Dakota, and the University of West Florida. Everything else — 147 loan-first, 241 cost-of-attendance caps, 32 mixed or conditional rules — falls between, and most of it is benign for most students: a loan-first rule trims debt before grants, and a COA cap only binds students already packaged near the full cost.

Two schools deserve a footnote rather than a label. The University of Rochester's rule is a different mechanic entirely — merit fills need, which only affects need-eligible students. And Macalester, though its policy is typed grant-first in our data, is actually protective: it shields merit awards entirely and the first $10,000 of outside scholarships.

Denominators: 19/453 stated policies (4.2%); 14/588 typed policies (2.4%). Every named school's classification carries a verbatim excerpt from its own policy page.

Finding 3 — Predictability

The 21-point predictability gap

About 66% of public colleges (198/301) and 67% of faith-based private colleges (133/198) publish an automatic, stats-based merit award — a grid a family can read their likely award off before applying. Among secular private colleges, it is 45% (101/224). That is roughly a 21-point predictability gap between public and secular-private colleges, with faith-based privates sitting even higher (n = 723 schools with published merit tiers; the sector differences are statistically robust, two-proportion z of about 4.8). A strong student can often predict a public or religious-college offer in advance; secular-private merit usually has to be competed for.

The precise phrasing matters: these schools publish an automatic, stats-based award — they do not "guarantee" four years of money. Renewal GPA floors and application deadlines still apply, and they are exactly what to confirm with the aid office.

Finding 4 — Price and opacity

The cheaper the school, the less it publishes

Among the 492 schools with both a published cost of attendance and a classified policy, lower-cost schools are far more likely to publish no outside-scholarship rule at all: 31–32% of sub-$40k schools leave the rule unstated (53/164 pooled; 43/140 on curated pages only), versus roughly 1 in 4 at the $55k-and-up end (17/94 in the $55k–$70k band; 18/150 at $70k+). The pattern holds within public, private, and religious sectors, and it is steep-then-flat — opacity stops falling around $55k rather than declining all the way up.

This is not "cheap schools hide your money." An unstated rule is an unpublished rule, nothing more — and the practical consequence is the same one email. At a lower-sticker school, assume nothing and ask the aid office directly how an outside award is applied.

The evidence behind it

Built to be checked

Classified policies quoting the school’s own words

99.5%

585 of 588 carry a verbatim excerpt from the policy page they were read from.

Documented awards linked to a school source

100%

All 7,231 awards we document (4,492 merit tiers + 2,739 named scholarships) cite a source URL.

Ledgered claims labeled and withheld, not asserted

12.5%

649 of 5,210 claims carry an explicit "not enough evidence," "needs confirmation," or "sources disagree" label (as of June 10, 2026) and are excluded from verdicts.

Bonus finding — Deep grids

16 public universities publish a 10-plus-tier merit lookup table

Some public universities publish their entire entry merit program as a stats grid. 16 public universities we track publish 10 or more automatic, stat-keyed tiers — Southern Utah leads with 22 — and 36 schools across all sectors publish 8 or more. At these schools, a student with known scores and GPA can read a likely award straight off the table before applying. Tier count measures how finely the grid is mapped, not how large the awards are — and an automatic entry award is not an automatic four-year award, so pair any grid read with a renewal check.