They trained at accredited schools and passed the exam-maker’s own official practice test. Then they failed a licensing exam that no party outside the exam-maker can inspect, rescore, or appeal — and at the country’s only historically Black college of veterinary medicine, the pass rate fell from roughly nine in ten to barely half in six years. A new NAVEC report assembles what the public record shows, and is careful about what it does not: it documents effects and structures, attributes every contested claim to its named source, and asserts no intent.
This is not an argument about diversity numbers, and it is not a claim that anyone has been caught doing wrong. It is narrower and, in a way, harder to answer. A group of candidates trained at accredited veterinary programs. They sat the exam-maker’s own official practice test — the instrument sold as a predictor of how they would score — and passed it. Then they failed the real licensing exam, in one direction, at a scale ordinary error struggles to explain. And because that exam can be inspected, rescored, or appealed by no one outside the organization that owns it, no independent party can say why.
A new NAVEC report, Despite Equal Merit, assembles what the public record shows about that pattern. It is disciplined about the line between fact and inference: it documents effects, structures, timelines, and statistics, and it attributes every contested allegation — of bias, of discrimination — to the named party making it, not to NAVEC. What follows is a readable tour of that evidence and, just as importantly, of its limits.
Begin with what the report is not claiming, because the distinction is the whole point. It is not measuring whether veterinary medicine has too few practitioners of one background or another. That is a real debate, but it is a different one. The question here concerns candidates of demonstrated equal merit — people who cleared the same academic bar, graduated from the same accredited programs, and passed the exam-maker’s own predictive instrument — and asks a single thing: when their outcomes on the licensing exam diverge sharply and in one direction, can anyone outside the exam-maker verify that the result is fair?
Framed that way, the merit objection falls away. These are not candidates about whom one could say they were unprepared; the exam-maker’s own practice test said otherwise. The report’s concern is not the gap itself but the impossibility of auditing it.
The candidates at the center of this report trained at AVMA-COE–accredited programs and passed ICVA’s own official NAVLE Self-Assessment — the instrument ICVA states “is predictive of later performance on the NAVLE.” Merit and readiness are not in question. Verifiability is.
The clearest place to see the stakes is a single institution. Tuskegee University operates the only historically Black college of veterinary medicine in the United States, and its graduates account for a substantial share of the nation’s Black veterinarians. As recently as 2018, roughly nine in ten of its graduating candidates passed the NAVLE — a figure reported in the public account of veterinarian Dr. Crystal Heath and consistent with the college’s historically reported rates. By 2024, declining year over year, that figure had fallen to about 51%, as reported by VIN News. The national first-time pass rate moved only a handful of points over the same period.
A forty-point collapse against a several-point national dip is the kind of divergence that demands an explanation — and there are ordinary ones that deserve a fair hearing. A college’s outcomes can shift with its finances, its clinical resources, its faculty, or its class composition, and Tuskegee has faced documented pressures on several of these fronts; its NAVLE results became one factor cited in a period of probationary accreditation. The report does not assert that any of this is proof of targeting. It asserts something narrower and, on the record, undeniable: no party outside ICVA can test which explanation is correct, because the school-level and item-level fairness data that would settle it is not published.
Part of the public argument turns on a precise question: does a human hand ever shape an individual candidate’s result? By ICVA’s own account, no — the NAVLE is machine-scored, and human judgment enters only upstream, where subject-matter experts write items and a standard-setting panel fixes the passing score. On that account, an allegation that someone reached in to alter one candidate’s answer sheet is inconsistent with how the exam works.
The report takes that claim seriously and then tests it against an episode ICVA itself published. In January 2024, after score reports for the autumn testing window had been released, ICVA discovered a problem in how a small number of reports had been assembled. It terminated access to all reports, had them corrected, and reissued them within hours — and, by its own statement, a subset of candidates had briefly received another candidate’s name, exam ID, and comparative performance data. The detail that matters is not the size of the error. It is what the episode proves: that a human administrative layer stands between the scoring engine and the candidate, that issued individual score reports are mutable, and that this layer operated through exactly the kind of undocumented discretion the “machine-scored, untouchable” narrative locates safely out of reach. That does not show anyone altered a result. It shows the public assurance is more porous than it sounds.
ICVA releases NAVLE scores roughly four to five weeks after a testing window closes, disclosing only that “a final quality control check” occurs. Every comparable machine-scored licensing exam — nursing’s NCLEX, medicine’s USMLE, pharmacy’s NAPLEX, the bar’s MPRE — publishes what its own wait is for: equating, item analysis, quality assurance. The NAVLE’s window is at the slow end of normal. The absence of any published explanation of what the weeks are for is not.
Running underneath every one of these threads is a single structural fact. The candidate agreement every examinee must sign reserves to ICVA the sole-discretion right to withhold, cancel, or invalidate a score through statistical analysis — with no rescore, no appeal, and no refund. The power to question a result runs one way only. A candidate who believes an individual score is wrong has no mechanism to test it; the organization that issued it has every mechanism to revisit it. By contrast, nursing offers candidates whose results are cancelled without a finding of irregular behavior a free retest, and the medical and bar exams offer paid score rechecks. On the NAVLE, the asymmetry is complete and documented.
The pattern at Tuskegee, the one-directional failure among candidates who passed the official practice test, the human layer the January 2024 episode exposed, and the unexplained release window share one feature: each is visible from outside, and none can be verified from outside. The report’s ask follows directly and is deliberately modest. Publish the school-level pass-rate and subscore data the medical and nursing boards already publish. Publish item-level fairness statistics — differential item functioning — across demographic groups. Open the scoring apparatus, and the January-2024 administrative layer, to an audit ICVA does not itself control. And give candidates a genuine route to review of an individual result.
ICVA, through counsel, denies that the exam discriminates or that its outcomes reflect bias. That denial, like the allegations it answers, cannot be tested against data the public cannot see — which is precisely the report’s point. A profession that conditions the right to practice on a single exam score should not have to take the fairness of that score on faith. Reform here does not require proving anyone acted in bad faith. It requires only that the exam standing between a candidate of equal merit and a license be open to the scrutiny we expect of every other gatekeeper.
Read the report — Despite Equal Merit NAVEC’s full examination of the evidence and attributed allegations of racial and ethnic targeting in the NAVLE: the Tuskegee case, the scoring forensics and the January 2024 incident, the statistical record, and what genuine independent verification would require.Sources: VIN News reporting on Tuskegee’s NAVLE outcomes and on the January 2024 NAVLE score-report incident; the public account of Dr. Crystal Heath regarding Tuskegee’s 2018 pass rate; ICVA’s own January 2024 announcements and published score-release language; the NAVLE Candidate Agreement; and the comparative score-release and appeal policies of the NCLEX, USMLE, NAPLEX, and MPRE programs. This article distinguishes documented fact from inference: it documents effects, structures, timelines, and statistics from the public record, asserts no intent, and attributes every contested claim — including allegations of discrimination or bias — to the named party advancing it rather than stating it in NAVEC’s own voice. NAVEC is a nonprofit advocating for fairness, transparency, and accountability in veterinary licensing.
← Back to Research & Reports