Methodology
Dataset verified 2026-09-27
Two classes of fact, labelled differently
This site publishes two kinds of claim and treats them differently, because they carry different risk.
Eligibility rules — who belongs to which compact, what rate the compact sets, what a state's residency test requires — are checked against the body that administers them: the compact commission (WICHE, MHEC, SREB, NEBHE) or the state's own higher-education authority. These are the site's load-bearing answers and every page lists the sources used, with the date each was read.
Dollar figures are compiled from each institution's published rate page. They are marked in the dataset with a basis of published_compiled rather than primary_verified, and every page that shows one says so. All 153 rate blocks currently carry the compiled basis; they are re-checked each June against the institution's own published rate, and promoted as they are confirmed.
What the figures include
- Tuition plus mandatory fees for a full-time student, per academic year, for 2025–26.
- Rounded to the nearest $100, because false precision on an annually-revised figure is worse than none.
- Excluded: housing, food, books, travel, health insurance, and course-specific differential tuition.
Differential tuition is the biggest omission and it is deliberate: engineering, nursing and business programs frequently carry per-credit surcharges that vary by campus and by year, and folding an average of them into a headline rate would describe no real student's bill.
One reference institution per state
Each state is represented at each level by a single named institution rather than a system average. An average across campuses of very different cost produces a number nobody is actually charged. The reference institution is named in every table and linked to its own rate page, so the figure can be checked at source in one click.
For four-year study the reference is the state's flagship public university. For two-year study it is the state system or the largest single system where the state does not price centrally. For graduate study it is the flagship's graduate school base rate.
How pair answers are derived
Pair outcomes are computed, not written. A rule engine evaluates each ordered pair against four things in sequence: same-state, bilateral agreements between the two states at that level, regional compact membership on both sides at that level, and portable home-state awards. Every applicable route is returned and ranked by estimated annual cost, with conditional routes — those gated on the major or the campus — flagged as such.
The engine is a pure function with no per-state special cases. Every branch it takes is driven by a field in the dataset, which means a data correction propagates to all 7650 pair pages and to the checker at the same time, with no manual editing.
Direction is preserved
Agreements are frequently one-directional and the dataset records which way each one flows. Texas waives non-resident tuition for students from named counties in four neighboring states and receives nothing in return, so the answer for Oklahoma-to-Texas is not the answer for Texas-to-Oklahoma. Level participation works the same way: Florida and Texas belong to the Academic Common Market at graduate level only. Treating an agreement as symmetric when it is not would be a fabricated fact, so the schema does not allow it.
Missing data is shown as missing
Where a value cannot be verified it is stored as null and rendered as "not published", never estimated and never quietly dropped. A route with no published figure still appears in the results table — it sorts last, but it is not hidden, because a student who cannot see it cannot go and look it up.
The build refuses incomplete data
Completeness is enforced mechanically rather than by review. The build fails if any of the following is true: fewer or more than 51 jurisdictions; any jurisdiction missing any of the 3 level blocks; any row without at least one source; any row without a valid verification date; any bilateral agreement naming a state that does not exist; any compact member outside the frozen registry. There is no override.
Update cadence
- Annually, each June — tuition tables, as institutions publish the coming year's rates.
- Semi-annually — compact membership and level participation, which change by legislative act rather than on a schedule.
- On report — any factual correction sent to [email protected], verified against the primary source and re-dated.
Download the data
Every figure and rule on this site is generated from three JSON files, one per level of study. They are free to reuse under CC BY 4.0 with attribution to instatecheck.com. Data verified 2026-09-27.
Limits worth stating plainly
- Campus-level participation is in the dataset only for WUE, from WICHE's own list of participating schools, and for the single-campus agreements it records (such as UNI's Neighboring State Rate). For the other compacts, whether a particular university offers the program, and for which majors, has to be confirmed with that university.
- Institutional non-resident scholarships are not covered. They are frequently the largest available discount and they are not published centrally in any comparable form.
- Approved-program inventories for ACM, WRGP and Tuition Break change annually and are specific to your home state. This site tells you the gate exists; the commission tells you whether your degree clears it.
- Projections hold tuition flat. Real tuition rises most years, so multi-year totals are a floor.