| Field family | Source | Vintage |
|---|---|---|
| LSAT / uGPA 25-50-75, acceptance, enrollment, scholarships, tuition, bar passage, employment outcomes, transfers | ABA Standard 509 Required Disclosures | 2011 – 2025 (15 yrs) |
| State attorney median / 25th / 75th wages | U.S. BLS OEWS (occupation code 23-1011) | 2018 – 2025 |
| State cost-of-living index | U.S. BEA Regional Price Parities | 2012 – 2024 |
| State fair-market rent | U.S. HUD Small Area FMR | 2019 – 2025 |
Every cell in Exhibit traces back to one of these four sources. The site is a re-organization of public data, we do not survey schools, we do not generate proprietary scores, and we do not estimate fields that the underlying source did not publish.
For each school we use its published 2025 LSAT and uGPA 25th and 75th percentile bands (the "middle 50%" of last year's entering class). When you enter your stats:
The single helper is fitClass(school, lsat, gpa). There is no second weighting, no adjustment for "soft factors," and no proprietary score, the rule above is the entire rule. If you enter only one stat we apply the rule to that one stat alone.
If a school has not published a 25/75 (rare, happens on new accreditations), it's excluded from your Safety/Target/Reach results until the next disclosure cycle.
The Net Price calculator uses each school's resident sticker tuition if your selected home state matches the school's state; otherwise the non-resident figure. We subtract the school's median grant amount (from the same 509 disclosure) when no fitClass-based estimate is available; with fitClass we use the band: Safety → school's 75th-pct grant, Target → median grant, Reach → 25th-pct grant.
Loan amortization uses the standard fixed-rate formula P · r · (1+r)n / ((1+r)n – 1) over your selected term. The 3-year living cost is (housing + other) × 9 academic months × 3 years. These are illustrative. Important: under the One Big Beautiful Bill Act, students entering from July 1, 2026 can no longer use Grad PLUS, and federal graduate borrowing is capped at $50,000/year ($200,000 lifetime). Any cost above that cap must come from private loans (typically higher rates, and no PSLF or income-driven repayment), so the “true cost” for 2026+ entrants can be materially higher than a pure federal-borrowing model implies. Confirm current federal and private loan rates with your lender before relying on the total cost number.
The headline First-Time Bar figure is the ABA-published first-time pass rate for the most recent reporting year (currently 2025, describing the 2024 graduating cohort). The 2-Yr Ultimate figure is a separate ABA disclosure with a longer measurement window, it captures graduates who passed within two years of graduation. The two figures are not directly comparable across cohorts.
Importantly: a school's bar on any given row describes graduates who entered roughly three years earlier, while the apps/offers/acceptance figures on the same row describe the most recent entering class. Treat the bar number as evidence about the school three years ago.
FTLT Employment = "Full-Time, Long-Term, JD-Required or JD-Advantage" employment 10 months after graduation, per the school's ABA 509 outcome report. We display it as both a percent of the graduating class and an estimated graduate count (graduates × pct / 100).
The 10-category employment breakdown (MegaLaw 500+ / BigLaw 251-500 / Mid 101-250 / Small 2-100 / Clerkships / Public Interest / Government / Business / Academia / Solo / Seeking) is reported by the school in the same disclosure. We rank rows by applicant-mental-model desirability rather than alphabetically. The "vs national average" delta on each row compares the school's percent against the unweighted average across all 197 reporting schools.
Every line chart on the site is built from year-by-year ABA 509 disclosures back to 2011. Where a metric was not collected in a given year (most notably the bar pass cadence change 2012-2017), the chart renders a dashed bridge through the missing years with hollow markers labeled "YYYY not reported (interpolated)". The interpolation is for visual continuity only, we never replace null values with synthetic numbers anywhere a real cell exists.
Two source files are reconciled, a hand-maintained "Historic" spreadsheet (treated as primary because it has been spot-checked against PDFs since 2011) and a machine-generated "v2" spreadsheet (used to fill gaps the Historic file does not cover). The two files agree on 91.8% of overlapping cells; remaining disagreements are documented in a separate file shared with the maintainers.
Every transformation is reversible from the original public disclosure. The methodology is public; the input files are public; the transformations are documented above.
A year-over-year scan of every school's tuition flagged abnormal moves (any single-year change above +18% or below −10%). Four were unambiguous single-year semester-vs-annual misreports and were corrected directly in the dataset; the replacement is interpolated from the school's neighboring years, pending a re-pull of the original 509 figure:
A larger set shows multi-year level shifts that we are deliberately not auto-correcting, because several are likely genuine tuition resets (a number of schools cut tuition during the 2014–2015 enrollment downturn) rather than reporting errors. Fabricating a "correction" would erase real history. These carry an "under review" flag on their school page until verified against the original disclosure: Baylor, Texas Southern, Tulsa, Ohio Northern, Roger Williams, Brooklyn Law, Iowa, Texas A&M, North Carolina Central, Mississippi, Elon, Howard, Toledo, Arizona, and North Texas at Dallas.
Trend charts extend the last reported year with a dashed 4-year projection and a shaded uncertainty cone. This is a model output, not a forecast of fact — it answers "if this school's own recent trend simply continued, where would it land?" and nothing more. A school can break its trend at any time; treat the dashed region as a sketch, not a promise.
Method (fit to each school's own history, never the national average):
The math is deliberately simple and transparent so anyone can reproduce or challenge it. It does not model policy changes, the economy, or a school's strategy — only its recent numerical trajectory.
The ABA publishes Standard 509 disclosures once per year, typically in the December–February window for the prior calendar year's data. We re-sync within the month following each ABA release. The "Last synced" date at the top of this page and in the disclaimer bar always reflects the most recent inline pull.
Wage (BLS), cost-of-living (BEA), and rent (HUD) feeds follow their own annual cadences, we re-pull each within a month of the respective agency's release.
—.Email blake@exhibit509.com or use the feedback form linked from the footer of any page. Data errors get fixed and the fix is documented in the next sync note.