New River Analytics · Accountability Suite

Is WV assessment quality getting better or worse?

West Virginia’s statewide residential assessment record, year by year, from 2003–2025. Each point is a direct, time-matched sales-ratio study — the assessor’s appraised value in the sale year divided by the actual arms-length sale price — rolled up across all qualifying sales statewide. We track the three core IAAO measures: uniformity (COD), vertical equity (PRB), and the level (median ratio).

Statewide only — and honestly framed. Only the statewide series is precomputed by year; a per-county-by-year breakdown is not available, so this page is the whole state, not any single county or assessor. Uniformity (COD) and vertical equity (PRB) are the quality measures we read for a trend. The level (median ratio) is shown for context only — it is confounded by WV’s mandated multi-year reappraisal cycle and by the housing market, so a falling level is not by itself a quality failure.

The headline trend

Over the 22-year window (2003–2025), West Virginia’s statewide uniformity has worsened: the coefficient of dispersion went from COD 21.8 in 2003 to 44.2 in 2025 (a rise of 22.4 points). Both endpoints sit well above the IAAO residential target of COD ≤ 15.0, so the roll has been non-uniform throughout — and the dispersion has only grown.

Vertical equity tells the same story: regressivity has deepened. PRB moved from -0.0128 in 2003 to -0.1404 in 2025. A negative PRB means lower-value homes are assessed at a higher share of market value than high-value homes; the more negative it gets, the heavier that tilt onto people who can least afford it.

Best uniformity year on record: 2003 (COD 21.8). Worst: 2025 (COD 44.2).

Uniformity (COD) WORSENED
21.8 (2003) → 44.2 (2025) · lower = better
Vertical equity (PRB) DEEPENED
-0.0128 (2003) → -0.1404 (2025) · 0 = equitable
Level (median ratio) CONTEXT
0.8216 (2003) → 0.7035 (2025) · cycle/market

Three measures, 20+ years

Uniformity — COD by year

How evenly comparable homes are valued. Lower is better; the dashed line is the IAAO residential target (COD ≤ 15.0).

Vertical equity — PRB by year

Price-related bias. The dashed line at 0 is perfect equity; below it the roll is regressive (lower-value homes over-assessed relative to high-value homes).

Level — median ratio by year (context)

Where assessments sit vs. market. The dashed line at 1.0 is full market value. Confounded by the reappraisal cycle and the housing market — not a quality grade.

Sample size — qualifying sales by year

How many arms-length residential sales back each year’s metrics. Thin early years carry wider uncertainty.

Read the early years with caution. The first few years rest on far fewer sales than later years (the typical year has 12,297 qualifying sales). Thin years — 2003, 2004 — have noisier COD/PRB estimates, so short-run wiggles there mean less than the long-run direction. The 2025 point is also a partial / still-settling roll year (n=6,203) and may revise.

How to read this

Statewide, by design. Only the statewide series is precomputed by year. A per-county-by-year trend is not available in the published data, so nothing here should be read as a single county’s or assessor’s trajectory.

COD (uniformity) is the average percent deviation of sales ratios from the median — the core IAAO measure of whether like homes are treated alike. Lower is better; the residential target is COD ≤ 15.0.

PRB (vertical equity) is the price-related bias coefficient: 0 is equitable, negative is regressive (low-value homes over-assessed relative to high-value homes), positive is progressive. IAAO considers |PRB| within about 0.05 acceptable.

Median ratio (level) is reported as context only. WV mandates a multi-year reappraisal cycle, and the 2020–2024 housing boom pushed assessments below market almost everywhere. So the level’s ups and downs reflect the cycle and the market far more than assessment practice — we do not grade it.

Source. Statewide time-matched residential sales-ratio study (avm/ratio_study.json → by_year). Appraised values are the 100% appraised basis, not billed taxes. See the site methodology and the Assessor Report Card for the per-county cross-section.