Contract Disputes by State — Methodology
DocsGenie Research · How this dataset was built, what it measures, and the limits of what it can claim.
This page documents the sources, calculations, and scope rules behind the DocsGenie state-level analysis of U.S. contract litigation. It is written so that a journalist, researcher, or analyst can evaluate and reproduce the work. The dataset is available as a free download (see Download & citation).
What the dataset measures
The core metric is contract case filings per 1,000 businesses, by state. Raw filing counts on their own mostly track how big a state is; normalizing against the number of businesses is what produces a comparable, map-able figure. We report the rate on two bases and corroborate it against population:
- Year-matched rate: 2022 filings ÷ 2022 establishments × 1,000.
- Current-volume rate: 2024 filings ÷ 2022 establishments × 1,000 (denominator lag disclosed).
- Lag-free corroborator: 2024 filings ÷ 2024 population × 1,000.
Data sources
| Source | Role | Vintage |
|---|---|---|
| NCSC — National Center for State Courts, Court Statistics Project | Numerator: incoming contract case filings, 2020–2024. Received directly from the CSP by data request. | 2020–2024 |
| Census SUSB — Statistics of U.S. Businesses | Denominator: business establishment and firm counts by state. | 2022 (latest available) |
| Census Population Estimates | Secondary denominator: per-capita corroboration. | 2024 |
| BLS OEWS — Occupational Employment & Wage Statistics | Legal-labor wage context (lawyers 23-1011, paralegals 23-2011, mediators 23-1022). | May 2024 |
Coverage and scope
This is not a national dataset. The NCSC extract covers states that report to the Court Statistics Project. In this data:
- 36 states/territories have 2024 filings (the headline cross-section).
- 29 states have all five years (2020–2024) — the clean backbone used for every trend claim.
- 17 states do not report and are excluded, including several large ones: California, New York, Ohio, Virginia, Washington, Tennessee, and Louisiana.
Composition — read this before quoting a number
The NCSC "Contract" category comprises: seller-plaintiff/debt collection, landlord/tenant, mortgage foreclosure, buyer-plaintiff, employment disputes, fraud, and contract-other. It excludes small claims, tort, and eminent domain.
Calculation rules
- Denominator anchoring: SUSB business counts lag ~2 years; the cleanest ratio pairs 2022 filings with 2022 establishments. The 2024-basis rate reuses the 2022 denominator with the lag stated, so any gap between the two rates is pure filing growth.
- Year-matching: Population Estimates are current; SUSB is not. Each rate names the exact numerator and denominator years.
- Descriptive only: All claims are descriptive (e.g., "filings rose X%"). No causal or prescriptive claims are drawn from the raw counts.
- Owned estimates vs. cited facts: BLS publishes wage rates, not hours per case. Any per-dispute dollar figure is a model we own and label as such; only the wage rate is a citable government statistic.
New York and California
New York is mapped from the NY State-Paid Trial Court Caseload dashboard by summing the lines that match the NCSC contract definition (contract, UCC, insurance, mortgage foreclosure, consumer credit, commercial, and housing), giving ~364,000 filings for 2024 — a mid-pack rate. NY granular case types exist for 2024 only, so NY has no 2020–2023 trend and its figure is likely a slight undercount. California does not publish a contract breakout; a data request to the Judicial Council is pending and will be mapped the same way when received.
Download & citation
The dataset is free to download and republish.
docsgenie_contract_disputes_by_state.csv— filings, rates, ranks, and coverage by state.docsgenie_legal_wages_by_state.csv— legal-occupation wages by state.DATA_DICTIONARY.md— column-by-column definitions.
DocsGenie Research, "Contract Disputes by State" (2026). Available at
docsgenie.net/research/contract-disputes.License: Free to republish, including charts and the dataset, with attribution and a link to the source page above.