Skip to main content
DLRestoreMap
Methodology

How this site is built

Per-state restoration rules change. Some change quietly (a fee schedule update); some change loudly (an all-offender IID expansion). The methodology below is how we keep claims honest and how we let you tell when a cell is verified versus when it's still on the research roadmap.

What counts as a primary source

  • State DMV portal — the official state government driver-services page (e.g. dmv.ca.gov, flhsmv.gov).
  • State legislature URL — the state-statute page (e.g. leginfo.legislature.ca.gov, flsenate.gov).
  • Cornell LII for federal preemption anchors (42 USC §666(a)(16), 49 CFR §383.51, 23 USC §164).
  • State IID-administrator portal where the program is administered separately from the DMV.

What is NOT a primary source

  • Justia state-by-state summaries — cross-walk only.
  • Nolo procedural articles — cross-walk only.
  • DUI-defense-firm SEO content — never load-bearing.
  • Generic news / explainer sites.
  • AAMVA tables — orientation reference, not authoritative on per-state implementation.

Unverified cells

If a cell could not be verified against primary source in the current research, the cell shows “Not yet verified” instead of a value. We do not paraphrase from secondary sources. We do not invent fee amounts. We do not estimate IID durations or SR-22 filing periods. A visible “Not yet verified” is honest; a fabricated rule is dangerous.

Per-cause matrices cover 10 states, chosen by combined population and variation richness. The pages of the other 41 jurisdictions render an "awaiting research" notice rather than fabricating a matrix.

Cost-stack methodology

The reinstatement cost stack adds up: state DMV reinstatement fee + IID monthly lease × duration + SR-22 premium differential + DUI school cost + treatment cost. The state DMV reinstatement fee is the only component with a single fixed dollar amount (state-statutory). Every other component is a range, and we mark ranges as ranges. Components that have not been verified to carrier or provider disclosure read “Not yet verified”.

See the cost-stack pillar for the full component breakdown.

Decoder engine

The decoder is deterministic: same inputs → same outputs, every time. Inputs are state code, suspension cause class, offense date (optional, enables date-aware branching for post-2023 court-debt-suspension reforms), prior suspension count, hardship interest, and employment-required-driving flag. The engine reads four state-matrix JSON files at build time. No network calls. No randomness. Validated against a 5-fixture test suite that runs in CI.

Privacy

The decoder runs in the browser. State, suspension cause, offense date, prior count, and hardship interest are sensitive financial-distress / legal-distress data and are never sent to a server. We log aggregate verdict counts only (state code, suspension cause class, hardship interest yes/no) — never offense dates, prior counts, or any combination that could re-identify a driver. See how the decoder works for the technical detail.