The index is designed to make its judgments visible: what is measured, how much each area counts, where the data come from and where the limitations are.
Some areas in the index, including reproductive policy, are politically contested. They are included because state law materially changes access, legal autonomy or measurable outcomes, not because the index assumes every woman gives them the same importance. The published ranking uses a documented set of default weights, and the personalized tool allows users to raise, lower or remove the influence of individual domains.
The score combines direct measures, source rankings, proxies and scores carried forward from V2. The descriptions below state what the current workbook can support and identify the domains whose original source trail is incomplete.
Laws and access related to pregnancy, abortion, medication abortion and reproductive healthcare, including whether women can make and act on reproductive healthcare decisions.
Women's earnings relative to men's earnings, plus a smaller economic score carried forward from V2.
A carried-forward V2 childcare score plus a 2025 assistance-system stability measure.
The Commonwealth Fund's 2024 state ranking for women’s health and reproductive care.
Oxfam America's 2025 Best States for Working Women policy ranking.
A state leave/care score carried forward from V2.
A carried-forward V2 safety score plus one current domestic-violence firearm policy.
A state civil/legal score carried forward from V2.
An indirect economic proxy, not a direct measure of housing cost or housing protection.
A state political score carried forward from V2.
Two state menstrual-product policies.
The domains carry different weights. A five-point gain in reproductive autonomy or economic equality has more effect on the overall index than a five-point gain in a 2% domain. The rankings page should therefore be read alongside the domain weights and underlying data.
Policy inputs can be current to 2026; wage and health results lag. The Data page lists the edition actually used rather than implying every input is the latest available today.
Direct percentages are min-max normalized, legal categories receive published point values, and some outside composite rankings are linearly converted to 0–100.
If source ranks 1 through 50 are converted to evenly spaced scores, adjacent score gaps are a modeling convenience. They do not show that real-world differences are equal.
For example, the reproductive domain is 55% abortion legal environment and 15% each for shield laws, Medicaid coverage and medication-abortion access.
Each domain score is multiplied by its published weight. The weighted domains are added to produce the final state score.
Alternative rights-heavy, economics-heavy and family-support-heavy weighting scenarios show how much the result depends on the default assumptions.
A difference of one or two positions does not prove one state is meaningfully better than another. The index is most useful for identifying tiers, tradeoffs and specific areas of strength or weakness.
Data years differ. Several domains use transformed source rankings rather than underlying measurements. Housing is an indirect economic proxy. Seven domains include scores carried forward from V2 without their original source chain in the V3 workbook. The weights, category points and proxy choices are editorial judgments, not an objective universal truth.
The personalized ranking tool preserves the published model as the default while letting users put more or less emphasis on any domain. That does not erase the methodology; it makes the normative choices explicit.