What metrics identify partisan gerrymandering like efficiency gap and mean-median?

1. What the headline metrics measure and why reporters use them

Metrics like the efficiency gap, mean–median difference, partisan bias and declination quantify different kinds of vote-seat asymmetry rather than trying to prove intent: the efficiency gap tallies “wasted” votes to estimate advantage, the mean–median compares the average district vote to the median district vote to show skew, partisan bias simulates a uniform swing to see which party would win at a neutral vote share, and declination looks for threshold‑related asymmetry around the win/loss line; PlanScore explicitly lists these four as its core measures and explains when each is more reliable ^[1].

2. How analysts combine metrics into a single judgment

Institutions such as Princeton’s Gerrymandering Project and PlanScore do not rely on a lone number; they score maps across multiple metrics and often supplement them with ensemble simulations or historical comparisons to produce a holistic grade or “report card” for partisan fairness and competitiveness ^{[2] [1]}. The Fulcrum describes Princeton’s Redistricting Report Card as converting partisan and racial performance data into grades — an approach that blends metrics with policy‑oriented thresholds ^{[3] [4]}.

3. Strengths and limits of the key measures

These metrics are powerful because they turn qualitative accusations into measurable asymmetry, but each has limits: PlanScore cautions that all four work well in competitive states but that the efficiency gap and declination are preferable when one party overwhelmingly dominates the electorate ^[1]. Princeton’s tools therefore treat metrics as part of a “cutting‑edge” algorithm and contextual review rather than as standalone proof of illegal gerrymandering ^[2].

4. How courts and litigants use (or don’t use) metrics

Metrics are influential in public debate and advocacy (used in media, tool outputs and litigation support), but the provided reporting shows courts still adjudicate on mixed grounds including racial‑gerrymandering claims and statutory standards; for example, a federal panel blocking Texas’s 2025 map cited substantial evidence of racial gerrymandering alongside political motives ^[5]. Available sources do not detail a single metric that courts uniformly accept as dispositive (not found in current reporting).

5. Practical tools journalists and citizens can use now

Free public tools cited in reporting — PlanScore and Princeton’s Redistricting Report Card — let users input or review maps and get instant scores on efficiency gap, mean‑median, partisan bias and declination, plus contextual grades and historical comparisons ^{[1] [2]}. The Fulcrum and similar outlets have also promoted these tools as ways for non‑experts to “identify partisan gerrymandering” and track mid‑decade map fights ^[4].

6. Political context: why numbers matter this decade

The stakes for these metrics rose in 2025 as multiple states pursued mid‑decade redistricting and lawsuits proliferated; reporting on Texas and other states shows both partisan and racial claims in play, and advocacy groups and journalists used metric‑driven analyses to label maps “egregious” or to grade them with failing marks ^{[5] [6]}. Democracy Docket and other trackers also frame seat‑vote disparities (percent seats vs. presidential vote share) as an intuitive way to show how maps can entrench advantage ^[7].

7. What readers should watch for when you see a single number

A single metric can be a useful signal but not a full verdict: trustworthy analyses report multiple metrics, explain which are robust given local competitiveness, and place numbers alongside simulations or historical baselines — exactly the practice Princeton and PlanScore describe in their public products ^{[2] [1]}. When outlets or campaigns highlight only one favorable number, ask whether other metrics or ensemble comparisons were run (not found in current reporting about that specific practice).

8. Bottom line for journalists and voters

Use the efficiency gap, mean–median, partisan bias and declination as the starting lexicon for assessing partisan skew, but demand a multi‑metric, comparative approach — the same approach PlanScore and Princeton promote — and remember courts treat numerical evidence alongside legal standards and racial‑fairness claims, not as automatically decisive ^{[1] [2] [5]}.