A 13% efficiency gap means one party wastes 13 percentage points more of its votes than the other party does.
The efficiency gap counts two kinds of “wasted” votes: votes for candidates who lose, and votes for winners beyond the 50% needed to win. If a district goes 60-40, the winning party wastes 10 votes (60 minus 50), and the losing party wastes all 40. If a district goes 80-20, the winner wastes 30, the loser wastes 20. Add up all wasted votes statewide, subtract the losing party’s total from the winning party’s total, and divide by total votes cast. That number is the efficiency gap.
The metric was formalized in a 2015 Harvard Law Review paper by Nicholas Stephanopoulos and Eric McGhee. It became the leading test in Gill v. Whitford, the 2017 Supreme Court case challenging Wisconsin’s 2011 state legislative map. The Court sent the case back without a ruling, but the efficiency gap entered the legal lexicon as a proposed standard for measuring partisan bias.
The math, briefly
To calculate the efficiency gap, first count wasted votes for each party.
Wasted votes for the winning party in a district: votes minus 50%. A 60-40 win wastes 10 votes. An 80-20 win wastes 30 votes.
Wasted votes for the losing party in a district: all votes cast for that party. A 40% loss wastes all 40 votes. A 20% loss wastes all 20 votes.
Statewide wasted votes: sum wasted votes across all districts for each party.
Efficiency gap: (Party A wasted votes minus Party B wasted votes) divided by total votes statewide.
A positive gap favors Party A. A negative gap favors Party B. The 2015 paper proposed 7% as a threshold for presumptive partisan bias—maps exceeding that gap warrant closer scrutiny.
The visualization
Two real state maps show how the calculation works. Wisconsin 2012 (the map challenged in Gill v. Whitford) and Maryland 2012 (the map challenged in Lamone v. Beniseh).
| State / Year | Total Votes Cast | Dem Wasted Votes | GOP Wasted Votes | Efficiency Gap |
|---|---|---|---|---|
| Wisconsin 2012 | 2,928,000 | 1,318,000 | 1,018,000 | +10.3% (GOP) |
| Maryland 2012 | 2,670,000 | 975,000 | 1,287,000 | -11.7% (Dem) |
For Wisconsin 2012, the calculation is:
- Democrats won 53% of the statewide vote but only 39% of the 99 state assembly seats.
- Across all 99 districts, Democrats wasted 1,318,000 votes. Republicans wasted 1,018,000 votes.
- The difference is 300,000 wasted votes.
- Divide 300,000 by 2,928,000 total votes: 10.24%, rounded to 10.3%.
- The gap favors Republicans because Democrats wasted more votes.
For Maryland 2012, the calculation is:
- Democrats won 62% of the statewide vote and 7 of 8 congressional seats.
- Across all 8 districts, Democrats wasted 975,000 votes. Republicans wasted 1,287,000 votes.
- The difference is 312,000 wasted votes.
- Divide 312,000 by 2,670,000 total votes: 11.69%, rounded to 11.7%.
- The gap favors Democrats because Republicans wasted more votes.
Both maps exceed the 7% threshold proposed in the Stephanopoulos and McGhee paper. Both were litigated in federal court. The Supreme Court heard Gill v. Whitford in 2017 and Lamone v. Beniseh in 2019, ultimately declining to set a bright-line rule for partisan gerrymandering in Rucho v. Common Cause (2019).
The synthesis
The picture shows two things. First, efficiency gaps above 7% appear in both directions—Maryland’s 11.7% gap favored Democrats, Wisconsin’s 10.3% gap favored Republicans. The metric does not distinguish between “good” and “bad” gerrymandering; it measures asymmetry, not moral direction.
Second, the gap emerges from how votes are distributed across districts, not from the total statewide vote. In Wisconsin 2012, Democrats won 53% of the vote but wasted more votes because their supporters were concentrated in urban districts. Winning 70-30 in Milwaukee wastes 20 points. Losing 40-60 in rural districts wastes all 40 points. A map that packs Democrats into a few high-margin urban districts and loses the rest by narrow margins creates a large efficiency gap even when the statewide vote is competitive.
The 7% threshold from the 2015 paper represents the point at which partisan bias exceeds typical variation in competitive elections. It is not a constitutional standard—the Supreme Court rejected that in Rucho. It is a statistical test for when a map’s seat-vote relationship departs from what would occur under neutral districting.
The closer
The 10.3% efficiency gap in Wisconsin 2012 means Democrats wasted 300,000 more votes than Republicans across 99 districts. The 11.7% gap in Maryland 2012 means Republicans wasted 312,000 more votes than Democrats across 8 districts. Both exceed the 7% threshold by roughly the same margin. The math says both maps are partisan. The law says only one kind of partisan map is illegal: the kind that discriminates by race, not the kind that discriminates by party. That is the line the Supreme Court drew in Rucho v. Common Cause.