Between 1970 and the start of 2025, only two U.S. states voluntarily redrew their congressional district maps between decennial censuses. That trend changed in the second half of 2025, however, as mid-decade redistricting battles unfolded nationwide. Ahead of the 2026 midterm elections, several states have rushed to enact new district plans to gain a partisan advantage. In the past few months alone, seven states have voluntarily proposed or adopted new maps.
The Freedom to Vote Act (FTVA) offers an antidote to this rise in unbridled partisan gerrymandering. Besides banning mid-decade redistricting, the proposed bill sets a uniform baseline for measuring partisan advantage in redistricting proposals, outlines when and how rules are violated, and crafts remedies for these violations. This analysis applies the FTVA’s test for partisan bias to analyze various plans used in one of the most contested states in redistricting: North Carolina. The results show that the FTVA test successfully identifies potentially partisan gerrymandered congressional plans — and that it can flag maps tainted by racial gerrymandering in some instances.
North Carolina has been deemed a “consistent center[] of intense partisan conflict over redistricting.” The state has been a party in multiple redistricting cases before the U.S. Supreme Court, including Shaw v. Reno in 1993, which overturned the state’s 1992 congressional maps for an overreliance on race when drawing districts. North Carolina has a lot of experience drawing congressional districts since the 2010 census. Since then, no congressional district plan has been used for more than two successive elections. The state legislature draws the maps, but the courts are actively engaged in reviewing redistricting plans.
In 2012, the state redrew congressional districts in advance of the 2016 election in response to a federal court decision finding that two districts were drawn predominately on the basis of race, which was affirmed by the U.S. Supreme Court in Cooper v. Harris. The 2016 districts were then vacated by a state court in 2019 to remove the partisan advantage in that remedial plan. The same year, North Carolina’s maps were again before the Court in Rucho v. Common Cause, which declared federal partisan gerrymandering challenges nonjusticiable.
The North Carolina legislature next drew a district plan for the 2020 elections. After the 2020 census, legislators were unable to draw a plan that satisfied state constitutional standards, so the state supreme court appointed two special masters to craft districts for the 2022 elections. This map was then replaced by the legislature in 2023, after a new majority on the state supreme court reversed its findings and followed Rucho, ruling that partisan gerrymandering was not reviewable by state courts. In October 2025, the legislature voluntarily adopted a new map ahead of the 2026 elections.
A recent academic study considering North Carolina election data from 2006 to 2024 shows that living in a competitive district increases election turnout. However, research by the Brennan Center and others has demonstrated that, compared with plans drawn by courts and independent commissions, legislatures are least likely to draw competitive districts — and that legislature-drawn districts are least likely to have competitive elections.
The federal FTVA would block anti-voter efforts in states; it came close to passage in 2022 but is not currently law. Section 5003(c)(3) of the FTVA sets out a test for identifying partisan gerrymandered congressional district plans. A map is entitled to a rebuttable presumption of partisan bias if one party has an advantage beyond a specific threshold, which uses the number of U.S. House seats apportioned to each state compared with seats won in the two previous presidential and Senate elections. In states with 14 or fewer seats, like North Carolina, the threshold requires an advantage of only 1 seat. If a district plan triggers the rebuttable presumption of bias spelled out in the FTVA, then the state may defend its district plan in court. However, if the judge finds the state’s explanation inadequate, then that judge can block the use of the district plan.
We have used the FTVA test in the past to assess district plans for evidence of partisan gerrymandering. To apply this test to the congressional district plans enacted in North Carolina since 2012, we use elections results — including data from the 2016 to 2020 presidential and Senate elections — collected by the University of Florida and Wichita State University’s Voting and Election Science Team. Because North Carolina has 14 congressional seats (the state had 13 seats following the 2010 apportionment), the threshold for triggering a rebuttable presumption of bias is 1 seat.
Table 1 reports the FTVA test results of the state’s six enacted plans since 2011. For the sake of clarity, we measure the results for each of the four reference elections (i.e., the 2016 and 2020 presidential elections and the 2018 and 2020 Senate elections) and report an average.
The outcomes are clear: Every district plan adopted by North Carolina’s legislature since the 2010 census would trigger the rebuttable presumption of partisan gerrymandering. The only plan that would not is the map adopted for the 2022 elections — the one drawn by court-appointed special masters in 2021. Had the FTVA been enacted in 2022, North Carolina would still be using this neutral district plan; the subsequent partisan gerrymanders enacted by the legislature would have been reviewed by a judge and likely blocked.
That the FTVA flags the 2012 plan (which was also used in 2014) as a partisan gerrymander is notable because a federal court ordered changes to this plan on racial gerrymandering grounds, a decision that the Supreme Court upheld in Cooper. If the FTVA had been in force after the 2010 census, then presumably a fair map would have replaced the districts used in 2012 and 2014 — and the whole of North Carolina’s struggle to have a partisan legislative majority draw a neutral map could have been avoided.
This analysis recognizes that applying data from elections after the adoption of the 2011 plan is contrary to the text of the FTVA, which calls for election data from the two previous presidential and Senate elections. When we use an alternative precinct-level dataset compiled by a team based at Harvard and Stanford Universities instead — which includes data from the 2004 and 2008 presidential elections and the 2004, 2008, and 2010 Senate elections — the 2011 plan still triggers the FTVA test’s presumption of partisan gerrymandering. The average seat bias is also very close to the result reported above based on the 2016–20 dataset: a Republican advantage of 2.68 seats.
To be clear, how the FTVA test would assess the plan adopted for the 1992 election (which the Supreme Court rejected in Shaw) is unknown without a dataset of Senate elections from the 1990s. That said, our results here show that the FTVA identifies a plan in North Carolina that is biased with regard to both partisanship and race.
