Skip Navigation

Extreme Maps

Key Fact: In the 26 states that account for 85 percent of congressional districts, Republicans derive a net benefit of at least 16 to 17 congressional seats in the current Congress from partisan bias.

Published: May 9, 2017

Using data from the 2012, 2014, and 2016 elec­tion cycles, Extreme Maps finds that partisan bias result­ing largely from the worst gerry­man­der­ing abuses in just a few battle­ground states provides Repub­lic­ans a durable advant­age of 16–17 seats in the current Congress, repres­ent­ing a signi­fic­ant portion of the 24 seats Demo­crats would need to gain control of the House in 2020. These “extreme maps” were all drawn in states under single-party control; the report finds that conversely, maps drawn by inde­pend­ent commis­sions, courts, or split-party state govern­ments had signi­fic­antly less partisan bias in their maps. 


Every decade, states redraw congres­sional maps after the decen­nial census. Redis­trict­ing allows districts to be rebal­anced, ensur­ing in theory that all districts are both equally popu­lated and repres­ent­at­ive. But redis­trict­ing also provides an enorm­ous oppor­tun­ity for politi­cians: the chance to redraw a district map means the oppor­tun­ity to gerry­mander and to manip­u­late a map to create a more favor­able set of districts for them­selves and for their party.

Congres­sional maps were last redrawn en masse after the 2010 Census, and accus­a­tions of gerry­man­der­ing in states nation­wide soon followed. Complaints about redis­trict­ing abuses ran the gamut from alleg­a­tions that some maps had been drawn to favor incum­bents to outrage at the sprawl­ing and unnat­ural shapes of districts in others.

This report focuses on one of the most egre­gious of these abuses: the manip­u­la­tion of district lines to give the party draw­ing the map a share of seats grossly at odds with statewide elec­tion results, thus ensur­ing that one party is overrep­res­en­ted and the other under­rep­res­en­ted in a deleg­a­tion.

To gauge where this type of gerry­man­der­ing is taking place and its magnitude, this report used elec­tion results in states with six or more congres­sional districts to assess the extent and the durab­il­ity of “partisan bias” — the degree of system­atic advant­age one party receives over another in turn­ing votes into seats. For this analysis, this report used multiple quant­it­at­ive meas­ures of partisan bias to exam­ine the 2012, 2014, and 2016 congres­sional elec­tions. It also looked at the rela­tion­ship between the body that drew the maps and the degree of bias observed. It is among the first analyses to use 2016 elect­oral data to exam­ine maps, and the first report of its kind to meas­ure maps using multiple meas­ures of bias and to identify the hand­ful of single-party controlled states that are respons­ible for nearly all of the bias in this decade’s maps.

Our key find­ings include:

This decade’s congres­sional maps are consist­ently biased in favor of Repub­lic­ans.

  • In the 26 states that account for 85 percent of congres­sional districts, Repub­lic­ans derive a net bene­fit of at least 16–17 congres­sional seats in the current Congress from partisan bias. This advant­age repres­ents a signi­fic­ant portion of the 24 seats Demo­crats would need to pick up to regain control of the U.S. House of Repres­ent­at­ives in 2018.

Just seven states account for almost all of the bias.

  • Michigan, North Caro­lina, and Pennsylvania consist­ently have the most extreme levels of partisan bias. Collect­ively, the distor­tion in their maps has accoun­ted for seven to ten extra Repub­lican seats in each of the three elec­tions since the 2011 redis­trict­ing, amount­ing to one-third to one-half of the total partisan bias across the states we analyzed.
  • Flor­ida, Ohio, Texas, and Virginia have less severe partisan bias but jointly account for most of the remain­ing net extra Repub­lican seats in the examined states.

Single-party control of the redis­trict­ing process is closely linked with biased maps.

  • The seven states with high levels of partisan bias are all states where one polit­ical party had sole control of the redis­trict­ing process. Court-ordered modi­fic­a­tions to maps in Flor­ida, Texas, and Virginia — all origin­ally drawn under sole Repub­lican control — have reduced but not entirely curbed these states’ partisan bias.
  • States where Demo­crats had sole control of redis­trict­ing have high partisan bias within state congres­sional deleg­a­tions, but the relat­ively small number of districts in these states creates a much smal­ler effect on partisan bias in the House over­all.
  • By contrast, maps drawn by commis­sions, courts, and split-control state govern­ments exhib­ited much lower levels of partisan bias, and none had high levels of bias persist­ing across all three of the elec­tions since the 2011 round of redis­trict­ing.

There is strong evid­ence that the bias in this decade’s congres­sional maps is not acci­dental. With the excep­tion of Texas, all of the most biased maps are in battle­ground states. These states routinely have close statewide elec­tions and a fairly even distri­bu­tion of partis­an­ship across most of the state — two factors that do not natur­ally suggest that there should be a large and durable under­rep­res­ent­a­tion of one polit­ical party.