Skip Navigation
Analysis

Does the Anti-Gerrymandering Campaign Threaten Minority Voting Rights?

We tested the theory — which has been put forward by both Democrats and Republicans.

October 10, 2017

When Pennsylvania Demo­crats went to the Supreme Court in 2004 to ask that Pennsylvani­a’s GOP-drawn congres­sional map be struck down as an unfair partisan gerry­mander, they drew oppos­i­tion from an unex­pec­ted source: fellow Demo­crats.

Alabama Demo­crats told the court in a brief they were concerned that ending partisan gerry­man­der­ing would “under­mine … the abil­ity of African Amer­ic­ans in Alabama to continue the effect­ive exer­cise of their newly won abil­ity to parti­cip­ate in the polit­ical process.”

In 2001, they poin­ted out, “African-Amer­ican repres­ent­at­ives pulled, hauled, and traded with their white colleagues” to achieve greater repres­ent­a­tion.

In short, polit­ical gerry­man­der­ing — in which it was taken for gran­ted that Demo­crats sought an advant­age — helped maxim­ize the voice of African Amer­ic­ans.

Flash forward 13 years. Echoes of the same argu­ment are being heard again as the Supreme Court once weighs the consti­tu­tion­al­ity of partisan gerry­man­der­ing — this time in the pivotal Gill v. Whit­ford case, involving Wiscon­sin’s polit­ical map. The oral argu­ment took place last week. (In the 2004 case, the court dead­locked badly, fail­ing to agree on where to draw the line on gerry­man­der­ing.)

Now it is the National Repub­lican Congres­sional Commit­tee that argues that ending polit­ical gerry­man­der­ing would hurt African Amer­ic­ans. “A [proposed] test for partisan gerry­man­der­ing creates conflicts between district­ing and Section 2 of the Voting Rights Act.”

Section 2 of the Voting Rights Act gives courts the power, in some circum­stances, to require the creation of major­ity-minor­ity districts, to ensure minor­ity voters have an equal chance to parti­cip­ate in the polit­ical process.

But is the claim actu­ally true? Is the goal of elim­in­at­ing partisan gerry­man­der­ing in conflict with the goal of making sure minor­ity communit­ies have an effect­ive elect­oral voice?

We looked at the issue empir­ic­ally, by analyz­ing more than two decades’ worth of elect­oral maps — from 1992 to 2006 — from 18 states that that had major­ity-minor­ity districts in this period.

Our find­ings upend conven­tional wisdom. The data categor­ic­ally demon­strates that minor­ity voting power can be protec­ted when partisan gerry­man­der­ing is forbid­den. Not only can the two goals coex­ist, but major­ity-minor­ity districts can help prevent minor­ity communit­ies from being used to maxim­ize partisan advant­age — by either Repub­lic­ans or Demo­crats.

Legis­latures created more major­ity-minor­ity districts in the 1990s, in part to comply with the Voting Rights Act

To set the stage for our analysis: Virtu­ally every­one acknow­ledges that the wave of major­ity-minor­ity districts created in the 1990s were wildly success­ful in finally giving African Amer­ic­ans and Lati­nos a mean­ing­ful say in who repres­en­ted them — often for the first time, and partic­u­larly in the South.

Before the 1990s, there were only 32 major­ity-minor­ity congres­sional districts in the United States (includ­ing just 11 in the South). Five South­ern states — Alabama, Flor­ida, North Caro­lina, South Caro­lina, and Virginia —had zero major­ity-minor­ity districts despite having very large minor­ity popu­la­tions. That changed in the 1990s as minor­ity groups, with support in some cases from Repub­lic­ans, were success­ful in using the Voting Rights Act to compel creation of major­ity-minor­ity districts as a remedy for vote dilu­tion — the spread­ing of black voters thinly across several districts to minim­ize their elect­oral effect­ive­ness.

After redis­trict­ing in the early 1990s, the number of major­ity-minor­ity congres­sional districts grew from 32 to 51, includ­ing 24 in the South. By the end of the 1990s, every South­ern state other than Arkan­sas had at least one such district.

Conven­tional wisdom has held that those major­ity-minor­ity districts had unin­ten­ded consequences, unfairly favor­ing Repub­lic­ans over Demo­crats. By pack­ing together highly reli­able blocks of Demo­cratic voters, the reas­on­ing goes, the new districts “bleached” surround­ing districts —made them whiter and more Repub­lican. And it is true that after the redis­trict­ing of the 1990s, many districts that had a history of elect­ing white Demo­crats flipped almost overnight and began elect­ing Repub­lic­ans.

Some pundits have even argued that major­ity-minor­ity districts were a signi­fic­ant factor in the 54-seat swing in House seats in the 1994 elec­tion, which gave the GOP control of the cham­ber for the first time in 42 years.

There is some truth to the conven­tional wisdom. That’s because whether you are gerry­man­der­ing to favor Repub­lic­ans or Demo­crats, minor­ity voters can be a useful focus of atten­tion. Minor­ity voters (espe­cially African Amer­ic­ans) are over­whelm­ingly Demo­cratic — help­ful know­ledge when you’re fine-tuning just how Demo­cratic any given district will tend to be.

And yes, the lack of minor­ity-major­ity districts before 1990 had been key to help­ing preserve the incum­bency of many white South­ern Demo­crats: Members of minor­ity groups were distrib­uted in a way that influ­enced elec­tions yet did not make it possible for them to elect their preferred candid­ate.

This did not escape the notice of Repub­lic­ans. Ben Gins­berg, general coun­sel to the Repub­lican National Commit­tee in 1990, said in 1995: “We began look­ing at the data, and we saw that white South­ern Demo­crats had domin­ated the redis­trict­ing process liter­ally since the Civil War, and that had created under­rep­res­ena­ta­tion for two groups— Repub­lic­ans and minor­ity voters.”

African Amer­ic­ans largely tell the same story about how white Demo­crats manip­u­lated the system. Rep. Bobby Scott, an African-Amer­ican Demo­crat first elec­ted to Congress after Virgini­a’s third congres­sional district was redrawn as a major­ity-minor­ity district, explainedto the New York­er’s Jeffrey Toobin that before 1990, the strategy for preserving arti­fi­cially large Demo­cratic major­it­ies had been “to keep African-Amer­ican percent­ages [in partic­u­lar districts] at around 35 or 40 per cent.”

But the fact that Repub­lic­ans derived some level of polit­ical bene­fit from growth in major­ity-minor­ity districts after 1990 does­n’t answer the cent­ral ques­tion. Sure, the creation of major­ity-minor­ity districts, along with the shift of South­ern white voters to the Repub­lican Party, may have helped south­ern Repub­lic­ans gain seats. But did they get more than their fair share of seats?

Look­ing at the evid­ence

To find out, we analyzed congres­sional elec­tion returns from the 1990s, 2000s, and 2010s, and meas­ured each party’s seat share using three stat­ist­ical tests used by social scient­ists to gauge the degree of partisan bias in elect­oral maps. Those tests are the “effi­ciency gap” (subject of much publi­city recently), the “seats-to-vote curve,” and the “mean-median differ­ence. The tests vary in their approach, but all meas­ure the degree to which a polit­ical party is able to trans­late votes into seat share. (If one party can consist­ently win a major­ity of seats despite attract­ing a minor­ity of votes, that can in effect lock the other party out of power.)

If a map scores high in partisan bias under any of these meas­ures, it’s a strong signal of partisan gerry­man­der­ing; if it scored high under all three, the signal is even stronger. Although some justices appeared skep­tical of these tests last week — Chief Justice John Roberts referred to “soci­olo­gical gobbledy­gook” — they hold out the hope of making the iden­ti­fic­a­tion of gerry­man­der­ing a more scientific process than it’s been.

We found little to no link between partisan gerry­man­der­ing and major­ity-minor­ity districts — and the evid­ence from the South was espe­cially compel­ling.

Although major­ity-minor­ity districts in the South more than doubled in the 1990s, South­ern states displayed low, and in most cases negli­gible, rates of partisan bias in their congres­sional maps. The one excep­tion was Texas, where congres­sional maps exhib­ited durably high rates of bias in favor of Demo­crats as a result of artful “crack­ing” of white Repub­lican suburban voters. (They were split among vari­ous Demo­cratic-lean­ing districts.) The chart below shows partisan bias in the South, from 1992 through 2000, using the effi­ciency gap:

The same trend can be seen in the congres­sional maps of non-South­ern states with major­ity-minor­ity districts (again, using the effi­ciency gap).

Only Cali­for­nia, Illinois, and New York had any degree of signi­fic­ant pro-GOP bias. All three only developed the high bias late in the decade. It is hard to attrib­ute the emer­gence of that bias to major­ity-minor­ity districts, however. Major­ity-minor­ity districts in the three states all were in heav­ily Demo­cratic areas, such as New York City, Chicago, and Los Angeles, where surround­ing districts contin­ued to elect white Demo­crats by safe and comfort­able margins.

Our study found that the same lack of a connec­tion between major­ity-minor­ity districts and partisan bias in the 2000s. The find­ing held, as well, in both South­ern and non-South­ern states.

But what about today? The answer again, for this decade, is no.

In fact, strik­ingly, the maps of this decade reveal that Repub­lic­ans have been using the same kind of slicing and dicing of minor­ity voters once prac­ticed by white Demo­crats to engin­eer a lopsided advant­age for their interests. The creation of more major­ity-minor­ity districts helped elim­in­ate arti­fi­cial pro-white-Demo­cratic bias in the 1990s. In the same way, major­ity-minor­ity districts might combat pro-Repub­lican bias in states like Texas today.

In short, there is simply no conflict between prevent­ing outrageous partisan abuses during redis­trict­ing and ensur­ing fair treat­ment for minor­ity communit­ies. We can do both — and we should.