Posted by Rob Richie, Drew Penrose, David O'brien, Author Maria Perez on June 23, 2017
This is the second part in a multi-part series on the projected impact of the Fair Representation Act on voting rights and representation for people of color.
The United States continues to fall short in fairly reflecting its diversity in elected offices. Our widespread use of a winner-take-all rule -- that is, one where the biggest group of voters can win 100% of representation -- presents a challenge to candidates who are not the preferred choice of a majority of voters. The problem of under-representation is less acute in the U.S. House than statewide races, but is still significant.
With the upcoming launch of the Fair Representation Act, legislation in Congress to end winner-take-all elections, in the coming week we will examine the state of representation of people of color in the current Congress and how it might change with enactment of the Fair Representation Act. The first entry in this series discussed African American representation in the South. This installment looks at the Fair Representation Act’s impact on Latino representation in five states in the American Southwest.
Current Demographics and Representation
As of the 2010 Census, just over half (53.6%) of the United States’ Latino population lived in five states: California, Nevada, Arizona, New Mexico, and Texas. Collectively, these states send 105 representatives to the House. The citizen voting age population ("CVAP") of these states is roughly one-third Latino (35.4%), so one would reasonably expect their House delegations to reflect that. In reality, however, Latino voters are consistently underrepresented across the region. Of the region’s 105 representatives, only 22 are Latino (21%).
Consider Texas. The most recent census found that Latinos made up 33.6% of its CVAP, but only nine of the state’s 36 congressional districts (25%) are majority Latino. (A federal court has found that three of Texas’ congressional districts were gerrymandered to dilute Latino votes, but litigation continues). Latinos hold five of these nine districts, along with one other seat, putting them at just 16.7% of Texas’ House delegation. Only 49% of voting age Texas Latinos live in a district where the Latino CVAP is large enough to meet the threshold to elect their preferred candidate.
In New Mexico, only one of the state’s three districts (33.3%) has a Latino plurality, even though Latinos make up 42.3% of the state’s CVAP. Just 36.5% of New Mexico’s voting age Latinos live in that district.
In Arizona, where 28% of the state’s CVAP is Latino, two of the state’s nine districts (22.2%) are majority or plurality Latino, and are home to 46.3% of the Arizona’s Latinos of voting age. Nevada, whose voting age population is 22.3% Latino, has no majority or plurality Latino districts.
The current congressional delegations of these states is more representative of their demographics, which shows the potential for Latino candidates having crossover appeal in this region. Two of New Mexico’s three representatives are Latino, as is one of Nevada’s four representatives. Both of Arizona’s two majority Latino districts are represented by Latinos (although no other districts in the state are). In addition to those five House seats, these states have two Republican governors who are Latino, in Nevada and New Mexico, and a Democratic U.S. Senator in Nevada.
In California, however, Latinos make up 33.1% of its voting age population, but are the majority or a plurality in 16 of its 53 congressional districts (30.2%) and hold 11 seats (20.8%). (That number increases to 14 if counting the three Portuguese-Americans who self-identify as Hispanic, bringing the proportion of representation to 26.4%. This would increase the region’s total to 25 Latino/Hispanic representatives, or 23.8% of the representatives from the region). Unlike Texas, any discrepancies between California’s population and its congressional representation cannot be blamed on a state legislature seeking partisan advantage. Its districts are drawn by the California Citizens Redistricting Commission, an independent body explicitly charged with drawing districts consistent with the Voting Rights Act of 1965 and forbidden from drawing districts with the purpose of favoring or disfavoring particular parties or candidates.
California’s decision to remove the legislature from the redistricting process is an important and welcome step towards fair representation but, as the numbers show, it alone is not enough.
The Fair Representation Act
Since congressional elections are winner-take-all, any demographic group that does not have a majority or significant plurality of a district’s CVAP is unlikely to elect the candidate of its choice, and will see its own political influence limited accordingly. The Fair Representation Act eliminates this problem through its use of ranked choice voting (“RCV”) and multi-winner districts. Ranked choice voting is a fair representation voting method, not winner-take-all. That means that a majority group can elect a majority of seats, but not all seats. Over 200 jurisdictions in the United States use a fair representation voting method, and in those places, Latino candidates can and do win in at-large elections, even in white majority jurisdictions with racially polarized voting.
In a fair representation voting method like ranked choice voting, any group of voters above the election threshold - less than a majority - has the power to elect a candidate of choice with their votes alone. The threshold goes down as the number of winners in a district goes up, which is why the Fair Representation Act prioritizes the use of 5-winner districts, making the threshold about 17% of the vote.
Under the FRA, Texas would go from 36 single member districts to eight multi-winner districts - six districts with five seats each and two with three seats. In one plan generated by Auto-Redistrict, each district has a Latino CVAP over the threshold necessary to elect at least one representative, and in some of these districts the Latino community would be able to elect two or three candidates of choice.
In this plan, the number of candidates Texas’ Latino population would be able to elect would go from 9 to 13, increasing its representation in the state’s House delegation from 25% to 36.1%, an amount much more representative of the Latino share of Texas’ voting age population (33.6%). Currently, 49% of Latino Texans live in a district where the Latino CVAP is above the threshold to elect a candidate. Under this plan, that would increase to 100%.
The number of representatives that California’s Latino community would be at or near the CVAP threshold to elect would increase to 19 of 53 (35.9%), significantly more proportional to the state’s voting age population, which is 38.9% Latino. The proportion of Latinos in California living in a district where their CVAP is above the election threshold would increase from 49% to 100%.
In Nevada, the number of representatives the state’s Latino population would have the CVAP to elect would increase from zero to one of four (25%), bringing it much closer to its Latino voting age population (22.3%). The proportion of Latinos in Nevada living in a district where their CVAP is above the election threshold would increase from 0% to 100%.
Under these projections, the number of representatives Arizona and New Mexico’s Latino populations would have the CVAP threshold to elect a candidate of choice would not change, but this assumes those state’s demographics remain unchanged between the 2010 and 2020 Censuses. An increase in those states’ Latino populations in the interim could easily change this. In New Mexico, for example, Latinos were 42.3% of the voting age population in 2010. If New Mexico’s Latino CVAP is greater than 50% in 2020, Latino voters will have the power to directly elect two representatives of their choice.
The proportion of Latino voters living in a district where their CVAP is above the election threshold in New Mexico and Arizona would change, from 36.5% to 100% in New Mexico, and from 46.3% to 81.2% in Arizona. Arizona is the only southwestern state where these simulated maps include even a single multi-winner district where Latinos are below the threshold to elect a candidate of choice with their votes alone, but even in that district, Latinos are more than half of the 25% threshold for election, giving them substantial influence in the district.
Factoring in Turnout Differences
Latino voters will see an increase in their voting power and representation even when controlling for the possibility of low voter turnout. Historically, turnout among Latino voters has often been lower than other ethnic groups, especially in this region. On the one hand, there is good reason to believe that Latino turnout would be more equitable under the new system. When Amarillo (TX) adopted a fair representation voting method for its educational board elections, turnout rose, and the first Latino representatives were elected to its board of regents and school board. Other stories from the more than 200 jurisdictions using fair representation demonstrate how turnout often increases when voters have more power.
Even assuming Latino voter turnout remains static, however, the number of representatives that Latino voters would have the power to elect would increase to 15 in California and 10 in Texas. Nationwide, there are currently 37 majority-Latino congressional districts, which the FRA would increase to 52. Accounting for voter turnout, 31 of the current majority-Latino districts have sufficiently-high Latino voter turnout to elect a candidate of choice, which would increase to 36 under the FRA. Even if there is no change in Latino voter turnout, Latino voters would have the power to elect five more representatives than they do now, and more as the Latino share of CVAP increases in these states.
Latino Representation Outside the Southwest
The FRA’s impact on Latino representation would extend beyond the southwest. In Florida, the number of districts where Latino CVAP is at or near the threshold to elect a candidate would increase from three to five. (These projections include districts where Latino voters are a plurality of the threshold necessary to elect a representative). In New Jersey, the number of districts at or near the threshold would increase from one to two. In New York, districts at or near the threshold would go from two to three and in Illinois from one to three. For the first time, Georgia, Colorado, and Connecticut would all have at least one district where Latino CVAP is at or near the level necessary to directly elect a representative when turnout was equitable.
Nationally, the FRA would increase the number of Latinos living in a district where their CVAP is high enough to select a representative from 36% of the Latino population to 74%, the largest increase of any ethnic group. As with every group, influence would increase even in places where Latinos are below the threshold to elect. Additionally, with ranked choice voting, there would be more opportunities for coalitions between the diverse ethnicities that collectively are categorized as “Hispanic” by the Census, and between Latinos and other racial groups. That is, even when Latino-backed candidates were below the victory threshold and lost, other non-Latino candidates very likely would seek second and third choice support from these voters - a unique feature of ranked choice voting compared to other fair representation systems.