Redistricting is a partisan activity. Democrats want to elect more Democrats, and Republicans want to elect more Republicans. Incumbents want to be re-elected, and good government advocates often want more competitive elections. But much of the public debate surrounding redistricting focuses not on partisanship, but on race. Why is this?
The main reason so much of redistricting battle centers around race is that a federal law regulates how New Jersey can draw its legislative and congressional districts. That law is the Voting Rights Act of 1965, or the VRA for short. The VRA includes many sections and provisions, but the part that is relevant to New Jersey’s redistricting fight is Section 2. Section 2 prohibits discriminatory voting practices, including vote dilution. The leading Supreme Court case on Section 2, Thornburg v. Gingles, set out a three-part test to determine whether the method (usually a legislative district map of some kind) used by a state, county, or local government in electing representatives to a governing body dilutes minority votes:
1. Is the minority group sufficiently large such that a geographically compact, single-member minority-majority district is possible?
2. Is the minority group politically cohesive?
3. Does racial bloc voting by the white majority enable it to defeat the minority group’s preferred candidate?
If the answer to all three of these questions is yes, then the vote dilution claim succeeds and the method or map in question is illegal. If the answer to one of these questions is no, the vote dilution claim fails and the method or map in question is legal.
Below the fold, I explore further the Gingles vote dilution standard and attempt to explain the impact Section 2 might have on redistricting at the state, county, and even local level in New Jersey. I make no normative arguments in this diary because I believe that, to understand the debate surrounding redistricting, it is important to understand the law governing it. The threat of litigation under the VRA affects line-drawing strategies of both Democrats and Republicans in redistricting. Politicians, leaders of minority groups, pundits, tea partiers, and other interested parties also deploy the VRA as a rhetorical weapon in the battle over redistricting. In doing so, some may overstate or understate its power to serve their own political goals. But who actually has a bullet in the chamber, and who is merely making empty legal assertions? Follow me below the fold to find out.
The first prong of the Gingles test is based on single-member districts. This means that, for State Assembly, this means that if a minority-majority geographically compact district with a population of 109,899 is possible, a claim of vote dilution depends only on whether the minority group is politically cohesive and whether racial bloc voting by the majority allows it to defeat the minority group’s preferred candidate. This is the case even though we will elect two Assembly members for each 220,000-person state legislative district. In some cases, the VRA may require a multiple member legislative district to be split. For example, in 2005, A United States District Court in South Dakota, finding that Native Americans votes were diluted in a two-member State House of Representatives district, ordered the district split into two single-member districts.
This hypothetical single-member district, which follows the Raritan river from Perth Amboy to new Brunswick, has a voting-age population that is over 60% Latino.
One idea floated by Latino groups in last month’s redistricting hearings was to combine New Brunswick and Perth Amboy in one district, increasing the Latino population in the district containing Perth Amboy from 31% to 36%. The commission might decide that this is a good idea and choose to draw such a district. But, if it did not, could Latino groups challenge the map in court and prevail on the grounds that it dilutes the Latino vote in Middlesex County? At first blush, a Section 2 challenge seems hopeless, because Latinos would not constitute a majority of the voting age population in the proposed district. However, it is possible to draw a hypothetical single-member Assembly district including Latino neighborhoods in Perth Amboy and New Brunswick in which Latinos exceed 60% of the voting-age population.
You can’t add different minority groups together to reach the 50% of voting age population threshold. This was the United State’s Supreme Court’s holding in the 2009 case Bartlett v. Strickland. The court’s decision does not say that coalition districts “violate” the Voting Rights Act. It does not mean, as some Republicans would have you think, that New Jersey’s current map (which contains several coalition districts) is illegal, or that the federal court decision which upheld it is no longer good law. It means simply this: a “coalition”-majority district (e.g. 30% black, 30% Latino, 40% white) cannot be the basis of a vote dilution claim.
Bartlett does limit the regions of New Jersey where Section 2 of the voting rights act will come into play. While it is possible to draw coalition-majority districts based around Atlantic, Cumberland, Camden, Burlington, Mercer, Union, Middlesex, Somerset, and Bergen Counties, minority-majority single-member districts as defined in Bartlett are only possible in the following parts of the state:
Section 2 does not require that a minority minority district be drawn even if a geographically compact one is possible. While the possibility of geographically compact minority-majority district is necessary for a successful claim of vote dilution, it is not sufficient. The minority group must also be “politically cohesive” and must demonstrate that the “white majority votes sufficiently as a bloc to enable it … usually to defeat the minority’s preferred candidate.” If a minority group is able to elect candidates of its choice in districts where it does not enjoy a voting-age majority, it cannot prevail on a claim of vote dilution.
In the last redistricting cycle, Democrats “unpacked” black voters in North Jersey in an effort to win more seats in that part of the state. Democratic map reduced the number of black and Hispanic majority districts, but increased the number of districts with either a plurality or significant minority of voters in one group or the other. Republican lawmakers joined with several black and Hispanic voters in Essex and Hudson County to bring a vote dilution challenge to the map in federal court. The District Court found that the legislative map did not dilute minority votes because minority candidates often enjoyed significant support from white voters in Essex and Hudson Counties. The court’s reasoning was vindicated in the June primary and November general elections, as minority candidates secured a record number of seats in the state legislature.
The less politically cohesive a minority group is, the harder it will be for that group to demonstrate vote dilution. We have already seen that a section 2 claim can fail because white voters cross over to support minority candidates. It can also fail because a minority group is not “politically cohesive”.
Just over half of the voting-age population is asian in this hypothetical single-member district, which is centered on Edison Township and includes parts of Woodbridge, Piscataway, and South Plainfield.
The Asian-American community in northern Middlesex County has grown fast enough over the past decade that a reasonably compact majority Asian single-member Assembly district is now possible. Therefore, Asian-Americans in that part of the state meet the first prong of the Gingles test. But they will have more trouble second and third prongs. Asian-Americans are both politically and culturally diverse. They don’t share a common history like African-Americans, and they don’t share a language like Latinos. Data from recent legislative and county elections seems unlikely provide much help in establishing either Asian political cohesiveness or white bloc voting. It seems likely that Asian-Americans will have to rely on voting patterns in local elections, particularly in Edison Township, if they are to have any chance of presenting a successful vote dilution argument.
That a minority group is underrepresented in a legislative body does not by itself constitute a violation of Section 2. Section 2 is not meant to ensure that a particular percentage of representatives belong to a certain minority group. The purpose of section 2 is to ensure racial minorities have the same opportunity as “other members of the electorate to participate in the political process and to elect representatives of their choice.” (emphasis added) Often, black voters will vote for a black candidate, and Hispanic voters will choose a Hispanic candidate. But sometimes, minority voters will choose the non-minority candidate, as they did in the 2007 Democratic State Senate Primary in the 33rd Legislative District. In that high-turnout race, incumbent Brian Stack beat challenger Silverio Vega 77-23. Stack carried Union City, the state’s most heavily Latino municipality (and where he also happens to be mayor), with an astonishing 91% of the vote.
Section 2 and Thornburg v. Gingles apply not only in state and congressional redistricting, but at the county and municipal level as well. That means that counties and cities must draw freeholder district and ward lines to comply with Section 2. The size of the single-member district is determined by the number of representatives in the entire body, not just the number that are elected on a district-by-district basis. Therefore, a single member district would have a population of just over 30,000 in Atlantic County, about 68,000 in Hudson County, and about 87,000 in Essex County.
A Section 2 challenge can succeed against a system that uses at-large elections. This is really a corollary of the single-member district rule. A town (or county) that elects five representatives to its governing body in at large elections essentially has one five-member electoral district. A single-member district would have one-fifth of the population of the larger district. If a single-member minority-majority district is possible and minorities are not afforded the opportunity to elect candidates of their choice, a vote-dilution challenge to an at-large system should succeed. In the last decade, the federal government has brought successful Section 2 challenges to at-large systems in counties and municipalities in Florida, South Carolina, Montana, Ohio, and New York.
New Jersey counties and municipalities with politically underrepresented minority populations could see their at-large freeholder or council election systems challenged in federal court. Black, Latino, or Asian majority districts are possible in Cumberland, Camden, Mercer, Union, Middlesex, and Passaic counties, all of which elect freeholders at-large. Challenges probably stand the greatest chance of success in counties municipalities with growing, and underrepresented Latino populations. Passaic County, Clifton Township, and the City of New Brunswick come to mind. Asian voters might bring a successful challenge in Edison Township, which elects all seven members of its township council at-large. Asians constitute over two-fifths of the voting-age population in Edison, but only one Asian sits on the council. Any racial divisions that may have characterized recent local elections would strengthen their case.