The Supreme Court surprised many observers of voting rights and immigrant rights this week by deciding to take up a case (Evenwel v. Abbott), in which plaintiffs argue that their votes are diluted—by counting less than the votes of citizens in other parts of Texas. For decades, voting districts have been drawn based on residents, some of whom are eligible to vote and some of whom are not. The Evenwel plaintiffs argue that they live in state senate districts in where a relatively high proportion of district residents are eligible to vote. By contrast, other Texas state senate districts have lower proportions of eligible voters, rendering actual votes in each of those districts more valuable.
The Court’s decision to take up the Evenwel case is surprising. For more than 50 years, it has maintained the principle of “one person, one vote,” and the federal government and state governments have taken this to mean drawing roughly equal districts based on Census counts of the entire population. But the total population numbers include many people who might normally be barred from voting, such as noncitizens, prisoners, felons, non-citizen immigrants, and children. The Supreme Court also declined as recently as April 2013 to hear a similar case involving local districts in Texas.
Given the Court’s sudden interest in the matter, many have speculated on the possible repercussions, with a mad scramble towards mid-cycle redistricting across the United States, a possible decline in the number of Latino elected officials, and the likely decline in political clout of inner cities vis-à-vis suburbs. The number of congressional seats would likely be unaffected: the Constitution states clearly that these will be apportioned by a count of free persons, based on the 14th Amendment’s promise of equal protection under the law and the 1924 Indian Citizenship Act. And yet the Court could decide to allow states to determine which populations deserve equal protection—excluding, say, immigrants, prisoners, and so on–when it comes to voting.
At the same time, observers have identified several problems if the Court allowed states to move away from population-based assessments. There are practical concerns, such as the lack of a reliable count of adult citizens: all the citizenship data we now have is based on sample estimates, whereas the decennial Census counts all persons. There are also substantive concerns over the loss of diversity in Congress, and over noncitizens losing political representation by proxy, particularly through family members who are eligible to vote.
But few observers have paid attention to how noncitizens, or ineligible voters more generally, are currently represented in Congress and other political institutions. For instance, how much do congressional districts vary in their proportion of citizens and noncitizens?
As we can see, House districts vary significantly in the proportion of residents who not voting-age citizens. In nine congressional districts (one in Illinois, one in Arizona, two in Texas, and five in California), such ineligible voters make up more than half the resident population. More generally, districts where more than 40 percent of the residents are ineligible to vote are predominantly urban, connected to large metropolitan areas like Los Angeles, New York, Houston, Miami, Chicago, and Dallas. Thus, changing our long-established method of counting “one person, one vote” would have a disproportionate impact on legislators with urban districts. It would also disproportionately affect Democratic seats (32 out of 38), with comparable Republican seats coming from Florida, metro Atlanta and Houston, and rural parts of California and Washington.
However, citizenship is only part of the reason that such districts might have a high proportion of ineligible voters. Many have a sizable proportion of citizens under age 18. Indeed, among the 38 House districts where the Citizen Voting Age Population dips below 60 percent, half the ineligibles are U.S. citizen children.
Moving beyond Congress, figuring out state legislative districts’ proportion of eligible voters is more challenging. We have no centralized and standardized source of data on citizen voting age populations across all 50 states’ districts. The Los Angeles Times has done an initial analysis and visualization of the proportion of state assembly districts and Los Angeles city council districts with high proportions of noncitizens. We might expect more local and regional analyses in the days ahead.
In the meantime, the best way to visualize local noncitizen representation (defined here most basically as those who are counted in the drawing of districts even if they are unable to vote) is to look at county-level maps of the citizen voting-age population.
As we can see, most counties in the United States have very high proportions of residents who are voting-age citizens. In the median county, 75 percent of residents are U.S. citizens of voting age. The Sunbelt has most of the counties with relatively low proportions of residents who are voting-age citizens. These include not only urban counties in California and Florida, but also many border counties in Texas and rural counties in the Mountain West.
In other words, any Supreme Court decision that allows states to base voting districts on eligible voters rather than total population could scramble political fortunes not just in the places we think of immediately, like Los Angeles, New York, Houston, and Miami, but also in many rural districts in states across the country.
Karthick Ramakrishnan is Professor of Political Science and Associate Dean of Public Policy. Sono Shah is a Ph.D. Candidate in Political Science, at the University of California, Riverside.