No, At-Large Voting Isn't Fair or Equitable.
Voters Should Rely on Data, Not Political Rhetoric When Deciding How to Vote on Knoxville City Charter Amendment No. 2
This November, Knoxville voters will decide on a referendum that could change how city council members are elected.
If approved, City Charter Amendment No. 2 would remove district-level elections and move to a fully at-large system, where all council members, including those running for district seats, would be chosen by the entire city. This would result six district seats (with residency requirement) elected through city-wide primary and general, and three at-large seats elected through city-wide primary and general. Residents would vote on all nine council seats, including in districts of which they are not a resident.
If rejected, district seats will be elected by district voters and at-large seats will continue to be elected city-wide. This would result in six district seats elected through district-only primary and general, and at-large seats elected through city-wide primary and general. Residents would vote on all three at-large council seats + their own district seat.
For a broader discussion about the referendum and the electoral systems used by other cities across the state, I recommend reading a recent column by Bill Lyons in the Knoxville News Sentinel.
Here are a few reasons why you should vote no on City Charter Amendment No. 2.
What Does Representation Even Mean?
As I explained in a recent op-ed for Knox News, an at-large electoral system will result in the city’s wealthiest and least diverse neighborhoods having the greatest influence over who represents any given district in the city. Proponents of the measure suggest this isn’t true — but the data is abundantly clear.
Take, for example, the last five primary elections in which only district seats were on the ballot. Each district accounts for roughly 20% of the eligible voter population.1 In a perfect world — a world in which voter registration and voter participation rates are the same across all parts of the city — each district would make up roughly 20% of the vote.
However, research shows there are persistent gaps in voter turnout rates across different demographic groups. Higher-income citizens vote at higher rates than low-income citizens. Homeowners vote at higher rates than renters. White citizens vote at higher rates than non-White citizens. These findings are corroborated by scores of studies and empirical research (see here, here, and here).
Disparities in voter participation help to explain why District 2 and District 4 — two of the city’s wealthiest and least diverse districts — have accounted for a disproportionate share of voters in recent elections. Despite making up only ~20% of the eligible population each, District 2 has on average accounted for 29% of primary voters while District 4 has accounted for 30%. Together, District 2 and District 4 have accounted for at least 55% of all voters in the last five primary elections, despite representing ~40% share of the eligible voter population.
Centralizing electoral power among just a handful of districts is the core issue. Under a fully at-large system, high-turnout districts would have disproportionate influence, enabling them to effectively select representatives for the entire city—even for districts that may not share their interests or priorities. In fact, assuming the current turnout trends continue, an at-large system would give Districts 2 and 4 more control over who represents District 1, for example, than the actual residents of District 1.
While the figure above illustrates the variation in turnout across the last five primary elections, the disparities in turnout can be even more dramatic depending on the year.
For example, District 3 residents accounted for just 9% of all voters in the 2021 primary election. If an at-large system had been in effect at the time, 91% of the votes in District 3’s primary would have been cast by people who don’t reside in the district — leaving the actual residents of District 3 with very little influence over who represents them on city council.
This inevitably leads to the question: what is the purpose of a district representative if most of the voters who elect them don’t even live in the district?
Moreover, the inequity of at-large voting is evident even under the city’s current electoral scheme, which gives voters city-wide veto power over the preferences of any given district. Data from past city elections proves this point.
As shown above, the winner of the District 3 primary has lost in the city-wide general election 57% of the time over the past four decades. Meanwhile, the district’s preferred candidate in the District 6 race lost the city-wide general in three consecutive elections in 1989, 1993, and 1997 — meaning District 6 voters did not have their preferred candidate representing them for a span of 12 years and the entirety of the 90s.
Example: Hypothetical District 1 Election with At-Large Voting
Imagine Candidates A, B, and C are competing for the District 1 council seat, but instead of a district-only vote, they must compete in a city-wide primary. Assume about 15,000 residents cast ballots, with voter turnout across districts mirroring the average of the last five elections.
Candidate A focuses heavily on engaging with District 1 voters, attending local meetings and knocking on doors. This grassroots effort pays off—Candidate A secures 65% of the votes cast by District 1 residents. However, despite being well-known and supported within District 1, Candidate A is relatively unfamiliar in other parts of the city, leading to the lowest overall city-wide vote count. As a result, Candidate A doesn't advance to the general election despite being the top choice within their own district.
Candidates B and C thus move on to the general election. In the general election, assume about 20,000 people vote, with district turnout following the same pattern. Candidate C is particularly unpopular among District 1 voters, having lost both the primary and the general election within the district by a significant margin.
However, Candidate C focuses their campaign on high-turnout districts, like Districts 2 and 4, where they build strong support. This city-wide strategy leads to a decisive win in those high-turnout areas.
As a result, Candidate C wins the city-wide election by a margin of 53% to 47% and becomes the next city council representative for District 1 — even though they are the candidate with the least support from the voters of District 1.
Highest Turnout Districts Are Among the City’s Least Diverse
Higher turnout districts also tend to be less diverse than the city at-large, according to census bureau redistricting data compiled by Knoxville-Knox County Planning.
Minority residents (i.e. non-white residents) represent roughly 30% of the city’s overall population. All city council districts, except for District 5 and District 6, are less diverse than the city overall.2 District 6 remains the city’s lone majority-minority district (i.e., a district in which the majority of the constituents in the district are racial or ethnic minorities), with minority residents accounting for ~57% of the district.
As shown above, the two highest turnout districts — District 2 and District 4 — have a lower (1) minority share and (2) Black or African American share of the population than the city overall.
Well-established threats to minority group voting rights are electoral schemes like 'at-large' districts that dilute the voting power of the minority group and make it virtually impossible to elect a candidate of the group's choice…
— NAACP, Make the Right to Vote a Fundamental Right Resolution
Put simply, if historical turnout trends continue, moving to an at-large system could result in the diminution of minority voting influence. This risk has long been recognized by prominent advocacy and civil rights organizations, and the NAACP Legal Defense Fund explicitly opposes at-large voting for precisely that reason.
Does the Proposed Amendment Violate the Voting Rights Act?
Because of their dilutive effect, many at-large voting systems across the United States have been struck down by courts under Section 2 of the Voting Rights Act.
Interestingly, there is even a relatively local example from a neighboring city in Tennessee. In the 1989 case Brown v. Board of Commissioners of Chattanooga, Tennessee, a federal district court struck down the City of Chattanooga’s at-large voting system, finding the system to be in violation of section 2 of the Voting Rights Act.
It is critical to note, however, that the mere potential for discriminatory outcomes under an at-large electoral system does not on its own constitute a violation of the Voting Rights Act.
Instead, courts have relied on the Gingles test — a legal standard derived from the 1986 U.S. Supreme Court case Thornburg v. Gingles — to adjudicate claims of racial vote dilution under Section 2 of the Voting Rights Act. The test establishes three preconditions that plaintiffs must satisfy:
Size and Geographic Compactness: The minority group must be large and geographically compact enough to constitute a majority in a single-member district.
Political Cohesiveness: The minority group must demonstrate that it is politically cohesive, typically voting similarly in elections.
Majority Bloc Voting: The majority must vote sufficiently as a bloc to usually defeat the minority's preferred candidate.
Rising Cost of Representation
An at-large system will raise the cost of running for office and thus exacerbate the role of money in politics.
Wealthier districts not only have higher turnout rates but also tend to raise and contribute more to political campaigns. Candidates in high-turnout districts often attract more donations, while those in lower-turnout areas tend to raise substantially less. Moving to an entirely at-large system would require candidates to raise vastly more money to get their message out to the entire city instead of just their friends and neighbors.
In a city where the cost of elections is already rising, this shift risks sidelining candidates who are more representative of Knoxville’s neighborhoods but lack the financial resources to compete, especially against against well-funded challengers.
Each district contains roughly 16.7% of the city’s total population. However, District 5 appears on the ballot during a different election cycle. When excluding District 5, each district contains roughly 20% of the city’s population that is eligible to vote in the primary election in which Districts 1, 2, 3, 4, and 6 appear on the ballot.
District 5 is excluded from the analysis because it appears on the ballot during a different election cycle and, thus, is not directly comparable to other districts in terms of turnout.