Note: This article is part four in a miniseries about how voters respond to alternative, and particularly proportional, voting systems. You can read the previous articles in this series here: Part 1, Part 2, Part 3.

Fair voting methods, and particularly multi-winner elections paired with proportional voting, provide a more robust solution to the problem of under-representation of people of color than traditional solutions.
Tweet This

Amarillo, Texas, is as quintessentially American as you’re likely to find.  The panhandle city of nearly a quarter-million is on Route 66, home to chuckwagon barbecue, rodeos, and a world famous Cadillac monument. It also has been home to another all-too-common fixture of the United States: the underrepresentation of people of color in elected bodies. In the 2000 census, more than a quarter of the city’s residents identified as Hispanic, Latino, or African American, yet no African American and only one Latino had ever been elected to the city’s school board.

The city took a bold step to change that in 2000 by doing away with its plurality voting system and instead holding its school board elections via a semi-proportional voting method called cumulative voting. In that one election, the city doubled the number of people of color ever elected to the board, with an African American and a Latina winning seats. Two years later, the city added a second Latina to its board.

The success of reform in Amarillo is not an anomaly. More than 200 jurisdictions in the United States have broken the winner-take-all plurality voting mold, and currently use proportional and semi-proportional voting, or single-winner ranked-choice voting (collectively I refer to these voting systems as fair voting methods). Many of these American jurisdictions turned to fair voting methods with the explicit intent of empowering voters of color to elect candidates of their choice.

More-proportional voting systems have changed the makeup of legislative bodies across the United States, from the Yankton and Sisseton-Wahpeton Sioux tribes in South Dakota, to Latino communities in New Mexico, New York, and Texas, and African American communities in Alabama, Illinois, and Massachusetts. Even non-proportional fair voting reforms, such as single-winner ranked-choice voting, have strong track records of boosting representation for communities of color, such as for Asian Americans in California’s Bay Area elections.

Despite the successes that pepper the US map, when most democracy reform advocates look for solutions to the persistent problem of underrepresentation of people of color in elected offices, they rarely think of changing the voting system. Instead, they turn to solutions that manipulate district lines but maintain unrepresentative plurality vote counting methods. Though racial districting—drawing one or more single-winner districts where a majority of voting-age adults in the districts are of a minority racial or ethnic group—can sometimes elect more candidates of color, it is also vulnerable to gerrymandering and demographic shifts, and often continues to short-change communities of color from representation proportional to their share of an electorate. Other times democracy reform advocates neglect to look at the voting method at all and instead focus on boosting turnout among communities of color. Though get-out-the-vote efforts are inspiring, they typically yield only modest effects on election outcomes.

Fair voting methods, and particularly multi-winner elections paired with proportional voting, provide a more robust solution to the problem of under-representation of people of color.

Semi-proportional voting leads to immediate and sustained wins for candidates of color

The most comprehensive research regarding how voting system reform influences representation for communities of color in the United States examines results of 96 elections using semi-proportional voting methods across 62 US jurisdictions. The study found that Latino candidates won seats in 70 percent of the cases in which a Latino sought representation. African American candidates won at least one seat in 96 percent of the elections where one or more African American candidates were on the ballot. By comparison, in nearly all of these jurisdictions almost no African American or Latino candidate had been able to win a seat under the previous plurality voting system.

The slightly smaller gains in Latino representation compared with that of African Americans are the result of a few factors. Likely the most influential of these is municipal laws that split up the seats across different election cycles. This split keeps the threshold for winning a seat out of reach for many Latino communities based on their share of the electorate. That electoral design flaw, which may well have been instituted by leaders in those cities precisely to stymie electoral success of candidates of color, is one Cascadia likely won’t be vulnerable to, as I’ll show.

Additional obstacles to fair representation for communities of color are common to all election systems, and include candidate recruitment, turnout among voters of color, and obstacles from skewed campaign finance systems. I’ll explore each of these in greater detail below.

Semi-proportional voting elects staggeringly more candidate of color in US elections than plurality voting

Many American cities, counties, and school boards use single-winner at-large elections to determine their representatives. That is, candidates run jurisdiction-wide but they pick one seat number and only run against candidates who also picked that number. They don’t have to compete with candidates who picked a different seat or position number. But each candidate has to win over the same voters to win a seat. For example, in 2018 Portland is electing two city councilors in at-large elections. Julia DeGraw and others are challenging Nick Fish for Position Number 2, while Jo Ann Hardesty, Andrea Valderrama, Loretta Smith, and others are competing for Position Number 3.

In contrast, over 200 American localities use semi-proportional voting methods, where all candidates run against each other and the top candidates from the entire pool win seats. If Portland used such a method, voters might be able to elect both Jo Ann Hardesty and Loretta Smith or both Julia Degraw and Nick Fish. Semi-proportional elections result in elected bodies that much more accurately reflect their electorates than the results of plurality, at-large elections.

The chart below, adapted from this research, shows election outcomes with each of these election methods in jurisdictions with relatively similar population sizes and in which African Americans made up the largest racial or ethnic minority group. (The average population size in the jurisdictions using semi-proportional election methods was 10,000; jurisdictions using at-large plurality elections had an average population size of approximately 40,000.) Researchers chose these relatively small population sizes because most jurisdictions using semi-proportional voting systems in the US today have relatively small populations. The horizontal axis below shows the percentage of the voting age population who were African-Americans, and the vertical axis shows the percentage of city council or school board seats African Americans won.

The orange line demonstrates how single-winner at-large elections exclude African-Americans from local leadership—even in jurisdiction where African-Americans make up fully half of the voting age population, they win a mere 11 percent of seats in single-winner at-large jurisdictions. Though other research shows that plurality, at-large elections yield somewhat better outcomes for African Americans in larger jurisdictions, plurality voting still fails to yield proportional results until African Americans comprise close to half of the population.

In contrast, the green line shows that in cities using semi-proportional voting, African-Americans win seats nearly in direct proportion to their share of the population, winning nearly half the seats when they make up half the population.

Original Sightline Institute graphic, available under our free use policy.

Fully proportional voting systems can produce even more equitable results

Semi-proportional voting methods have changed the face of government in the American cities, counties, and school boards that use them. They elect a far more representative number of candidates of color than single-winner plurality elections. What’s more, because they don’t create winners and losers based on district lines, they pull the rug out from under gerrymandering, reducing its power.

Fully proportional voting methods would likely yield even better election results than semi-proportional systems, while retaining the benefits of being gerrymander-resistant. Unfortunately, only two US jurisdictions currently use fully proportional voting systems: Cambridge, Massachusetts, and a few city boards in Minneapolis, Minnesota, use multi-winner ranked-choice voting, also known as single-transferable voting (STV) (residents of Amherst, Massachusetts recently decided to adopt STV as well, but have yet to implement the system). As a result, data confirming the results of fully proportional voting in US elections are sparse.

Anecdotally, STV has been highly successful in diversifying representation on Cambridge’s city council, with the city retaining consistent African American representation since 1980 (and in all but one election cycle since 1963). A study comparing the actual 2013 Cambridge city council elections, conducted via STV, with a simulation of that year’s results under a winner-take-all bloc voting system found that without proportional voting the city would have had much less turnover and much less diversity on its council, retaining two more of its incumbents and failing to elect two councilors of color who won seats under STV.

  • Our work is made possible by the generosity of people like you!

    Thanks to Clay Veka for supporting a sustainable Northwest.

  • Historically, at least 24 cities in the US have used STV for local elections; African Americans and other candidates from racial, ethnic, and party minority groups had widespread success in winning seats with this system. In fact, in many cases, it was these successes that led to the repeal of the system. For example, New York City’s decade long experiment with STV from 1936 to 1947 revealed the effectiveness of PR in giving more third parties and other minority groups a voice in city office. The system effectively stopped the democratic party’s near total control of city council with only two-thirds of the popular vote, electing representatives from the Communist, Fusion, American Labor, and Liberal parties. STV was also responsible for the city council’s first female member and first two African American city councilors (one of whom went on to become the state’s first African American congressman). Unfortunately, this success in bringing more voices into the city’s policy conversations may have also led to the system’s ultimate repeal as the two major parties ultimately saw PR as a barrier to their own power.

    Changing the threshold of exclusion opens the door for diversity

    Proportional and semi-proportional elections open the door for minority groups of any type—racial, ethnic, religious, party, etc—to win seats by creating a lower threshold of exclusion. That is, they make it possible for candidates to win seats with a smaller minimum percentage of the total votes. In proportional and semi-proportional voting races, the higher the number of seats up for election in the same cycle, the lower the threshold. Conversely, the fewer seats a jurisdiction elects per cycle, the higher the threshold to win one of those seats. So, a five-winner school board race would have a lower threshold—candidates would need roughly 17 percent of the vote to ensure a win—than a two-winner race, in which candidates need at least a third of the votes to win a seat. Proportional and semi-proportional elections almost always have lower thresholds than elections held via plurality voting.

    Many of the jurisdictions in the research I examined above that administered elections via semi-proportional voting and in which Latinos formed the predominant minority group, ran only a few seats per election cycle, creating a relatively high threshold of exclusion for these seats. This threshold was likely the key driver behind Latinos winning representation in only 70 percent of the cases in which they sought election. Though that was still a major improvement over Latino representation under plurality voting, the success rate was below that of African Americans who enjoyed elections with more seats and therefore lower exclusion thresholds.

    Any move to proportional voting methods will result in thresholds lower than under Cascadia’s current election methods. Most Washington and Oregon elections have a threshold of 50 percent plus one to win a seat. For example, in a top-two election, the type currently used by every jurisdiction in Washington state, a candidate must receive 50 percent of the vote plus one to win. In contrast, a multi-winner proportional race in which voters elect even just three candidates together would require just half this amount, or only 25 percent of the votes plus one.

    Oregon conducts most of its elections via “pick one” plurality ballots. Many other Oregon cities use bloc voting. In both cases winners simply need a  plurality of votes to win, which depends on the number of candidates on the ballot. The more candidates on the ballot, the lower the winning threshold can be. In practice, however, most races see few candidates on the ballot, resulting in a high vote threshold to win a seat. As a result, it’s often nearly impossible for any minority group, even those representing a large portion of the electorate—say, 20 or 30 percent—to gain a voice. Oregon’s plurality elections almost never elect candidates from third parties, and rarely elect candidates of color.

    Instant-runoff voting can also result in more candidates of color winning seats

    Though single-winner ranked-choice voting, also called instant-runoff voting (IRV), is not a proportional voting system, it also can support the election of more candidates of color compared with at-large plurality elections. It does so because ranked voting allows members of a minority group to avoid splitting their votes among similar candidates, which can happen in any plurality election, whether at-large or districted (even majority-minority districts are not immune). For example, in San Francisco’s 2006 Board of Supervisors race for District 4, the initial count of only first-round votes found that Asian American voters split their votes among four Asian American candidates, resulting in a non-Asian candidate winning all of the first-round votes. But after the subsequent rounds of counting, the ranked voting system consolidated Asian American voters’ preference on their consensus candidate, Ed Jew, an Asian American candidate who won the the seat.

    Other hurdles for electing candidates of color common to all voting systems

    Proportional electoral methods are a powerful reform that can give voice to voters and candidates who go unheard in single-winner at-large elections. But, electoral method reform alone can’t solve everything. Three other factors can still impede fair representation for people of color no matter what voting system a jurisdiction uses.

    1. Candidate recruitment

    First, in order to elect candidates from minority groups—whether racial, ethnic, religious, party, or any other minority—candidates from these groups have to run. That may sound obvious, but under-recruitment can be a major reason for lack of representation, even with reforms aimed at boosting minority representation. In the study I cited above, many of the communities in which Latinos made up the predominant minority group struggled with Latino candidate recruitment, making it impossible for Latinos to gain representation even though the voting system would have supported them to do so.

    2. Minority group turnout

    A second barrier for members of minority groups, including people of color, is voters of these groups must fill out and return their ballots in order to gain representation. When turnout of a particular minority group is far below its share of the electorate, it will be difficult for that group to win fair representation no matter what voting system it uses. Unfortunately, in Cascadia as elsewhere in the United States and Canada, turnout rates among underrepresented minority groups (such as people of color and younger voters) is consistently lower than among plurality voter groups (such as White, older voters). The good news is that, as I covered in a previous article, proportional voting systems help overcome this barrier; they tend to lead to pronounced and sustained boosts in turnout amongst people of color, helping those voters get their fair share of representation.

    3. Lack of public financing

    A third barrier to fair representation with any voting system is a broken campaign finance system. The current system favors candidates who go into the race with already well-established lists of wealthy friends and potential donors. Candidates without these connections are at a disadvantage from the word go. To level the playing field, jurisdictions need to implement public funding programs, which make it possible for candidates to run viable campaigns without fat rolodexes or spending all their campaign time dialing for dollars. Seattle’s democracy voucher program or Portland’s public match system both offer successful models for giving serious candidates from all backgrounds the chance to reach voters and have a voice.

    Conclusion

    By taking on the challenge of voting system reform, Cascadia could join the ranks of jurisdictions across the continent addressing the root cause of unequal representation for communities of color: “pick one” plurality elections. In fact, with fully proportional voting like STV, Cascadians could see even more representative election results than those enjoyed with semi-proportional voting systems in 200 places across the United States, including Amarillo, Texas.

    Thank you to George Cheung, Liz Dupee, and Alex Hur for reviewing this article and providing helpful comments and insights.

    Note about additional sources

    In addition to the citations I link to in the text above, I relied on two other sources that are not archived online anywhere. Both pertain to election outcomes from cumulative voting in Peoria, Illinois. Here are the full citations:

    Aspin, Larry and Hall, William. (1996) Cumulative Voting and Minority Candidates: An Analysis of the 1991 Peoria City Council Elections. The American Review of Politics, Vol 17: 225-244.

    Aspin, Larry. (2001) Cumulative Voting and Straight Voting: An Empirical Comparison. The American Review of Politics, Vol 22: 55-91.

    Note about single-winner districts

    Another common election system American cities, counties, and states use to boost representation for communities of color is majority-minority single-winner districts. When jurisdictions draw district lines to maximize the population of a particular minority group within a specific district, these voters have a better chance of electing a candidate of their choice to represent their district. For example, African American voters in majority-African American districts typically win better representation than under single-winner at-large elections.

    This reform, however, isn’t intended to give minority communities proportional representation outcomes; rather, it’s meant to boost representation based on the number of majority-minority districts that map makers can create and courts require. Often, due to geographic or other factors, it’s not possible to draw a proportionate number of majority-minority districts to fairly represent racial and ethnic minorities. For example, people of color in Portland are not sufficiently concentrated in any particular part of the city to allow line-drawers to draw them into a majority-minority district. Some recent studies suggest that single-winner district plans may in fact reduce political power and representation for communities of color (for examples, see here, here, here). However, due to data limitations, I did not discuss the impacts of majority-minority districting plans on election results in this article.