(The following article first appeared in Blueprint for Social Justice. It is based on the 1995 book Proportional Representation and Election Reform in Ohio, edited by Kathleen Barber and published by Ohio State University Press. The book is available from the Center for $39.)

The True Experience of Proportional Representation in American Cities

by Kathleen L. Barber

In Ashtabula, Ohio, an Irish Catholic was elected to the city council for the first time. Polish Americans were elected in Toledo. In Cincinnati, Hamilton and Toledo, path-breaking African Americans were elected to their city councils, while in Cleveland the one African American on council was joined by others as the black population grew.

When did these breakthroughs in representation of urban minorities in America happen and why? The forgotten history of proportional representation elections in the United States holds the answers, and may provide the key to solving the puzzle of fair representation in the future as well.

The rules of the electoral game as it is played in the United States have an impact on participation in public policy decisions that few worry about or even recognize. Yet the traditional use of a winner-take-all election system excludes from representation large numbers of Americans whose candidates simply lose in plurality contests.

Plurality electoral systems, whether in single-member districts or at-large, provide simply that the candidate or candidates with the most votes will win. In the familiar single-member district election, with two candidates in the race, up to half of the voters may be losers. If there are three or more candidates, close to two-thirds of the voters may not win any representation in the elected governmental body.

In a plurality at-large election, such as a seven-member at-large city council, the majority may sweep the field, filling all the seats, to the total exclusion of minorities. A minority organized to vote strategically, that is, to cast one or two vote s only, and for the same candidates, may win a seat or two. But more often when a minority candidate is elected in this electoral context, it is the majority's choice of who should represent the minority.

The right to vote without a right to representation is not inevitable; it is simply a hallmark of the plurality election system.

The racial gerrymander -- districts drawn to facilitate representation of otherwise excluded minorities in city councils, state legislatures, and Congress -- has been successful in expanding representation in some situations. However, not only is the racial gerrymander merely a partial remedy for under-representation, but its use is increasingly restricted by court decisions finding that in the attempt to remedy previous wrongs it unconstitutionally handicaps other groups.

Growing awareness of these problems has spread interest in electoral systems that are capable of providing representation not only to a plurality or a majority of the voters, but to most voters. Such systems are taken for granted in most of the democratic world, and they are generally known as proportional representation.

Party list systems and the Single Transferable Vote (PR/STV), sometimes called "choice voting" or "preference voting," can deliver representation to 90% or more of the voters, yet they are little understood in this country. Many people believe them to be strange "European" ways of voting. But PR/STV was developed in nineteenth century England (also the source of plurality voting practices). Furthermore, in the Progressive Era, PR elections were used in 22 cities in the United States. In a curious case of omission and ne glect, their history was lost for decades. Myths grew up around PR to explain and perhaps to justify its repeal in 21 of the 22 cities that used PR for local elections. Only in Cambridge, Massachusetts have council elections by PR survived.

In the 1980s a project was undertaken to recover that lost history and analyze what actually happened in a group of American cities that used PR/STV elections between 1915 and 1960. Inspired by the late Leon Weaver of Michigan State University, political scientists Dennis Anderson of Bowling Green State University, Ronald Busch of Cleveland State University and I, joined by urban historian Robert J. Kolesar of John Carroll University, gathered election statistics and information about the characteristics of council members in five Ohio cities before, during and after the duration of proportional representation council elections.

The Roots of Proportional Representation

The social, economic and political context for the adoption of PR is critical for understanding its electoral outcomes in American cities. The idea of proportional representation is rooted in late eighteenth century debates about democracy. In both th e United States and France, revolution was in the air, and ideas about voting and representation emerged from resistance to reigning oligarchies. Popular demands for participation in governance stimulated fears that majority tyranny would replace the tyranny of the minority. French political theorists and mathematicians such as Mirabeau, Condorcet and Saint Just understood that suffrage was the key to democracy, and that different methods of casting votes would have different consequences. They developed a variety of electoral systems that would produce majority rule but would also ensure minority representation. The educated minority was their principal concern.

Although these theorists themselves died in jail or under the guillotine in the Terror -- the very mob rule they had feared -- they profoundly influenced the nineteenth century struggle for representative government. Alexis de Tocqueville in France an d John Stuart Mill in England became leading proponents of proportional representation. Tocqueville advanced the cause of party list voting, while Mill advocated adoption of Thomas Hare's Single Transferable Vote, which he called "the greatest improvement of which the system of representative government is susceptible."

In the new United States, the term "proportional representation" was used to mean representation of the states in proportion to their population, as distinct from one state, one vote. Federal debates over democracy -- who should vote and how votes should be cast -- were settled by devolution of these important questions to the states. The American framers, like the French theorists, were worried about potential tyranny of the majority, but they relied for its prevention on the institutional checks of the Constitution, and on the "social check" of conflicting interests, which James Madison argued would prevent the dominance of any one interest. Representation in the new American governments was generally determined by state adoption of the territorially-based English practice of plurality voting, whether in single- or multi-member districts.

By the second half of the nineteenth century, the inability of minority parties or minority interests to gain representation in winner-take-all elections led to a search for other ways of voting. Alternative electoral systems such as PR/STV, Cumulative Voting (CV) and Limited Voting (LV) gained adherents in the United States and fueled repeated efforts to try new voting methods. The two dominant parties, however, controlled Congress and most state legislatures, the very institutions which held the power to enact or prevent election system change.

Ohio was especially fertile ground for reform agitation. After the Civil War, political loyalties remained rooted in the powerful emotions of the war years, and Republicans and Democrats were closely divided in the state. Election rallies were militant affairs because with plurality voting the conversion of just a few voters could swing a legislative or congressional district from one party to the other. As a result, there was little relationship between the share of votes won by a party's legislative or congressional candidates and the share of seats won in Ohio's Assembly or in the congressional delegation. Frequent post-election shifts in party control of the legislature spurred repeated redistricting of congressional seats as the new majority part y attempted to maximize its power for the next election.

The Progressive Reform Agenda

Partisan misrepresentation and party duopoly were the reigning political conditions leading to the Ohio Progressives' commitment to election reform. Because the dominant party machines in Ohio between 1890 and 1920 were (with a few exceptions) Republican, the Progressive leadership was composed primarily of urban Democrats, independent Republicans, and an unusually large component of labor leaders. Their agenda featured such measures as initiative, referendum and recall, popular election of senators, w omen's suffrage, railroad rate regulation, municipal ownership of utilities, protective labor legislation, tax equity, and home rule for cities.

Much of this agenda was achieved in 1912 through state constitutional amendments. Home rule for cities opened up the party-controlled political system to grass roots action. The local initiative became a critical tool enabling urban residents to wrest control of their cities from "the bosses." At that time, the power of the machines was built on the votes collected through small favors in the wards of the major cities, and financed both by contributions from businessmen seeking franchises and tax benefits, and by kickbacks from illegal operators in gambling and prostitution who bought protection for their activities. It was opposition to this power structure that united the reformers.

In 1914, the National Municipal League, the official voice of Progressive urban reform, endorsed both the city manager plan and election of city councils by PR/STV, stressing the linkage between the two elements of reform. If administration were to be professionalized, it became critical to have a council fully representative of the people to make the policies which the manager would carry out. In 1915, Ashtabula, Ohio became the first American city to adopt PR elections.

In subsequent adoptions of home rule charters, four more Ohio cities -- Cleveland, Cincinnati, Hamilton and Toledo -- chose proportional representation elections for their city councils. Most "reform" cities in the United States at the time adopted the city manager plan with the intent of improving administrative efficiency, but simultaneously adopted small, plurality at-large councils. What was different about the PR cities was the firm conviction of their charter sponsors that efficient administration had to be implemented in the framework of more representative -- and therefore more responsive -- policy-making. This meant doing away with winner-take-all voting systems.

While many Progressive reformers aspired to have cities run like business corporations, most in the PR stream of the Progressive movement rejected the implied analogy. John R. Commons, for example, argued that a corporation is created to produce dividends, while a city is "a compulsory corporation into which men are born." He wrote in Proportional Representation (1907):

In a political corporation different classes of citizens often have different interests. Therefore all interests and classes should be represented in its administration. In what direction its sovereign powers shall be employed is a political question, involving justice and expediency as well as business. Shall taxes be levied to protect health, to extend free schools, to cleanse the slums, to buy water-works or street-car lines? -- these are a few of the political questions which cities must consider (pp. 200-201).

Commons' analysis of group representation demonstrates the flexibility of PR to meet diverse and changing interests over time. He noted the difference between John Stuart Mill's concern about representation of the educated minority and his own worry: "PR is no longer needed to defend the rich against the poor. Its problem now is to defend the masses against the monopolists." In Ohio at the turn of the century the interrelated pressures of rapid industrialization, immigration and urbanization created t he explosive conditions Commons sought to address.

Did Anything Change?

With this brief account of the setting for the adoption of PR in the five Ohio cities, let us turn to the findings about the impact of changing electoral systems. Three major areas of impact were the subject of evaluations and debate in adoption and repeal campaigns. These were:

Voter turnout: did it rise or fall with the adoption of PR/STV elections?

Characteristics of council members: what kinds of changes, if any, occurred in the demographics or socio-economic status of successful candidates in PR elections as compared to council members elected before and after PR?

Stability in governance: did the enhanced potential for group representation through PR elections lead to greater or less conflict or dissensus in the way councils made their decisions?

Because the contemporary sources of information about PR were largely polemical, we chose verifiable measures to the extent possible with available sources. We collected voting data from the five cities on three elections before the implementation of PR/STV systems, the PR elections themselves, and three elections following repeal. PR systems were utilized in the five cities for council elections for overlapping periods: Ashtabula, 1915-29; Cleveland, 1923-31; Cincinnati, 1925-55; Hamilton, 1927-59; and Toledo, 1935-49.

Turnout. Advocates of PR elections argued that the expanded opportunity to have one's vote count in PR elections was a significant incentive to vote. Typically, as many as 90 percent of the voters could expect to have a candidate elected that t hey had supported; turnout would grow. Opponents, on the other hand, argued that the complex ballot with its multiple ranked choices, as well as the lengthy count before results could be known, were deterrents to voting; turnout would decline.

Limits on the availability of population and registration data made comparisons difficult, but it is fair to say that the findings for four of the cities were inconclusive. The change in elections systems did not have an independent effect on turnout. Turnout rose and fell with local issues and candidates in all three periods before, during and after PR elections. In one city, Toledo, a measurable drop in turnout occurred during the period of PR elections. Because the PR years in Toledo included World War II, we considered the possibility that the dislocations of both military personnel and factory workers might have affected turnout in Toledo. However, turnout for presidential and congressional voting rose in Toledo during the war, as it did in the country as a whole, even while it was declining in Toledo municipal elections. With the exception of Toledo, then, the change in electoral systems did not affect turnout.

Characteristics of Council Members. Here, the findings were significant. Data were collected on the ethnicity, race, sex, party affiliation, occupation and to a limited extent education of council members in the five cities under three electoral systems. Replacing large ward-elected councils (Cleveland, Cincinnati and Toledo) or smaller mixed councils elected by a combination of districts and at-large (Ashtabula, Hamilton) by small councils elected at-large by PR changed the composition of their members. Generally, professional people replaced the small businessmen who predominated in the pre-PR councils. Labor leaders replaced blue collar workers. The level of education of council members rose in the large cities (where education data were available). Incumbents also were more commonly reelected. These changes could be attributed in part to the larger constituencies created by the at-large feature of PR elections. In Cleveland, where four large districts, each electing 5, 6 or 7 members, were drawn, for a total PR council of 25 members, the increases in status, while measurable, were less dramatic.

After PR was repealed, the four cities that kept their small councils but elected them by plurality at-large saw a further rise in socio-economic status of council members, and a greater tendency toward incumbent success. With Cleveland's return to a 33-member ward-based council, professional representation remained constant, but blue collar workers gained seats at the expense of council members with business ties.

PR ballots as well as post-PR systems were nonpartisan in all five cities, but we found little separation of local electoral politics from the tides of national party affiliation. The Republican party was dominant in Ohio during the years of the earlier PR systems, and councils were generally dominated by Republican-endorsed candidates. But the GOP was weakened by successive splinter movements and by the sharp blow of the Great Depression. Both the earlier Republican hegemony and the growth of Democratic strength in the 1930s were reflected in local elections under all electoral systems.

Surprisingly, minor party candidates failed to win elections on the PR ballot. Only one, a Socialist candidate in Ashtabula, was elected in 1915. Before World War I, however, Socialist candidates were also elected in Ohio in plurality elections to both city councils and school boards. In subsequent years, the party's opposition to the war and the postwar Red scare limited its electoral opportunities under any method of voting. Although both Socialist and Communist party candidates ran for local office in Ohio in the 1920s and 1930s, none was elected again in the PR cities. Support for the minor party candidates appears to have been too thin to reach the ten or twelve percent of the vote, either on first choice or through transfers, needed to win a sea t on the nine- or seven-member councils, or in the Cleveland multimember districts. Independents, however, were occasionally elected by PR.

The most important diversification through the PR/STV ballot brought significant minorities onto the councils in these Ohio cities. African Americans were elected for the first time in Cincinnati, Hamilton and Toledo, while in Cleveland the one black pre-PR council member (a "ward boss" in the Republican machine), was joined by others elected in proportion to the growth of the African-American population in the city during the 1920s.

In Ashtabula, Irish Catholics, a residentially dispersed minority apparently unable to elect either a ward member or an at-large representative to the city's former mixed council, gained a voice through PR elections. In Toledo, Polish Americans had a similar experience, winning one or two seats in each citywide PR election.

These results were not necessarily welcome at the time; critics attacked PR for producing "racial and religious blocs" on the councils of the PR cities. In Cincinnati and Toledo, the prospect of PR-elected councils electing a minority member as mayor loomed large in the final and successful repeal campaigns. In most of these cities, after PR/STV was repealed, minorities lost their foothold in public office in the winner-take-all plurality at-large elections that supplanted PR.

Only in Cleveland, where the 33-ward council with single-member plurality elections replaced PR in 1931, did minorities hold on to their newly-won voices in public life. Cleveland's residential patterns included pockets of ethnic voters clustered in separate neighborhoods and increasingly segregated blocks of African Americans in wards where minority candidates began to run against each other. The combination of small districts and housing segregation was the key to preserving the minority representation on the city council which had been achieved in the 1920s through the PR ballot.

The impact of PR on women's representation was less consistent than the positive outcome for ethnic and racial minorities. Except for Ashtabula, where no women ran for city council in the PR period and the issue appears not to have been raised, Ohio w omen were active workers in charter campaigns. After winning suffrage in 1920, these and other younger women often became candidates for local office, and some were elected. Only in Cleveland, where no women were elected to council before PR or for over a decade after its repeal, did women reach significant numbers on the PR councils. The carryover of feminist commitment from a strong suffrage movement and support from well-organized professional women's groups seem to have made the difference in this rap idly growing industrial city.

When the PR/STV electoral system was challenged in court as a violation of both the Ohio and the U.S. Constitutions, it was a woman from Cleveland, Judge Florence E. Allen, first woman on the Ohio Supreme Court and later first woman on the federal court of appeals, who wrote the opinions upholding its validity (Hile v. City of Cleveland, 107 O.S. 144, 141 N.E. 35 [1923], writ of error dismissed, 266 U.S. 582 [1924]; Reutener v. City of Cleveland, 107 O.S. 117, 141 N.E. 27 [1923]).

Stability in Governance. Opponents of PR in Ohio did not go away over the years between 1915 and 1960. Repeated repeal issues were brought before the voters, usually by political leaders who had been ousted from power in the transition from the old to the new way of voting. Three times in Ashtabula and five times in the other four Ohio PR cities, challenges to PR were brought to the ballot, finally leading in each case to success. Although the politics of these initiatives differed, a common theme that seemed to strike a response from the voters was fear of political fragmentation.

The appeal to fear was often built on the success minorities experienced with PR, allegedly forming the "racial and religious blocs" discussed earlier. By the 1930s, this threat was linked to Nazi Germany, where the proportional representation elections of the Weimar Republic were blamed for the rise of Hitler.

Facts could easily be cited to highlight the differences between PR/STV and the German PR/Party List system, between the German and American political cultures, and between Germany and the other stable and democratic PR/party list electoral systems of Europe. But the facts never seemed to catch up with the charges, and the strategy was effective in repeal campaigns. Were these urban American PR councils actually vulnerable to conflict and dissensus in their operations? Local newspaper stories from the PR period reported significant public achievements and progress in dealing with the cities' problems of urban growth and service delivery. Only in Ashtabula, a small city deeply divided by Prohibition and its enforcement, did the press consistently highlight council disputes and deadlock.

To move beyond impressions from the media, we collected data on non-unanimous votes on substantive issues throughout the three electoral system periods in our five cities. Procedural votes were omitted from the tabulation (usually unanimous), but both ordinances and resolutions were included. Factionalism actually declined significantly in Hamilton and Toledo with the introduction of PR/STV elections. In Cleveland and Cincinnati, there were no significant differences in the percentage of divided votes on council issues under different electoral systems. Only in Ashtabula did conflict increase. A minority of "wets" on a previously "dry" council may account for the modest rise we found there in non-unanimous votes. When Ashtabula's PR system was repealed in 1929, it would not be long before Prohibition itself disappeared and with it the cause of local factionalism.

Dispelling the Myths

These historical findings of what actually happened in a group of American cities using PR elections should serve to dispel some of the myths that have developed around proportional representation. In most cities, turnout did not decline. If some were discouraged from voting by the ranking of choices on the ballot, others were attracted to the polls by their enhanced opportunity to be represented. These Ohio cities did not tumble into factionalism because minorities won a share of council seats. In most of the cities, the incentive to candidates to draw second- and third-choice votes from supporters of competing candidates led to less stridency and more cooperation in campaigns and subsequent governance.

This history has lessons for our time. The right to vote, guaranteed by constitutional amendment long ago and slowly implemented by statutory enforcement, has grown to embrace the right to representation as well. No longer are Americans satisfied to appear at a voting booth and cast a ballot without the realistic expectation of winning a voice in public decisions. The 1982 amendments to the Voting Rights Act of 1965 addressed this problem of minority vote dilution which occurs in winner-take-all electoral systems. The amended statute requires that minorities have the opportunity to elect candidates of their choice. It specifies that "proportional representation" of groups in the population is not required, but neither is it prohibited. The search for standards by which fair and effective representation can be judged has been ineffectual in part because of our simple ignorance of alternative electoral systems.

It is widely believed (by judges and citizens alike) that single-member districts, however grotesquely drawn, are the sole remedy for exclusion of minorities from a fair share of seats in the nation's legislative bodies. If this were true, segregation in housing would be an electoral imperative instead of a social ill. But it is not true. As our case studies of American PR cities have shown, minorities may be residentially dispersed but politically united, and are capable of choosing their own representatives when they are provided with appropriate electoral tools.

The culprit responsible for exclusion is plurality, winner-take-all voting, not the at-large feature. If conducted by a preference ballot that empowers minorities as well as the majority, at-large elections expand the range of choices for all voters. As the U.S. Supreme Court draws the noose ever more tightly around districts drawn to expand minority representation, the alternative of proportional representation becomes ever more compelling. As our book concludes:

"Free voting," Cleveland's Tom Johnson called it a century ago, attempting to capture in a phrase the fluid opportunities implicit in proportional representation. In the microcosms examined here, not all social, economic, and political problems were solved, nor should electoral systems be expected to bring utopia, but adoption of the proportional principle for American elections would create opportunity for a more inclusive, less polarized democracy.


Kathleen L. Barber's book, Proportional Representation and Election Reform in Ohio was published in 1995 by The Ohio State University Press. 383pp. ISBN 0-8142-0660-3.