New York City’s local elections would benefit greatly from reform. The city has a budget larger than all but a handful of states and powers that profoundly shape the everyday lives of its residents, yet it suffers from low voter turnout, poor election administration, outsized special interest influence, and a city council elected largely through uncompetitive general elections.
This report describes New York City’s electoral system, including its recent introduction of ranked-choice voting (RCV) for local primary elections. It then turns to the shortcomings of the city’s system, including chronically declining voter participation in local races, a city Board of Elections rife with ineptitude and scandal, and New York State’s fully closed primaries, in which only voters registered in a particular political party may vote in that party’s elections.
To address these concerns, this report proposes a series of reforms for New York City’s electoral process. Some are long-overdue necessities, and others are long-shot overhauls. The underlying goal is to reinvigorate local democracy by encouraging broad participation, democratic deliberation, and robust political competition.
- Board of Elections Reform
The state legislature should require that commissioners have relevant election administration expertise, allow the mayor to nominate commissioners, extend removal powers to officials beyond the governor, require the state Board of Elections to issue regulations and exercise oversight, and replace the statutory bipartisan staffing requirement for technical and clerical positions with a transparent, skills-based hiring program and strict ethical codes.
- On-Cycle Elections
New York’s local election cycle should move away from its current off-year timing to match the national even-numbered election-year cycle in order to boost turnout, which is consistently higher in elections with state and national races, particularly presidential elections. On-cycling would also save the city resources, as the federal funding used for federal races could cover Election Day expenses for local races.
- Include Endorsements on the Ballot
To help voters identify the candidates who best match their values and preferences, candidates’ names should appear alongside endorsements from high-level local officials and local organizations that qualify as endorsers. Providing immediate informational signals also helps ameliorate the issue of voters selecting candidates on the basis of race or gender, as signaled by candidate names.
- Abolish the Offices of Public Advocate and Borough President
The public advocate’s role is largely symbolic. Its stated purpose—holding the mayor accountable to the people—is unnecessary, given the mayor’s prominence in New York’s government and media. Borough presidents have for more than three decades, following the abolishment of the Board of Estimate, lacked meaningful political power. Abolishing these offices would enable councilmembers to serve as singular liaisons between their constituents and city government, rendering them more accountable to their voters.
- Implement Final Five Voting (FFV) or Reintroduce Proportional Representation (PR)
This report’s final section proposes two transformational changes to the structure of elections for the New York City Council, both of which utilize ranked-choice voting. The first would introduce an improved variant of Final Five Voting. All registered voters, regardless of their political party, would vote for candidates in a qualifying primary, and the top four vote-getters—again, regardless of party—would advance to a general election that uses ranked-choice voting.
The second change calls for reintroducing proportional representation (PR), in which the sub-groups that comprise the electorate would more closely match the composition of seats on the council. PR would be achieved via multimember districts and the single transferable vote, the multi-winner variant of ranked-choice voting.
This report’s appendix contains party-list options to obtain proportional representation, which necessarily involves political parties as part of the city’s electoral architecture. The first option involves an open-list system, whereby parties obtain seats according to their share of the vote, but voters can select the priority of candidates each party elects. The second option, for mixed-member proportional representation, would have voters select most councilmembers through elections for single-member districts. Each political party with a sufficient number of votes would also fill a number of council seats, in proportion to its share of the vote, so as to achieve proportional representation in the council as a whole.
All New Yorkers, regardless of background or political persuasion, share an interest in a more representative and innovative government. Dynamic political competition can best achieve these ends. Without electoral reform, such competition is all but impossible. In proposing reforms, this report seeks to boost political competition and the benefits that flow from it.
New York City’s local elections consistently rank among the bottom in the United States. When New Yorkers go to the polls, they can expect administrative blunders, poor voter turnout, and boring, uncompetitive general election races. Though the city has 8.5 million inhabitants, and 5 million active registered voters, most mayoral elections are decided by about 1 million registered Democratic voters in the primary election. Races for the New York City Council frequently sport even lower turnout; some do not even feature an opposition candidate in the general election. New York is proud to be one of the world’s most diverse cities, but its political arena does not reflect the diversity of its voters’ views—the hegemonic Democratic Party has long dominated the city council.
New Yorkers’ failure to show up at the polls does not appear to result from their widespread satisfaction with city government. An October 2021 Emerson College/PIX11/NewsNation poll, taken shortly before the local general election, found that 57% of likely voters believed that the city was headed in the wrong direction. That did not translate to high turnout: the election saw a mere 23.3% of registered voters cast a ballot, a historic low.
Elections do more than simply select the next slate of officeholders. Their manner and structure create incentives for the way representatives conduct themselves and respond to stakeholders’ concerns. Without broad democratic participation, elected leaders may support policies that favor the factional and special interests that help them win elections, not necessarily those that benefit the general welfare. Poorly structured elections impose costs on the public at large—costs that, in New York, can be felt in issues as diverse as housing and public safety.
For example, long-standing tradition grants city councilmembers an effective veto over any rezonings that occur in their districts. They often exercise this “aldermanic privilege” by halting new housing and business development in response to acute pressure from not-in-my-backyard groups. New York consequently loses opportunities and residents to other cities less hostile to new development. Former councilmember Carlos Menchaca, for example, added so many conditions to the approval of his district’s proposed Industry City development, including major concessions to the powerful hotel workers’ union, that investors walked away. These concessions to antidevelopment special interests happen regularly, even though 71% of New Yorkers, according to a 2021 Manhattan Institute poll, support making the housing approvals process easier so as to facilitate more housing development. Local electoral reform can help the city achieve its perennial needs for economic growth and housing.
Greater political competition should yield political leaders who are more attuned to local needs and more responsive to voter preferences, providing benefits to many different groups in the process. Research on political competition reveals a statistically significant and strong positive effect on state income and economic growth.
Cooperation, too, has a place in politics, but New York’s local electoral system largely discourages it. A dominant Democratic Party need not fear that voters will back its rivals enough to threaten its power. It thus has little to gain from across-the-aisle compromise. The state’s closed primaries, moreover, give conservative- and independent-minded New Yorkers little practical incentive to register as anything other than a Democrat. Those who opt not to register with any party—the second-largest group in the city, numbering about 879,000 active registered voters—simply cannot vote in any primary election.
New York City’s local elections were not always this way. Following a successful 1936 city charter amendment, elections for city council from 1937 to 1947 used a system of proportional representation via multi-winner ranked-choice voting, in which candidates were elected in proportion to their total boroughwide vote. The result, according to law professor Lani Guinier, was not only increased turnout but a greater number of “energetic and public-spirited candidates.”
As a diverse set of politicians and parties entered the council, they invigorated its politics. This era saw the first black member of the council, Adam Clayton Powell Jr., as well as its first female member, Genevieve Beavers Earle. Substantive and energetic debates entertained and educated New Yorkers, who listened rapturously to radio broadcasts of proceedings from the council chamber. An estimated 750,000 residents tuned in to what became one of the most popular radio programs of the period.
This was not a mere historical quirk. Vigorous, engaging, and broadly popular local politics can become the norm once again. But doing so requires changes to the way New York structures and administers its elections.
This report provides voters, organizations, and elected officials options and ideas to improve local elections, ranging from long-overdue necessities to long-shot overhauls. While it focuses on New York City, many of its findings and suggestions can be applied to cities across the United States.
At the outset, it is necessary to note that no single electoral system is ideal; all involve trade-offs and depend on the circumstances of each place. Even if voters from all political stripes agree that the current system is severely lacking, they will not always agree on solutions. To be clear, overhauling the electoral system may necessitate an attitude shift for residents. Some may perceive contentious council debates as democratic deliberation; others may perceive the debates as chaotic squabbles.
Any efforts to achieve electoral reform will face powerful obstacles. Even if greater political competition would represent a net gain for the majority of city residents, it will invariably pose threats to the beneficiaries of the current system, particularly incumbents and highly motivated special-interest groups. During New York’s years of proportional representation, it survived two unsuccessful attempts at repeal; modern-day reform would likely engender similar opposition.
Voters and political leaders will have to decide, based on political feasibility and philosophical preferences, how to structure elections so as to align the interests of the people with those of their representatives. Slow, incremental reform may be the only attainable avenue forward, but it is nonetheless worth the effort to identify paths toward improvement.
That said, reform should strive to achieve four important goals, all of which transcend political divides:
- Broad popular participation. All else being equal, greater democratic participation allows elected officials to enjoy a more secure mandate for their agenda. It also signals greater public confidence in the electoral process as the primary way to resolve political differences.
- High-quality voter deliberation. An electorate of well-informed voters making an active and deliberative choice yields a stronger mandate for politicians. Voters benefit from knowing about the policies at issue in each election.
- Alignment of interests and preferences between voters and representatives. Voters should feel that their preferences and concerns are not shut out by the political process and are represented by some elected officials. Citizens alarmed by growing public disorder, for instance, should be able to address their concerns with a receptive councilmember. A mismatch between the priorities and preferences of the populace and their representatives will cause voters to feel that the political process is failing them, which, in turn, leads to disengagement.
- Avenues for fair political competition and cooperation. At every election, voters should feel that there is a meaningful and viable alternative to the party or coalition in power and that therefore their vote makes a difference. The advantages that incumbents enjoy should not be perceived as insurmountable to challengers. No party or coalition should feel dominant enough to eschew competition and appropriate collaboration.
Context: New York’s Fusion-Ticket Balloting and Ranked-Choice Voting
Before addressing NYC’s election-related problems, it is necessary to note some features of its current electoral system. Under the state’s fusion ballot system, used in general elections for national, state, and local offices, a single candidate may be nominated by multiple parties and thereby run on more than one ballot line simultaneously. For each election, city voters see ballot lines for perennial minor parties like the Working Families Party and Conservative Party, as well as idiosyncratic, election-specific parties like the “Stop de Blasio” and “No Homeless Shelters” parties. Minor parties may nominate their own candidate or one from another party.
Allowing these parties to cross-nominate major-party candidates achieves several aims. Fusion-ticketing can avoid the “spoiler effect” associated with nominating separate third-party candidates, i.e., running two ideologically similar candidates who would split voter support and improve the odds of an ideological opponent. If a third party nominates someone other than a major-party candidate, however, the spoiler effect remains. Candidates running on more than one party line may also benefit from the perception of greater legitimacy, as well as from simply having multiple opportunities under which voters may select their name.
To general election voters, this system provides information about candidates who align with the third party’s positions, which tend to be narrower and more focused than those of the major parties. Pro-union voters who know about the Working Families Party’s labor-centric platform, for example, can vote for the Democratic candidate on the WFP ballot line. Those frustrated with a major party, but supportive of its candidate, can register such discontentment by voting on the minor-party line. Finally, voters might signal their chief concern by voting, for example, for the easily accessible “Rent Is 2 Damn High” ballot line.
Having more parties involved in general elections increases political competition; major party leaders have to weigh the prospect that smaller parties will not cross-endorse their candidate, thus raising the possibility of defeat. These pragmatic considerations influence the platform that major party candidates adopt.
The major downside of fusion-ticketing, from a voter’s perspective, is greater ballot complexity and thus greater chances for confusion. Others claim that the system leads to political patronage; a major-party candidate may, for example, appoint leading members of a third party to government positions in exchange for its support. It should be noted that, even if these arrangements sometimes have the appearance of corruption, many third-party leaders have specific competencies that lend themselves to particular government positions. A top member of the Green Party may, for example, be uniquely well suited to address pollution policy.
Powerful officials have sometimes resented the increased competition stemming from fusion-ticketing; then-governor Andrew Cuomo and Democratic Party chairman Jay Jacobs, for example, sought to quintuple the number of votes needed to secure a party’s ballot position in the subsequent election, from 50,000 to 250,000. Following a legal challenge and political pushback, this figure was reduced to 130,000, which the legislature passed in 2020 for the next year’s budget.
Ranked-choice voting (RCV) is another noteworthy recent feature of NYC’s primary elections. In 2019, NYC voters approved an amendment to the city charter that introduced a single-winner version of RCV (sometimes known as the alternative vote (AV)) for local primary and special elections, which was first utilized in June 2021. Voters may, but are not required to, rank up to five candidates; candidates with the fewest votes get successively eliminated, and their votes are redistributed proportionally to their voters’ second-choice ranking, until a candidate secures a majority. General elections are conducted using a traditional single-choice vote. In theory, RCV should give longer-shot and independent-minded candidates a greater opportunity to win, insofar as they can run without being spoilers.
While RCV carries benefits over the single-choice system that it replaced, it is not a panacea for New York’s electoral issues. RCV allows voters to express their preferences sincerely, without worrying about the spoiler effect. But greater complexity comes with more risk of ballot invalidity. What, then, does the overall picture surrounding RCV look like?
The emerging consensus among electoral-reform groups is neatly summarized in a recent report by Lee Drutman and Maresa Strano at the New America think tank. On the whole, it recognizes that RCV is a relatively modest reform, especially in the single-winner form that New York implemented. RCV shows promise in counteracting the spoiler effect of third-party candidates.
Insofar as RCV encourages more candidates to run, it should promote accountability and scrutiny for incumbents. Competition can also secure concessions from leading candidates on issues important to sizable minorities of voters. A well-known environmentalist candidate, for example, might withhold support from a leading candidate until he promises to improve the city’s plastics recycling infrastructure, thereby giving him an incentive, in theory, to adopt policy positions that would appeal to voters likely to rank the environmentalist above him. A wider selection of candidates can also translate to greater voter participation, as citizens who prefer long-shot candidates are likely to remain engaged if their favorites are still in contention, although empirical evidence linking RCV to higher voter turnout is mixed.
RCV may also encourage more civil campaigns. During the 2021 Democratic primary, for instance, Andrew Yang encouraged his supporters to rank Kathryn Garcia second. This went unreciprocated, however, and more evidence is needed before a definitive statement about campaign civility can be made—but it nonetheless shows that preference-swapping deals are possible under RCV.
What is more certain is that RCV infrequently changes election outcomes. Analyses of past RCV races have found that the candidate with the most votes after the first-round tally has a high chance to win the election. Though Eric Adams won the first Democratic primary using RCV in a crowded field of challengers, he led in every round of the tally, indicating that he may have also won in a traditional single-choice plurality system.
As for voter satisfaction, a recent report by Erin Carman and Jay Wendland for the Unite America Institute found that New Yorkers were generally pleased with their first experience using RCV in last year’s primaries. A formidable 77% of respondents wanted to see RCV used in future city elections. Among mayoral voters, 84% ranked more than one choice; 45% ranked the maximum of five. While only 26% ranked the full five city council candidates, largely because of a lack of information about those running, 70% ranked at least two candidates. By and large, voters appeared to understand the way that RCV works, with 77%–80% of voters across different racial backgrounds reporting that they understood the system very or extremely well. Only 20% of voters said that they did not understand the system. As Drutman and Strano suggest, RCV will likely come to feel natural for New York voters as they continue to use it over time.
Two phenomena associated with RCV—ballot exhaustion and bullet-voting—are cause for concern. RCV ballots are said to be exhausted when they do not make it to the final-round tally because all their choices have already been eliminated. The chances of exhaustion increase when voters do not rank all their available choices. In the 2021 NYC Democratic primary, 14.7% of all ballots for mayor were exhausted. Sometimes, ballots do not count because of voter error—e.g., leaving ballots blank or mismarking them. When voters rank two candidates first, their vote does not count; this is called a fatal overvote. In last year’s primary, 1.2% of ballots were invalidated because of a fatal overvote. When New York used RCV’s multimember variant in elections from 1937 to 1947, well over 10% of ballots were routinely invalidated, largely because voters used prohibited “X” marks. Sustained voter education efforts and technological means of counting good-faith marking errors can diminish this issue.
Bullet-voting, relatedly, occurs when a voter either selects one candidate and leaves the other choices blank, or ranks the same candidate multiple times (with only the highest rank counting once). This may occur intentionally—multiple choices for the same candidate may signal emphatic approval—or in error, the result of misunderstanding. According to Campaign Finance Board voter analysis, 5.6% of Republicans (who only had two mayoral candidates) and 3.7% of Democrats bullet-voted in last year’s primaries.
As one recent study of ranked-choice races observes, high rates of ballot exhaustion make it more likely that a candidate will win the election without receiving a majority of total votes cast. However, as Drutman and Strano note, this hasn’t been a major issue in U.S. RCV elections thus far: “[W]hen one considers all RCV elections in the United States, most of them do result in a final-round winner with a majority of total votes cast.”
Finally, there are good reasons to believe that alternatives to RCV—such as the open-list proportional representation model, championed by Jack Santucci—would carry significant advantages. (These worthy non-RCV proposals are discussed in the Appendix.)
All that said, eliminating RCV in New York so soon after its implementation would likely prove difficult, both practically and politically. RCV can also be adapted to obtain proportional representation by using the single transferable vote, its multimember variant, without disruptive changes to the way New Yorkers vote in primaries today. This form of RCV could produce a city council that reflects the share of votes that smaller parties receive in an election. As a pragmatic matter, therefore, the reform proposals for New York considered in the latter part of this report assume RCV as a part of the city’s electoral architecture.
The Flaws of NYC’s Election Landscape
New York’s elections suffer from multiple overlapping structural problems. Before identifying avenues of improvement, it is necessary to address some flaws in the city’s current system.
Low Voter Turnout in Primary and General Elections and Low-Information Local Races
Despite the introduction of RCV and efforts to expand early voting and absentee mail-in voting throughout the last decade, voter turnout remains low in NYC. A small minority of active voters determine the outcomes of NYC’s local elections. Indeed, typical primary turnout rates make it possible for a candidate to effectively win a citywide race with support from only 9% of all registered voters. According to the Campaign Finance Board’s latest voter report, the overall turnout in local primaries has varied over the last decade, from a low of 14.9% in 2017 to a recent high of 26.5% in 2021. Not included in those percentages are registered voters who are ineligible to vote in primaries, either because they are not registered with any party, or because they are in a party that did not hold a primary. When the 1,013,427 primary voters in 2021 are measured against the 4,919,037 total registered voters, that percentage drops to 20.6%, some six percentage points lower than the board’s statistic would suggest.
General election turnout is bleaker still. The 2021 mayoral general election witnessed a mere 23.3% of registered voters heading to the polls, a historic low. Turnout in city elections has declined sharply and consistently over the decades, as shown in Figure 1 below.
NYC Turnout in Mayoral General Elections: 1953–2013
Yet as the chart’s early years show, low turnout was not always the norm in NYC local elections. The mayoral general elections held in the 1950s saw an impressive 93% of registered voters cast a ballot in 1953 and 91% in 1957. Notably, between 1921 and 1965, New York State’s literacy law required prospective voters to pass an English literacy test in order to register to vote, which had the effect of discouraging registration.
Compounding the problem of low turnout is the low-information nature of city council races; voters generally do not know much about candidates’ policy views. Obtaining such information is costly to voters, who will not undertake the effort unless the perceived payoff from doing so outweighs the costs. District-specific races for council candidates, who are typically obscure outside their local areas, often generate little media buzz, even in New York’s relatively rich local media environment. Given the city’s exorbitantly expensive media market, which includes parts of New Jersey, council candidates cannot afford to launch television ads to raise awareness.
Aside from the relatively small cohort of voters motivated by policy or affinity, many decide based on party identity, name recognition, and, most troublingly, a completely irrelevant factor. Some studies find that voters without informational cues on Election Day, such as party labels, vote on the basis of race and gender, as signaled by perceptions of candidates’ names. All else being equal, candidates with surnames that begin with the first letter of the alphabet outperform those whose names begin later in the alphabet. Name recognition appears to matter far more than policy positions; according to David Birdsell, in city council elections, “[t]he information base available to all but the most actively engaged voters is very small, making name recognition perhaps the key variable in victory.”
Why is low voter turnout problematic? After all, thoughtful commentators like Errol Louis contend that, while reasonable efforts should be made to facilitate voting, New York’s low turnout may result from the well-known phenomenon whereby 20% of the people in organizations accomplish 80% of the significant work. If the majority of residents cannot be bothered to vote, the argument goes, let it be.
But the combination of low voter turnout and low-information local races calls into question whether electoral outcomes truly reflect the deliberate will of the electorate. To take an extreme example, if all voters made completely random choices on Election Day, those they elect could not fairly claim to wield power on the basis of the popular will. Much the same can be said about the results of an election where most voters lack any relevant information. When elected officials secure their authority through a small minority of voters, the perception of legitimacy can likewise attenuate, at least in the minds of some residents.
Showing up at the polls is one way that voters signal their belief in the fairness and legitimacy of the electoral system as a whole; low turnout, by contrast, might be interpreted as tacit disapproval on the part of those who believe that their vote is inconsequential.
Research by Zoltan Hajnal and Jessica Trounstine also suggests that lower local election turnout tends to reduce the representation of racial minorities in city councils and mayors’ offices. Specifically, higher turnout leads to significantly higher levels of representation of Latinos and Asian-Americans on city councils.
Finally—and crucially—low turnout races give special-interest and single-issue groups disproportionate power. This is especially true for public-sector unions, whose members stand to gain from officials they help elect. As Manhattan Institute senior fellow Daniel DiSalvo writes: “The phenomenon of high public sector worker turnout, driven in part by self-interest, combined with the apathy of average citizens, is very significant for public policy, especially at the local level.”
In her groundbreaking recent book on interest groups’ impact on local policies, Sarah Anzia finds that in cities that use collective bargaining, public-sector unions—especially those representing police and firefighters—are among the most active groups seeking to obtain favorable municipal policy. This influence is apparent not only in generous pay and benefit packages for municipal workers but also, more subtly, in the ways that public employees approach their jobs. Union leaders have a strong incentive to push for lenient working conditions and greater insulation from performance-based accountability for their members. Political leaders, for their part, often acquiesce to these demands because of their less obvious and immediate impact on budgets, especially if doing so will avert political opposition or even a strike. Anzia finds that greater activity by police and fire unions was associated with a $13-per-capita increase in police spending and a $20-per-capita increase on fire protection spending. Given New York’s population of 8.5 million, that would correspond to additional spending of around $110.5 million on police and $170 million on fire protection.
To be sure, low turnout is the norm in most American municipal elections, a somewhat counterintuitive phenomenon. As Daniel J. Hopkins writes in a Manhattan Institute issue brief examining the causes of big-city voter turnout declines, “Low voter engagement with city politics is something of a paradox since cities control a variety of policy levers that can have major impacts on citizens’ day-to-day lives—and also since cities’ smaller sizes relative to states means that blocs of voters are more likely to influence election outcomes.”
In particular, the reach and scale of New York’s local government makes it uniquely important in the lives of its residents. While election-related issues are not exclusively to blame for New York’s governmental problems, they create incentives for elected leaders that make it more difficult to tackle political problems, especially those caused by special interests. The city’s over–$104 billion 2023 budget is larger than that of every state government except California, New York, Texas, and Pennsylvania. The city Department of Education alone commands a budget of $38 billion, instructs some 1.1 million students, and employs more than 300,000 full-time employees. Unlike much of the country (where school districts are the norm), New York’s state legislature has, for the last two decades, permitted the mayor to administer the city’s public education system, the largest school district in the country. That arrangement has recently come under question. In June, the state legislature granted a two-year extension of control to the Adams administration but signaled that this might be the last renewal of mayoral control.
If New York does away with mayoral control of the schools, the need for electoral reform will become only more acute. School-board races held under the current electoral system are likely to be very low-turnout affairs, and candidates for these boards are susceptible to capture by special interests, particularly New York’s teachers’ union. Michael Hartney, for example, has recently shown that candidates who receive teacher-union endorsements are more likely to win school-board elections—but are not necessarily better qualified. Absent elections that promote accountability, self-dealing by board members may become the norm. Indeed, scandal and corruption marked many school boards in New York’s past, with some that operated as the “personal fiefdom” of their board members.
Closed Partisan Primaries
New York State is one of nine to utilize fully “closed” primary elections, that is, limited to voters registered with each party. For NYC, this restriction is problematic for at least four reasons. First, meaningful political competition occurs only within the Democratic primary, allowing special interests to wield disproportionate influence. Second, voters are discouraged from registering with their preferred political party and from participating in general elections that they know to be uncompetitive. Third, without on-ballot informational cues, low-information voters cannot distinguish between primary candidates’ policy stances, a particularly thorny problem in light of the nonideological nature of many local issues. Fourth, a lack of political competition in the general election impedes policy innovation and partisan representativeness in the city council.
As mentioned above, New York’s local primaries generally feature low turnout—last year’s 26.5% turnout among major-party voters was enough to mark a multi-decade high. Turnout is often even lower in the general election. Registered Democrats in NYC constitute about 68% of all active registered voters, according to February 2022 data, so winning the Democratic primary very often implies an easy victory in the general election. The Campaign Finance Board voter analysis notes: “The energy that fueled higher-than-average participation in the June primary did not carry through to the fall, which continued a trend of non-competitive, low-interest general elections for local office.”
For the city council, uncompetitive races have long been the norm, even during the years when the city was led by a Republican mayor. In the five local elections between 1989 and 2005, Republicans won the mayor’s office four times; yet Democrats controlled 44–48 of 51 city council seats, thus ranging from 86% to 94% of the body.
Nor do unaffiliated and third-party voters show up in particularly large numbers on general election days, largely because they believe that the result is foreordained. In assessing the reasons for the low turnout in the 2021 general mayoral election, Manhattan College political science professor Margaret Groarke said: “The lack of competitiveness and the related lack of campaigning is enough to explain the turnout…. [P]eople aren’t mobilizing voters a whole lot, and voters don’t have a sense the vote matters as much.”
The number of votes cast in last year’s mayoral general election was 1,147,555, only slightly more than the 942,031 (first-round) votes cast in the mayoral Democratic primary (again, limited to Democrats). Among newly registered voters, primary turnout was 39.8% in 2021—well above the 26.5% figure for all voters—but general election turnout was only 18.5%—well below the 23.3% overall turnout rate. New registrants were, reasonably, far more interested in a meaningful contest.
Many voters correctly anticipated that the mayoral general election was a foregone conclusion and thus not worth the effort to participate. In fact, this view is entirely rational: aside from the pull of civic duty, what benefit do voters gain that would compensate for taking time out of their Tuesday or even mailing in an absentee ballot, only to participate in an election that they know to be uncompetitive?
Closed primaries further facilitate one-party rule by disincentivizing voters from affiliating with the political party that most closely represents their policy preferences. Most obviously and importantly, those who do not register with any party—the second-largest group of city voters—cannot vote in any primary, and the same is true for those belonging to a party that did not host a primary. Changing one’s party to participate in its primary must be done on or before February 14, four months before the primary. Across New York State in 2020, Democrats gained some 68,000 net new members from those who switched their affiliation from Republican. One grassroots group even formed specifically to urge more to do the same, in the hopes of swaying the Democratic primary.
In limiting participation to party voters, closed primaries likewise discourage political candidates from reaching out to those outside the party fold to build new electoral coalitions. Because non-Democratic candidates generally stand little chance of securing a general election win, either for the mayor or the city council, few non-Democratic voters turn out, either in the primary or general election. Out of NYC’s roughly 500,000 registered Republicans, only about 60,000 voted in the June 2021 primary; primaries for the Conservative and Working Families Parties were not held.
Even after gaining two seats in the 2021 city council elections, Republicans command five seats out of 51, or about 10% of the body. Yet in 2020, President Donald Trump, whose political brand was, in general, deeply unpopular with New Yorkers, nonetheless received 23% of NYC’s vote in a high-turnout race. Similarly, according to my tabulation of all city council general election results in 2021, Democratic candidates secured about 76.6% of all validly cast votes in all the general election races for city council in 2021; Republicans (including ballots cast for them under the Conservative Party label) obtained 19.8%; independent third-party candidates 3.1%; and write-ins the remaining 0.5%.
The 3.1% share of the vote that went to third-party candidates (who were not also major-party candidates) indicates a not-insignificant level of support for politically unconventional councilmembers. Under proportional representation, a 3.1% share of a 51-member council would have received at least one seat (assuming, of course, that a single party received that entire share). Yet the council does not feature any independent third-party members.
In sum, for all practical purposes, anyone who is not a registered Democrat has no say in the only election that matters. Among the roughly 3.76 million registered Democrats in NYC, approximately 384,000 are inactive, meaning that they never vote. Another 1.7 million have not voted in a primary in over a decade. This implies that, at most, the remaining 1.68 million active and semi-active Democratic voters—in a city of some 8.5 million people—decide nearly all of New York’s officeholders.
But when these voters cast their vote in the decisive primary, they have little information on the ballot to guide their choice. In contrast to the general election, where party labels signal some information about candidate policy preferences, there is no way for voters to distinguish between candidates in a closed primary. June 2021’s Democratic primaries, for instance, sported a crowded slate of candidates, including some endorsed by the Democratic Socialists of America; others focused on single issues; some were more conventional, center-left types; and some were local neighborhood leaders. Yet the Democratic ballot contained only candidate names, without any additional information that would help voters identify which contender best aligned with their preferences.
The lack of candidate-level policy signals on ballots is particularly problematic for municipal elections. Unlike at the national level, at the local level, parties are often not the primary locus of political affiliation. Civic group affiliation, homeownership, union membership, and racial and ethnic identification represent alternatives to party identity. Homeowners, for example, may diverge sharply on social views yet share common aims of preserving their property values and limiting the externalities associated with greater real-estate development. Many constituencies will simply not know about or have well-developed preferences about local issues; the same homeowners would likely not know or care about public-sector collective bargaining or the details of local tax policy.
Nonetheless, the two major national parties predominate in local elections, including in New York City. But voters’ national-party identities are formed almost exclusively around national-level issues, not local issues. Municipal governments simply do different things from federal and state governments. The most important policies at the local level, such as those governing land use and zoning, economic development, real-estate taxes, and schools, overlap little with policies that get attention in Washington, DC, which include immigration, foreign policy, health care, and national security. Indeed, many local issues simply do not have clear ideological fault lines. Matters of routine governance, such as the efficiency of the city bureaucracy, park maintenance, food vendor permitting, and housing development, for example, do not bifurcate neatly along party lines. Anzia states:
[M]any local policy issues, such as policing, are cross-cutting and not a comfortable fit within the national party system. Other local issues, services, and priorities seem to have little partisan dimension, such as efforts to attract jobs to the city or having a well-functioning sewage system. It is not clear how national-level partisanship or ideology would be connected to residents’ preferences on these local matters.
Furthermore, election law prevents national parties from developing local-level brands that could better adapt to these circumstances; voters cannot register as, say, a Republican in local races and a Democrat in national ones. That hinders third-party registration, especially given the salience of party in national politics—New York’s Independence Party, for instance, does not compete meaningfully in national elections. Even within the two-party structure, the city’s minority Republican Party cannot develop a sub-brand that appeals to voters who would favor fiscal discipline and smaller government but with different social views from the national party. And the longer the minority party does not exert political power, the more difficult it becomes to persuade voters to give the party a chance, as voters have no experience upon which to base their expectations.
Because councilmembers in noncompetitive districts win general elections on the basis of party, far more experienced and qualified candidates from opposing parties never make it to office. For example, in 2005, at the peak of Mayor Michael Bloomberg’s popularity, Republicans nominated Joel Zinberg, a Yale-educated lawyer and cancer surgeon who supported the mayor’s agenda, to compete against Jessica Lappin, a first-time candidate, for a council district on the Upper East Side that Bloomberg had won decisively. Despite endorsements from the mayor, the New York Post, and the New York Times, Zinberg lost by about a 2–1 ratio, almost entirely on the basis of party. The same could be said about third-party opponents.
A lack of political competition has profound consequences for municipal policy. Some academic studies suggest that low levels of political competition reduce the perceived costs of engaging in corrupt activities. One-party rule allows for governance without the moderating influence or policy innovation that competition engenders. Schleicher aptly sums up the issue:
The problems created by a lack of partisan competition are not only felt statically, but also dynamically. That is, each set of elected officials is less representative than it should be and there is an inefficiently low amount of policy development, candidate incubation and interest group mobilization opposed to the dominant party over time. Political parties do more than put candidates up for office—they serve as the fulcrum for the creation of ideas about governance and for the development of future political leaders. They also organize groups into politically effective coalitions. In a one-party city, there is too little of each of these.
Poor Election Administration
New York City’s Board of Elections (BOE) offers, in essence, a modern-day glimpse into the way that city government functioned under Tammany Hall. The 10-commissioner board, split evenly among Democrats and Republicans, comprises appointees selected by the political parties from each of the five boroughs; the city council merely rubber-stamps their approval. Neither commissioners nor the staffers they hire need have any election administration qualifications or experience. According to a 2021 state senate report: “Elections Commissioners are entrusted with significant responsibility, yet there are essentially no standards in place for qualifications or training…. There are no standards for conflict of interest, or opportunities for the public to understand how and why Commissioners are appointed to their roles.”
Commissioners have no reason to fear accountability or removal. The governor has the exclusive power to fire commissioners for cause. But there is little practical reason for a governor to exercise this power, and thus it has never been used. The 10-member board allows each commissioner to evade individual accountability—after all, failures can always be attributed to other members. Firing commissioners also necessarily entails political risk. Removing a political ally threatens the relationship between governors and their party, potentially jeopardizing the support that they will receive for reelection. And firing a commissioner from the other political party will almost certainly lead to allegations of partisan gamesmanship, especially as commissioners from both parties share responsibility for blunders.
New York’s statutory election law requires that every position, even technical and clerical roles, be filled by a Republican and a Democrat. The resulting duplication hampers efficiency and accountability, raising costs without necessarily improving the level of expertise at the board. The goal of this duplication is to prevent one party from securing an advantage through fraud, but there are better ways to achieve this end, including through contract, ethics oversight, and the imposition of criminal liability. The bipartisan requirement instead creates a thriving spoils system. Because commissioners hire lower-ranked employees, often as thanks for political favors or loyalty, those in junior positions similarly need not fear that their bosses will hold them accountable. In 2010, Village Voice reporter and columnist Tom Robbins characterized BOE as a “Bizarro World,” in which employees with criminal histories and evidence of malfeasance never get fired and sometimes get promoted, concluding: “The safest place to hold a job in New York City—bar none—is the Board of Elections.”
The fruits of this structurally unsound arrangement are evident in nearly every election cycle. Long lines supposedly caused by ballot scanners malfunctioning in humid weather, poorly trained poll workers, illegally purged voter rolls, and improperly tabulated election results are just some of the most blatant problems that New Yorkers have put up with over the past decade. As the state introduces new methods to vote, the city board is ill-equipped to handle these tasks. During the 2020 presidential election, nearly 100,000 city voters received incorrectly printed absentee ballots, rendering them defective. Partially as a result of these inadequacies, the Massachusetts Institute of Technology Election Data and Science Lab’s Election Performance Index consistently ranks New York near the bottom of states for election administration.
Ineptitude trickles down to poll-site operators and coordinators, temporary workers charged with administering Election Day procedures. A 2013 city Board of Investigation report found several instances of poll-worker misconduct, including instances of voters being told to “vote down the line,” or for all the candidates of one party, in order to avoid ballot malfunctions. Likewise, a 2017 report by the city comptroller’s office found that BOE did not engage a sufficient number of poll workers and standbys to ensure that every poll site was fully staffed. Out of 156 sites sampled in the three elections held in 2016, poll workers in 82—more than half—did not follow state and BOE guidelines. Three-quarters of sites, or 118, had insufficient staffing. When all compliance issues were accounted for, a stunning 90% of all poll sites were noncompliant. In the words of the report, “these deficiencies increase the risk that registered voters will not be provided with the opportunity to vote and to have their votes counted.”
Possibilities for Electoral Reform in New York City
In light of the low bar set by the city’s electoral system, there is much room for upside. This section will discuss several possible avenues for reform. Some are practical, urgently needed, and enjoy broad political support; others are experimental, aspirational, and thus less plausible. Readers should note that the reforms to the city Board of Elections and to on-cycle local races, while seemingly more modest, would have immediate and dramatic positive results for a wide swath of New Yorkers. In fact, overhauling BOE might be best understood as a prerequisite for other reform, given that the board’s current arrangement is inadequate to manage even today’s elections.
Reforms to the Board of Elections
Without substantial and long-overdue reforms to the Board of Elections, New York’s elections will almost surely continue to suffer from irregularities and mishaps. While certain structural changes to the board require an amendment to the New York State Constitution, several possible avenues for reform are available through ordinary state legislation. In May, the state senate passed a suite of election-reform measures—including the substance of some of the proposals below—but the state assembly adjourned before calling the legislation for a vote, thus placing these efforts on hold until at least the next legislative session.
Article II, Section 8 of the New York State Constitution requires that all “officers charged with the duty of qualifying voters, or of distributing ballots to voters, or of receiving, recording or counting votes at elections, shall secure equal representation of the two political parties.” Modifying this bipartisan requirement therefore requires a constitutional amendment, which, in turn, involves passage by both houses of the state legislature in two consecutive sessions, followed by a majority vote of the statewide electorate in a ballot initiative. Consequently, amendments would take years to enact and would face several practical and political obstacles, rendering them less effective at achieving short- and intermediate-term change.
The most transformative constitutional reform would restructure BOE on a nonpartisan basis, made up of members with election administration expertise. Whether ostensibly nonpolitical experts would represent an improvement over party-based commissioners is a matter of philosophical preference; there is no ideal answer. But such a measure is likely to face opposition from Republicans, who represent a minority of city voters but command half the commissioner seats. Political opposition may also come from Democratic lawmakers who are closely affiliated with the party committees in their districts and who use the board as a way to reward supporters.
Even with the constitutional requirement, there is still a great measure of flexibility in how to staff the city board. Splitting the composition of commissioners evenly between experts and party officials, for example, may achieve both bipartisan fairness and administrative competency. Democrats might nominate Amber McReynolds, currently a governor of the U.S. Postal Service, who has extensive experience in election administrative issues and voter access initiatives, particularly vote-by-mail. Republicans, for their part, might select election integrity expert and National Review columnist John Fund. While the partisan commissioners may be expected to serve in their capacity full-time, these experts could serve in a part-time capacity so as to allow them to continue in their existing positions. Such a structure is implementable without a constitutional amendment, provided that the two major political parties each select expert commissioners.
There are other ways to amend the state election statute that do not require a constitutional amendment but would quickly achieve meaningful BOE reforms:
- The number of city commissioners should be reduced from 10 to four, or perhaps two. Fewer commissioners would allow for greater accountability, as each would bear a larger share of responsibility for election administration. While a reasonable argument exists for reducing the number to two, four commissioners would allow for a greater range of experience and expertise on the board. (The state election statute sets two commissioners as the default for each county board of elections, though local law can expand this to four.) Regardless, fewer commissioners would represent a major advance toward greater responsibility and accountability.
- The salaries of the discharged commissioners and their staffers could be redirected toward more effective uses, such as procuring more and better-quality ballot scanners and hiring more qualified poll workers. Such a reform would more accurately reflect NYC’s status as a single governmental entity, rather than as a collection of five distinct counties.
- The salaries of the discharged commissioners and their staffers could be redirected toward more effective uses, such as procuring more and better-quality ballot scanners and hiring more qualified poll workers. Such a reform would more accurately reflect NYC’s status as a single governmental entity, rather than as a collection of five distinct counties.
- Election administration expertise should be a job prerequisite for commissioners and staff. Even if the board structure is split between experts and party officials, requiring a credential or prior experience would better enable the agency to manage the complexity of running elections for millions of New Yorkers. Commissioners could also be required to complete coursework or training for election professionals within a short time of accepting their office. Instruction could be administered by the state BOE, or through an approved college-level curriculum. The Network of Schools of Public Policy, Affairs, and Administration, for example, has developed a curricular framework in election administration that can serve as mandatory training for board commissioners and high-level staffers.
- For lower-level positions, posting all open job positions and conducting a merit-based hiring process would likewise facilitate greater transparency and more competent workers.
- For lower-level positions, posting all open job positions and conducting a merit-based hiring process would likewise facilitate greater transparency and more competent workers.
- The state BOE should be required by law to issue a series of rules and standards for local boards, which would facilitate greater administrative competency and consistency across New York State. Despite having a mandate in New York election law to issue instructions and rules to local boards, the state board merely produces rules when statutory changes necessitate updates. Arizona’s election statute, by contrast, requires the secretary of state to produce an election procedure manual before the end of each year preceding a general election that contains the latest legal changes and rules developed by the secretary’s office to ensure accurate, fair, and efficient election administration.
- The exclusive power of county party leaders to nominate qualified commissioners should instead be given to politically accountable officials, such as the mayor. Half-measures could also include allowing party leaders to submit nonbinding recommendations to the mayor for his ultimate nomination. Because the bipartisan requirement would remain for commissioners, the mayor would have to nominate some individuals from the opposition party, but this has precedent in bodies like the U.S. Postal Service’s bipartisan governing board. Such a move would provide an incentive to nominate capable candidates, given that any failures would be at least partially attributable to the nominator.
- Similarly, the mayor and city council should be granted for-cause removal powers over commissioners, which would prevent corruption and ineptitude from continuing unchecked. For example, the mayor can nominate candidates for removal, and the council could require a two-thirds supermajority vote to remove (a simple majority vote might invite politically motivated firings). With more avenues for removal, citizens and organizations could pressure their elected representatives to hold poorly performing commissioners accountable. Necessitating buy-in from both the mayor and the council should spread political risks associated with upsetting county-level partisan leaders across multiple decision-makers.
- For BOE positions that perform tasks that do not fall under the constitutional bipartisan requirement, the statutory bipartisan staffing requirement should be amended and replaced by other means of ensuring employee objectivity and fairness. These might include imposing criminal liability, stricter transparency and ethical oversight, and provisions in an employment contract.
- Poll-worker training should be professionalized and standardized across sites to guarantee a consistent and seamless experience for every voter, regardless of location. This training should cover commonly occurring situations, such as Election Day signatures that do not match the registration book, requests for affidavit ballots, and end-of-day procedures for processing machines and ballots.
- Much of this training should be hands-on, not conducted in a classroom. Workers should not spend time on Election Day learning how voting machines work, when voters wish to cast their ballot in a timely fashion. Poll coordinators also need to be aware of these processes and cannot merely delegate these responsibilities to workers.
Moving New York’s Local Elections On-Cycle
NYC’s local elections are held “off-cycle,” or on odd-numbered years that disconnect them from races for state and federal office. Throughout the 19th century, local elections shifted between off- and on-cycle years numerous times. But in 1894, Article XII of the New York State Constitution was amended, and its language survives today as Article XIII, Section 8, providing that all elections of city officers “shall be held on the Tuesday succeeding the first Monday in November in an odd-numbered year.” To change the local election cycle for New York City, therefore, a constitutional amendment is necessary, but the benefits of on-cycling local races outweigh the burdens of doing so.
Out of all the reforms discussed in this report, moving local elections on-cycle would likely have the single greatest impact in boosting voter turnout. In fact, there is evidence that on-cycling local elections would likely double today’s turnout figures. Requiring voters to make just one trip to the polls—a trip that they likely already make for national elections—reduces the costs of voting. In her review of the impacts of election timing during New York’s years of shifting local election cycles, Anzia finds: “Throughout the period of off-cycle elections, voter turnout in city elections consistently fell far below turnout levels in gubernatorial and presidential elections.”
On-cycle elections may lead to a greater descriptive similarity between the electorate and their representatives—i.e., those in office look more like their constituents. For example, all else being equal, black Americans elect more black representatives in places with on-cycle elections. And higher voter turnout tends to secure greater minority representation in city councils and mayors’ offices, specifically among Latinos and Asian-Americans.
Having more voters from a broader swath of the electorate secures a closer congruence between voter policy preferences and those of their elected officials. Michael Hartney and Sam Hayes, for example, find “clear and consistent evidence that citizens who reside in school district governments that hold on-cycle elections are more likely to be represented by school boards that share their ideological orientation toward fiscal policy making.” In a Manhattan Institute issue brief, Hartney concludes: “When local governments are elected on-cycle, the average American is simply better represented by his local municipal officials.”
Holding municipal elections on the same day as presidential elections would leverage the disproportionate attention given to national news by the press and voters. Partisan national issues provide a stronger impetus for most voters than local-level matters, which tend to be less ideologically driven. For example, even though there was no realistic chance that Donald Trump would win a plurality in New York State, 61.9% of the city’s registered voters—more than 3 million New Yorkers—went to the polls in November 2020. That figure is nearly three times greater than the general election for the mayor a year later, even though Eric Adams’s victory was likewise all but assured.
A unified election calendar would dilute the voting power of special interests. Anzia finds that, all else being equal, school districts with off-cycle elections provide experienced teachers an average of 3% more in base salary, compared with those with on-cycle elections. Hartney similarly writes that candidates backed by teachers’ unions were more likely to win school-board elections in off-years. More moderate candidates would likely benefit from moving elections on-cycle, as a higher voter turnout would likely dilute the inordinate voting power that highly motivated special interests and factions have in low-turnout, off-cycle primaries.
But the benefits of on-cycling elections would extend well beyond participation and representativeness. Separating local and national races means that the city must shoulder the entire cost of its local elections. Consequently, on-cycling would allow the city to piggyback the costs of local elections on the federal funding that it receives to administer national elections.
This should improve efficiency and thereby save tens of millions of dollars annually. Consider that the city BOE currently employs 517 full-time employees, partly because of the need to have elections every year. BOE’s budget for FY 2017, 2019, and 2021, respectively, ranged from $121 million to $173 million to $131 million. Holding elections every other year would theoretically allow BOE to cut its full-time salaried employees and hire more temporary workers. This would limit county party bosses’ ability to use BOE positions as rewards for political allies.
Aside from the opposition of incumbents and special interests that benefit from off-cycle races, the major impediments to on-cycling elections are greater election administration complexity and potential voter confusion. Because more races would be decided on Election Day, ballots would be significantly longer, likely spanning both sides of several pages. The time needed for poll workers to explain directions and for voters to cast their votes for more races will increase. Consequently, more polling sites and better technology would be needed in order to avoid long lines, but those moves would be beneficial regardless. Redesigning New York’s ballot layout could also ameliorate these issues.
In NYC specifically, another major barrier to on-cycle elections stems from the city’s move to permit its more than 800,000 lawful noncitizen residents to vote in local elections. As this report goes to press, it appears that the law enabling noncitizen voting may not take effect. In late June, a state court judge enjoined it, holding that noncitizen voting was unconstitutional per the state constitution, unlawful under the state election statute, and lacked the municipal referendum necessary under the Municipal Home Rule Law. The city has filed an appeal to the intermediate court, and review by the state’s highest court, the Court of Appeals, appears likely. It is not yet known whether Governor Kathy Hochul’s replacement for retiring Chief Judge Janet DiFiore would tilt the odds in favor of upholding the noncitizen voting measure.
But in the event of reversal or changes to state law, it is worth noting the particular difficulties that noncitizen voting poses for on-cycle elections. These new voters must receive a specially designed local-only ballot—or else risk voting in prohibited races, which can subject them to federal criminal sanctions and jeopardies to their immigration status. Hosting local, state, and federal races on the same day would significantly increase the risk of noncitizens voting illegally; preventing that risk entails greater election administration complexity. For example, Drew Spencer, then a fellow at the election-reform think tank FairVote, testified before the city council in 2013: “Takoma Park elected officials seem wary of the possibility of consolidating local elections with congressional elections precisely because doing so might make voting rights for non-citizen residents too challenging to administer.”
As a practical matter, especially given the ineptitude of today’s city BOE, noncitizen voting and off-cycle elections represent mutually exclusive choices. But if the purpose of New York’s effort to enfranchise new voters is to engage more residents in local democracy, it would accomplish that goal far more effectively if it instead opted to move elections on-cycle, not enfranchise noncitizens.
In sum, on-cycle elections secure benefits for voters and the municipal treasury without major downsides. According to Hartney, the evidence “unambiguously shows that the benefits of consolidation far outweigh any demonstrated costs. Consolidation is therefore the reform that does no harm while eradicating several clearly established deficiencies hampering off-cycle local governments.”
Providing More Information to Voters Before and On Election Day
Providing voters with more information about candidates before and on Election Day would help them make higher-quality decisions at the ballot box. This is especially true for local races that may not generate much media coverage and in crowded closed primary races, where party affiliation does not sufficiently capture substantive policy differences. A study by Boudreau, Elmendorf, and MacKenzie reveals that providing a voter information guide and printing endorsements on ballots strengthen the congruence between the policy views of voters and the candidates they select, with the strongest effect seen in voters with low levels of knowledge about local politics.
Proving information-rich signals on ballots is an immediate and low-cost way to guide voter decision-making. This would reduce the likelihood of voters making choices based on irrelevant or prejudicial factors, such as alphabetical order or perceptions of candidate names and their race and gender. Research suggests that providing informational cues on ballots “virtually erases the effects of candidate demographics on voter behavior, even among voters with high levels of racial and gender prejudice.” Thus, informational cues not only can make elections more representative of voters’ preferences, but they can improve fairness, too.
More concretely, Elmendorf and Schleicher propose giving high-level executives, such as the mayor or governor, the power to make on-ballot endorsements. They also suggest printing the existing partisan balance in government on ballots, which would enable voters to connect recent news to the party in power and to better select whether to continue or to change the existing regime.
The incumbent mayor might, for example, designate his preferred successor, which would signal to voters that such a candidate represented continuity with the current administration. The comptroller’s endorsement could likewise signal a candidate who espouses fiscal responsibility. While this might appear to strengthen the advantages of incumbency, the endorsement is only as strong as the endorser. Polls leading up to the 2021 primary revealed that Mayor Bill de Blasio’s unpopularity made his endorsement a political liability, not an asset. Moreover, by more closely linking the efforts of past and future mayoralties, on-ballot endorsements may nudge incumbent mayors to take on long-term projects whose benefits will be felt only after their administration ends.
Going further, organizations like local newspapers, unions, and civic groups might provide endorsements. Given the affiliations other than party that voters find salient, these endorsements can better ensure that voters’ political preferences match those for whom they cast a ballot. Allowing for endorsements by groups focused exclusively on local issues would help correct for the lack of relevance of national party brands for local policy. It would also strengthen the political power of local intermediary organizations, providing residents with a stronger incentive to take part in such groups.
The most important administrative issues inherent in this proposal involve thresholds and gatekeeping: Which officeholders and groups should be allowed to make on-ballot endorsements? How many endorsements should be possible, so as not to overwhelm voters and lengthen ballots excessively? By what process should endorsers qualify?
These are practical questions that can be answered adequately only through consultation with election administration officials and input from various stakeholders. Answers will vary from place to place. But one way to administer on-ballot endorsements could be as follows: some high-level city officeholders, such as the mayor and the comptroller, could endorse as-of-right. Organizations could qualify by obtaining a sufficient number of signatures of registered voters, in a way similar to candidates seeking to get their names on the ballot.
Those seeking to endorse city council candidates would require a lower threshold to qualify than, say, candidates for mayor. To avoid an unpopular political leader or organization from strategically endorsing a political opponent, candidates should be given the option to reject the endorsement.
A cap on the number of endorsements could limit the space that endorsements occupy on the ballot, reduce the chances of adding more pages to the ballot, and equalize the perceived effect of the endorsements (a candidate with 10 lower-impact endorsements, e.g., may appear more impressive to voters than a single endorsement from the New York Times). Candidates might, for example, be given a maximum of four endorsements. If they receive more than that, they would be required to choose which among the four they would like to have printed. This would signal the candidates’ priorities to voters. Endorsers might be given the chance to rank candidates; if a candidate ranked first rejects the endorsement, the second-choice candidate would then be given the option, and so on.
California already attempts to provide voters some information by allowing primary candidates to include a three-word designation of their choosing on the ballot under their names. While this gives candidates the chance to appeal directly to voters through an immediate cue, it has been the subject of gamesmanship and perennial lawsuits. A standardized endorsement signal from third parties would avoid these issues.
In addition to these immediate signals at the ballot box, the New York City Charter requires that the Campaign Finance Board produce a voter information guide for local races. This guide effectively gives voters all that they need to know in the weeks leading up to the election. A copy is mailed to each registered voter; but the city could also nudge residents to be aware of this important source of information by mentioning it briefly on recordings of the 311 system’s greetings, on municipal tax statements, and by advertising it on city bus stops and subways, along with QR codes for download.
In California, the secretary of state’s office issues an official information guide that likewise lays out candidates’ statements; during recalls, it includes the proponents’ statement and the elected official’s answer. Nonprofit groups such as the League of Women Voters, the California State Library, and the Common Knowledge Group collaborate to issue a nonpartisan Easy Voter Guide that describes the American system of governance and the propositions on the ballot. Similar collaborations in New York could yield benefits in mobilizing voters sympathetic to such groups.
Abolish New York’s Redundant and Low-Value Political Offices
Some elected offices in NYC produce little value, often serving the interests of their holders, not voters. In particular, the positions of public advocate and borough are largely ceremonial roles that provide opportunities for politicians to make a name for themselves before running for higher office. These positions do not provide practical executive experience, making their holders particularly ill-suited for the positions that they later seek to win. Consequently, the offices of public advocate and borough president should be abolished.
New York’s borough presidents were once highly influential officials, members of the powerful but long-defunct Board of Estimate, which had the authority to shape New York’s budget, allocate city contracts, and vote on land-use issues. After the U.S. Supreme Court declared the Board of Estimate unconstitutional in 1989, borough presidents remained, though stripped of most of their powers. Today, they largely serve as a liaison to city hall for borough-specific concerns—something that also falls within the purview of councilmembers’ duties. Borough presidents also disburse a portion of the city’s capital budget (5%, or about $4 billion in 2020) to fund projects in their respective boroughs. Their offices have an expense budget totaling more than $25 million of the approximately $100 billion city budget, a relatively modest figure, but this could instead fund approximately $500,000 in participatory budgeting for each city council district.
The public advocate is technically the second-highest-ranking city officer and fills the amorphous role of ombudsman between the public and city government, tasked with overseeing city agencies’ performance and investigating citizen complaints. In reality, the relatively powerless position allows for public criticism without the need to grapple with practical trade-offs. Jumaane Williams, the current public advocate, is the city’s foremost critic of the New York Police Department, having personally defied police authority at protests and threatened to use his (doubtful) authority under the city charter to freeze the city’s ability to collect revenue that would fund the department. He most recently lost the Democratic primary race for governor of New York.
Former mayor Bill de Blasio, among the most unpopular political leaders in the city’s recent history, was previously the public advocate before becoming mayor. This was also the case for Letitia James, New York’s current attorney general, who garnered national headlines by quixotically prosecuting a climate-change-related civil case against ExxonMobil on behalf of New York, only to receive the court’s rebuke for making “hyperbolic” claims. In sum, the office is suited for political grandstanding, not developing the acumen and competencies necessary to become an effective and pragmatic leader.
City leaders from both parties have, in fact, called to abolish the public advocate’s office. In 2009, for example, the city council agreed to reduce the office’s budget from $2.9 million to $1.8 million. That same year, Mayor Michael Bloomberg called for eliminating the position, saying to a local editorial board: “You should get rid of the public advocate. It’s a total waste of everybody’s money.”
These sentiments resurfaced among Democratic councilmembers nearly a decade later. A trio of Bronx members were among the sponsors of a bill that would have eliminated the office. This is noteworthy because the Bronx is by far the city’s lowest-income borough; these representatives did not believe that the public advocate produced sufficient results for their constituents to justify continuation, despite the ostensible purpose of the position to represent “the needs of New York’s most vulnerable and underserved.” In an interview, councilmember Ruben Diaz Sr. said that the “stepping-stone position” was “unnecessary” and that its holders “don’t do nothing.”
Similar calls have been directed at the borough presidencies. During a 2010 hearing to amend the city charter, State Senator Pedro Espada called for the abolition of the office, despite having run for the Bronx’s presidency in 2001. Also in 2010, Republican David Casavis ran for Manhattan’s presidency; his primary campaign promise was to abolish borough presidencies.
Eliminating these positions would streamline city government and save resources. New York’s strong mayor is a singular figure whom voters can hold responsible across all five boroughs. The city comptroller is well suited to hold the mayor and city council accountable for greater fiscal responsibility, rendering the need for two watchdogs unnecessary. Members of the city council—who represent districts within a single borough—should simply represent the appropriate concerns that the public advocate and borough president would espouse.
The Fork in the Road Ahead for City Council Elections: Two-Party or Multiparty Democracy?
Thus far, the suggestions in this report have largely assumed the continued existence of the current electoral system for New York City Council races: ranked-choice voting in closed primaries; and single-choice plurality voting in general elections. While there are gains to be made within the confines of that system, the city would be better off pursuing structural reform to council elections. New York’s history with proportional representation shows that a different electoral system can result in a transformative shift in the council’s importance and relevance to city residents.
At the start, it is necessary to point out that different electoral systems are associated with either a two-party system, familiar to Americans, or a multiparty system, more commonly found in other countries, in which several parties often win seats without a majority and then form legislative coalitions. Each contains trade-offs, and the perceived desirability of two-party versus multiparty rule will depend on voters and elected officials’ philosophical preferences and risk tolerances.
But, as discussed earlier, local politics are shaped by different dynamics from those in play at a national level—both in terms of the political identities of voters and the substantive issues. This makes it difficult to assemble a local party platform that is broad in scope and consistent in approach, which is compounded when national-level party identity interferes with local deliberation. At the same time, local government is well suited for innovative political solutions, in terms of election structure and the policies that local governments adopt. Cities, even more than states, can serve as the “laboratories of democracy” that pave the way for solutions at other levels of government.
Those who wish to vote their preferences uncompromised by pragmatic political considerations will likely prefer a system of multiparty democracy. But under such an arrangement, securing seats for one’s favorite party in a legislative body does not inexorably translate into enactment of one’s preferred policies. Without a single party in power, multiple parties must compromise to form governing coalitions, such that no voter knows how the results of an election will translate into actual policy outcomes. Consequently, it becomes more difficult to hold any particular party accountable for successes or failures—one member of the coalition can always lay blame on another. When the coalition cannot reach consensus on a suite of policies, ineffective gridlock can persist until the next election.
Multiparty systems may work differently in local settings from the way they work at the national level. Smaller local parties, each focused on narrower issue subsets, can attract voters predominantly concerned with those issues and then form coalitions that would approximate the population’s priority of concerns. Thus, smaller parties need not take platforms so broad as to invite intra-party factionalism.
The disadvantages of a multiparty system might not be felt as strongly in the context of New York’s strong-form mayoral government. The mayor is independently accountable to voters by virtue of a separate election. His successes and failures—say, in reducing crime, promoting economic development, and providing high-quality city services—are largely separate from the goings-on in the city council. Because the city council is institutionally weak, legislative accountability is arguably less important. It is more important that the council produce legislation marked by thoughtful deliberation and buy-in from multiple constituencies. During the council’s period of proportional representation, multiple parties, contested elections, and raucous debates did not prevent the body from passing legislation, very often with large majorities.
A two-party system, on the other hand, gives voters greater clarity on how those in the winning party intend to implement campaign promises. In theory, voters can hold the governing party directly responsible for failures and successes at the next election. When the mayor and the city council are from the same party, as in New York, it is also easier for the mayor to secure the legislation that can accomplish his agenda.
Yet within two-party systems, factions necessarily vie for control over the party’s agenda, rendering it susceptible to takeover by a faction that is out of step with the majority of city voters. This arguably occurred when the most progressive members of New York’s city council agreed to redistribute $1 billion from the NYPD’s budget toward other city services; a number of black councilmembers opposed these measures as “colonialization” and “political gentrification.”
Minor parties might vie for influence on single issues such as the environment or labor, but in two-party systems, they don’t stand a plausible chance at winning a seat. Voters may feel that a lack of meaningful choice makes participation fruitless or even illusory, dampening turnout and narrowing the field of political competition.
Opening up political competition within a two-party context might, however, allow alternative majority coalitions to form. Concretely, that could imply a situation where incumbent moderate Democrats, fearful of losing races to Republican challengers, eschew the faction on their left. Instead, these moderates could partner with Republicans on a multidimensional platform of lower taxation, higher police and public safety funding, and streamlined city agencies and service provision. In this way, the new alternative majority would deliver a more broadly popular platform than an exclusively Democratic-leftist one, which would permit such moderates to run against opponents on a different basis from those who supported a coalition with progressives.
Finally, to achieve major structural reform, an amendment to the city charter and changes to state laws would be necessary. Amending the charter typically involves the mayor or the city council establishing a Charter Revision Commission, either for a specific set of proposals or for a comprehensive “overhaul.” The commission conducts public hearings and places specific proposals for an up or down vote at the next general or special elections. Replacing closed primaries can either be achieved through a city charter amendment (which was attempted in 2003) or by amending the state’s election statute.
Option 1: Retain Two Parties with a “Final-Four” Nonpartisan Primary and Ranked-Choice Voting
The first transformational option, one continuous with the current two-party electoral system, would retain single-member council districts and introduce a variant of “Final Five” voting (FFV), a proposal popularized by Katherine Gehl and Michael Porter. This system entails the following structure: (1) a “nonpartisan primary,” in which all candidates appear, regardless of party, in a preliminary election open to all registered voters, who use non-ranked voting; (2) the top four or five vote-getters in the primary, regardless of party, advance to a general election that uses ranked-choice voting to select a single winner. But because FFV presents the same informational challenge as a fully nonpartisan election, my proposal suggests supplementing party labels on ballots, designated by internal committees or processes, along with the aforementioned on-ballot endorsements made by executives and local interest groups. I further propose that four (not five) candidates advance to the general election, in which each party can endorse only one candidate for the party label.
Before addressing the substance of this proposal, it is necessary to address a terminological point. The colloquial name for FFV’s preliminary election is a “nonpartisan primary,” but this term is somewhat mischaracterized. Traditional primaries (whether open, closed, or blanket), by definition, are processes by which parties select their nominees. A so-called nonpartisan primary, by contrast, does not select party nominees at all. It merely winnows down the (frequently large) number of candidates who run initially to a smaller, more manageable subset in the decisive general election. The term “qualifying round” better captures this concept.
A qualifying round election comes with some advantages over hosting a single general election. Most obviously, it avoids overloading general election voters with too many candidate names. It also appeals to those who believe that having two separate elections allows voters to refine their choices through a longer deliberative process and thereby achieve the “right” results. While primaries are frequently centered on candidates, political scientist Jack Santucci makes a strong argument for framing primaries as a means of curating the issues that will be relevant in that election cycle: “Primaries matter not for candidate ‘extremity’—but for the issue substance of electoral competition.” This played out in the June 2021 New York City Democratic primary. Large swaths of Democratic voters disapproved of the progressive push to defund the NYPD, and the issue featured prominently in the June primary, which likely helped Eric Adams win the nomination.
Some jurisdictions already employ variations of this system. For example, under California’s “top-two” system—which is not technically FFV but uses a similar nonpartisan primary—all candidates, regardless of party, appear on the same primary ballot, and the top two vote-getters advance to the general election. Alaska’s final-four option expands the number of candidates who proceed to the general election to four, which has several advantages, such as allowing for greater voter choice among members of party factions and third parties. In both cases, several members from the same party can compete against one another in the general election. California uses single-choice voting in both the primary and general election, whereas Alaska uses RCV in general elections, rendering its system a closer fit for New York, which uses RCV as well, though in primaries only. The use of RCV in primaries is relatively unusual.
How many candidates, then, should advance to the general election in New York? To answer that, it is necessary to consider the potential for exhausted ballots, or ballots that do not count in the final tally because their selections have already been eliminated. In New York’s first experience with RCV, about 15% of mayoral ballots were exhausted, and 55% of voters ranked four or fewer candidates for the mayoral race. Approximately half of voters ranked three or more city council candidates; 30% ranked one, and 20% ranked two.
Unsurprisingly, many voters ranked few candidates in council races, given their lower-information and lower-stakes nature. Voters who opted not to rank more than one candidate felt that they did not have enough information about other candidates, especially for fourth and fifth choices, and especially in uncompetitive districts. Without on-ballot information cues, this problem will likely persist.
In sum, given that a minority of voters ranked fewer than five candidates in both the mayoral and city council races last year, New York electoral reformers should consider allowing the top four qualifying-round candidates to advance to the general election, where voters would be given a maximum of four ranked choices (if the city adopts RCV for general elections). A top-four primary for council races in New York would adequately balance the need to enable diverse political voices and backgrounds with the ability of voters to remember and judge candidates’ positions. By contrast, a two-winner runoff would limit political competition to a progressive and a center-left candidate in many districts. A maximum of three ranked choices, used in Minneapolis, may likewise overly constrain mayoral primary voters and is now uncommon relative to four- or five-winner options, according to Drutman and Strano. Voters may find it easier to learn about and rank four candidates, rather than five, rendering them less likely to give up prematurely and potentially exhaust their ballot.
As for party labels in qualifying round elections, some jurisdictions retain them next to candidates’ names, whereas others eliminate party designations altogether. The major disadvantage of removal is the consequent lack of information for poorly informed voters. As mentioned previously, candidate names, standing alone on the ballot, can lead to voters making choices on the basis of their perceptions of candidates’ race, gender, or even alphabetical order. Party labels provide an important signal for voter decision-making on Election Day. Schleicher states: “The use of nonpartisan elections takes the little information that the heuristic of party provides and removes it entirely.”
Keeping party labels in qualifying rounds introduces a challenge: Who decides what a candidate’s label should be? On the one hand, allowing political committees or some other internal party mechanism to select the party’s nominee would represent a major boost to the power of party insiders. On the other hand, some states—notably, Washington—allow candidates to list the political party of their preference, which does not constitute an endorsement by that party. While the Supreme Court has declared this practice constitutional, it dilutes the party’s ability to signal that the candidate aligns with its platform. The party may not want anything to do with a candidate who nonetheless affiliates with it on a ballot. Voters, likewise, cannot be sure that the candidate with a given party label is actually in line with that party’s values and policy proposals. Weakening parties to this extent may lead to a loss of party cohesion within the city council, raising the prospect of unstable coalitions and unruly lawmakers who need not heed party discipline.
Because of the vital importance of information on the ballot and the need for parties to overcome coordination problems in the legislature, the better choice is to allow internal party mechanisms to select one candidate to receive its endorsement for both primary and general elections. This would preserve the current incentive structure, in which candidates promise a degree of loyalty to the party’s policy agenda, in exchange for the party’s organizational support. With greater political competition, the risks associated with more powerful insiders abate somewhat, as parties, at least in theory, have an incentive to nominate candidates with the greatest chances of winning in a general election.
Critically, the aforementioned proposal for on-ballot endorsements would provide additional means of informing voters on Election Day, effectively supplementing the party cue. In fact, such endorsements would signal more than mere party identifiers, which cannot distinguish substantive differences between candidates within the same party. The quantity and reputation of endorsers would enable even uninformed voters to identity which candidate is likely to align with their values and policy preferences. Longer-shot, unorthodox, and first-time candidates will have a greater incentive to participate in a primary that is open to all voters and advances the top four vote-getters, not merely a single-party designate.
The ranked-choice general election should certainly include party labels next to candidate names, and each political party would select only one candidate to receive the party designation. This would be beneficial for two reasons. First, it would preserve the party mechanism’s ability to organize a political program for which voters can hold the party responsible. While a nonpartisan primary enables smaller parties and independent candidates to exert greater influence relative to a closed primary, the Democratic Party will likely continue to enjoy a majority in the council, which requires linking it to candidates in a way that promotes party-based accountability. The risks of having committees or caucuses select each party’s nominee are also lower in general elections. Voters have already curated their top-four choices through the primary, and the pressure to win more competitive general elections should incentivize parties to select philosophically compatible candidates who have the best chances of winning. Such an approach also accords with Elmendorf and Schleicher’s suggestion that parties “should be treated as private associations with the designated public function of enabling a low-information electorate to steer public policy and hold the governing coalition accountable for results.”
Second, party labels would perpetuate New York’s fusion voting system, which, in general elections, allows minor parties to nominate major party candidates or those of their own choosing, and in which voters can select the candidate from the party that most closely aligns with their preferences. Elected officials and organizations that qualified to provide endorsements in the primary election could automatically qualify to endorse candidates in the general election, giving voters more information upon which to select their rankings.
The prospect of voter confusion due to too many candidates on the ballot becomes less of a concern, thanks to on-ballot endorsements. To address this problem further, the number of signatures needed to appear in the city council primary can be increased. This has some precedent; in 2020, Governor Cuomo signed a budget that included tripling the number of signatures necessary to secure a candidate’s place on the ballot for statewide office, from 15,000 to 45,000. NYC’s signature requirement for city council primaries halved from 900 to 450 because of Covid-related concerns, but this can be substantially increased to keep out candidates whose trivial chance of being elected fails to outweigh the benefit of additional voter choice.
Option 2: Reintroduce Proportional Representation Through the Single Transferable Vote
The second electoral proposal would allow for proportional representation (PR) by electing several councilmembers from a smaller number of geographically expanded districts. This system would use the Single Transferable Vote (STV), the multi-winner version of ranked-choice voting. PR-based electoral systems aim to reflect the subgroups that constitute the voting population within the composition of seats in a representative body. Multimember seats allow smaller groups of voters to secure representation than is possible under single-member districts. For example, in a multimember district, voters troubled about environmental matters could rally behind a Green Party candidate and elect that candidate to one of the seats; in a single-member district, by contrast, a major-party candidate would be almost certain to win.
The number of candidates elected in each district can vary; as more seats are added and the geographical boundaries enlarged, a smaller percentage of the vote is needed to secure a seat. Therefore, the greater the number of seats per district, the more likely that the result will reflect smaller subgroups in the voting population; that is, larger, more numerous districts obtain PR more closely. Table 1 sets out what minimum percentage of the final vote each winning candidate must receive as the number of seats in the district increases:
Minimum Vote Share Required in Multimember Districts
The optimal number of candidates per district will vary by place and context; in general, that decision will depend on balancing a few key practical and political considerations. As more seats are added to a district, the larger the geographical bounds of the district will have to become. Local representation aims to reflect the descriptive makeup and preferences of the voters within particular neighborhoods, so larger districts may dilute the purpose and benefits of local representation.
With New York’s relatively large council of 51 members, three- to five-member districts may be small enough to encompass proximate neighborhoods with similar demographic compositions. During New York’s historical period of STV, the number of seats per district changed with voter turnout; this should not be repeated. Each district should be drawn and fixed with a consistent number of councilmembers. Moreover, because STV uses voters’ ranked preferences to elect several candidates, it would eliminate the need for primary or qualifying round elections. A single general election in the fall using STV would place all qualified candidates on the ballot, regardless of party, and all registered voters would be eligible to cast a ballot.
Historically, more seats per district troubled those concerned about communists gaining power. Critics of PR from the left have also claimed that the system enables radical right parties to gain a foothold in government, often in uneasy coalitions, such as New Zealand’s 2017 shotgun marriage between the Labor Party and New Zealand First, a nationalist party. In NYC, however, the electorate’s composition reduces the likelihood of far-right candidates winning seats; in any event, the number of seats per district can be set lower to further reduce such risk.
As for how STV works on Election Day, voters rank their preferred candidates as they did in last year’s primary. In a four-member district, for example, they could rank up to four candidates. On-ballot endorsements, including from parties, can guide voter decision-making and link candidates to party agendas. A vote quota (commensurate to the percentages in the table above) is established to win a seat: if a candidate’s first-choice votes meet or exceed the quota, that candidate wins a seat. Any “surplus” votes above the quota are distributed proportionally to the second-choice candidates selected by the winning candidate’s supporters. (If several candidates cross the quota threshold, the order will start with the largest vote-getter.) If there is at least one additional seat to be filled with no candidate over the quota, the last-place candidate is eliminated, whose votes are redistributed among the candidate’s second choices. This process of elimination and redistribution repeats until a candidate reaches the threshold, until all remaining seats are filled.
There are several rules to allocate surplus votes for candidates who receive votes over the quota. While the details are best left to local decision-makers, it is worth demonstrating how this works with an example of a simple “fractional” surplus transfer using the second-choice rankings of the winner’s supporters. To avoid the issue of which votes are surplus, all the ballots that selected the winner are transferred, but at a reduced value, to equal the number of surplus votes, and allocated proportionately to their second choices. After the surplus votes have been allocated, if all candidates have met the winning threshold, the election ends; if not, the next-lowest-ranked candidate is eliminated, and the process repeats.
This system is very different from the “block preferential” style of RCV used in Utah, which appears similar to STV on the surface but leads to wildly different outcomes. Block preferential voting does not obtain proportional representation; it essentially acts as a series of sequential single-winner RCV races. After a winner is selected, all his supporters’ second-choice ballots are allocated, not just the surplus, which, given the likely similarities between voters’ first- and second-choice candidates, has the effect of favoring majorities.
An example might be helpful to illustrate how the single transferable vote would work in practice in a multimember district. I avoid using numbers, for the sake of simplicity. Imagine that three candidates are to be elected in a multimember city council district. Five candidates are running: Adams, Bloomberg, de Blasio, Giuliani, and La Guardia. On Election Day, enough voters rank La Guardia and Bloomberg first to win them seats, with La Guardia coming in first. Both now have surplus votes that must be allocated. La Guardia’s ballots are scaled down to the number of his surplus votes and distributed proportionately to his voters’ second choices. Most of these surplus votes go to Adams, but they aren’t enough to push him over the threshold. Bloomberg’s surplus votes then undergo the same process, and after being redistributed to his voters’ next-highest choices, Adams gets closer but does not yet have enough. The candidate with the fewest votes, de Blasio, now gets eliminated, and his votes are distributed to his voters’ next-highest choices. After that distribution, Adams secures enough votes to cross the threshold and win, and Giuliani is eliminated.
While PR may sound like a novel or radical proposal, it would, in fact, not be the first time that NYC has used such a system. As mentioned in the introduction, New York employed STV for city council elections from 1937 to 1947. It almost immediately produced benefits compared with the system it replaced, which largely mirrored today’s single-member districts, chosen by plurality.
Most important, by ending the hegemony of the Democratic Party, PR boosted political competition and more accurately reflected voters’ preferences. This did not exclusively favor conservatives. Democrats retained a majority, but the American Labor Party (ALP), a local party associated with the socialist movement, quickly became the second largest in the council, with Republicans in third place. Other political parties also gained seats, such as City Fusion (which propelled Mayor La Guardia to victory in 1933), the Insurgent Democrats, and Citizens Non-Partisan; though these parties were not always cohesive or internally strong, at least four were represented throughout New York’s period of PR.
The proportionality gap, used in FairVote analysis of STV in New York, also shrank, indicating a more representative local legislature. This measure effectively captures the total difference between the percentage of the vote that each party receives and the percentage of seats that the party obtains in the representative body. According to FairVote’s report, the five aldermanic elections preceding STV’s implementation averaged a proportionality gap of 14%; by the time STV was abolished, that had dropped to a mere average of 1.8%. That meant that STV produced a closer match between voter preferences and candidate positions. With a representative body comprising members with sharply divergent viewpoints, from conservative Republicans to avowed socialists, lively, even chaotic, council meetings turned into popular spectator events, and radio broadcasts from the chamber attracted an estimated—and astounding—750,000 listeners.
Despite such competition and political diversity, legislation still managed to get passed. In the 1946–47 term, 80% of bills passed unanimously, and consensus-building became a regular part of council business. Some of the laws passed during the city’s experiment with PR included the creation of a central traffic commission, a smoke-control bill, and a rent-control regime. In sum, reducing single-party dominance and introducing electoral contestation achieved the fruits of competition: a higher quality and quantity of ideas, better governance, and energized democratic engagement.
Why, then, did New York voters opt to repeal STV-based PR in 1947—even though they had voted to retain it three times in the preceding decade? The conventional wisdom has long been that the rising fears of communism in the U.S., culminating in the postwar Red Scare, turned voters away from multiparty local politics. Communists had, after all, been elected to the council from 1941 onward, and changing geopolitical dynamics meant that they could not remain.
But Santucci has convincingly demonstrated that there was far more to the story of STV’s demise in New York. For starters, Republicans had not supported the repeal of PR in its early years because ALP’s cross-endorsement allowed Republicans to keep control of the powerful Board of Estimate (which was subsequently dissolved in 1989). In the mid-1940s, ALP split into two factions based on issues of race and urban redevelopment, which destabilized the council and ended the unified party cross-endorsement for Republicans in board elections. Seeing in ALP’s collapse a unique opportunity to coalesce all opposition under its party banner, Republicans turned against PR. Newspapers and business interests followed, which, when combined with fears of communism, led to PR’s repeal.
Today, a New York City Council elected through proportional representation might well look similar to the one from a century ago. The Democratic Socialist wing of the Democratic Party may, for instance, splinter to form a leftist party, largely mirroring ALP. In turn, the Republican Party might lose its status as the default opposition party, replaced by an assemblage of smaller parties that span a number of issues and viewpoints. Such an environment would present minor parties with a challenge and an opportunity; they would need to craft policies that are distinctive yet resonate with enough voters to secure electoral wins. Another possibility includes moderate Democrats collaborating with Republicans to a greater extent than today.
Reintroducing multimember districts would yield a supplemental benefit in limiting the practice of aldermanic privilege. Several councilmembers from the same district could not be expected each to wield a veto—doing so risks halting new development across the city. A modified aldermanic regime would likely require a majority of members to block a rezoning, if not end the veto altogether. Multi-councilmember districts would thus increase transaction and organizational costs for housing development opponents.
The main impediment to the implementation of STV today would likely be a combination of Democratic unwillingness to change a system that has long worked in their favor and Republican hesitancy to embrace perceived electoral innovations. Leaders within the Democratic Party, for instance, may fear splitting into fiscally conservative and liberal parties. Republicans may fear losing their party’s status as the exclusive opposition. Yet there are reasons to suggest that STV would serve both major parties’ interests. The Democratic Party is likely to retain a majority of council seats, even under PR (as it did from 1937 through 1947), and allowing its socialist faction to splinter into a separate party would alleviate the institutional pressure to accommodate leftist policies. Republicans already suffer from structural disadvantages, such as a closed primary, so electoral reform represents the prospect of political upside.
In thinking through avenues of PR-based reform, it is important to heed the lessons of New York’s experience with, and ultimate repeal of, STV. The two major parties have incentives to combat proportional representation and thereby eliminate political competition from smaller parties. A lack of significant mechanisms for parties to maintain cohesion between their supporters, candidates, and agenda risks a party split like ALP’s. In turn, that opens the door for major parties to gain enough momentum to see through repeal. These risks can be ameliorated by, for example, allowing each party to give one candidate an on-ballot endorsement, thereby providing voters information to guide their rankings while also allowing a candidate-based electoral system. For policymakers, the key takeaway is the need to balance party cohesion with measures that would strengthen candidates’ power to win elections without a party’s backing. Santucci captures this warning well:
For reformers, two-party politics is an overarching constraint. This limits adoption paths at the national level, directs reform energy into the states (to win “demonstration cases”), and makes it tempting to design reforms that attack party politics in principle. In short, two-party politics pits reform against parties. But reform is not a party. Post-reform change in a party system can expose problems with non-partisan rules, leading to frustration in opposing camps. A reform that permits candidates to defy their own parties does not sit well with competition between parties.
New York City’s elections are ripe for improvement, and New Yorkers deserve better. If the city is to adapt better to a rapidly changing environment—from Covid-19, to macroeconomic instability, to stiffer competition from cities in the Sun Belt—its electoral system should encourage the dynamism that flows from political competition. Innovative policy options and local political coalitions should be given the opportunity to shape the New York of the mid-21st century.
Fundamentally, local democracy risks losing its perception of legitimacy if there are not meaningful opportunities to compete against the governing regime. As Drutman explains:
Democracy requires both the magnanimity of the winners and the consent of losers. The losers must believe that no loss is permanent and that the process is fair enough that they can regroup and build new coalitions to win the next election. The winners must accept that any win is also temporary and restrain from using their majority powers to give themselves any permanent advantages. The continuation of democracy depends on a shared sense of a fair process.
The ideas and proposals in this report represent a starting point for voters and political leaders to build fair and inclusive electoral systems, and are certainly not the last word. While Election Day quickly passes, the ramifications of elections—and the way they are structured—affect the everyday lives of New Yorkers. In putting forward moderate, politically feasible, policy options like BOE reform and on-cycling local elections, this report seeks to identify areas where broad consensus could yield tangible and immediate policy changes. And by proposing different electoral systems for the city council, it aims to provide voters and elected officials with a sense of the possibilities that can be achieved, given sufficient political will.
Electoral reform comes in many varieties, but at its core, it is not about merely the optimization of a technical problem. It aims, instead, at achieving justice—giving to citizens what is due to them in a democratic society. In the words of Guinier:
Election reform, including proportional representation, is not however, primarily about electoral rules. . . . It is about transforming how power itself is exercised and shared. It is about opening up a different kind of conversation, during and following elections, because elections become a forum for voters choosing their representatives and expressing their ideas. It is about giving people in motion an incentive to stay in motion. It is about giving citizens their due.
The most important recommendation in this report, therefore, is: voters and city leaders should embrace electoral reforms and expend the effort necessary to see them through.
Appendix: Party-List Proportional Representation as an Alternative to Obtaining Proportional Representation
In the main document, I proposed two electoral systems that retained ranked-choice voting, largely on the pragmatic grounds that New York has already invested a great deal of expense and effort to introduce and popularize its single-winner form (also known as the alternative vote). The Single Transferable Vote (STV), a multi-winner variant of RCV, can obtain proportional representation without major disruption to the mechanism that New Yorkers have already begun using, at least in primaries.
But proportional representation can be obtained through alternatives to STV. In fact, list-based methods account for 80% of PR systems worldwide. Unlike STV-based PR, in which voters select candidates, list-based proportional representation counts votes at the party level, and parties receive seats in proportion to their share of the vote.
In list-based systems, internal party mechanisms select the names that appear on a ballot list. This helps ensure legislative cohesion and party discipline, once elected. Depending on the variant of list-based system, voters may choose a party or a candidate. If they select a candidate, their choices determine the order of who gets elected from the party’s list. This makes list-based PR systems fundamentally different from STV’s candidate-based nature. As Santucci states: “Elections in list systems are fundamentally contests among parties.”
Party-list systems are foreign to most Americans, so it is worth briefly describing some ways in which they work. The simplest method, called a closed-list system, has voters simply select a party—not a candidate—on Election Day. Internal party processes would then select the representatives who are sent to the legislature, in proportion to the party’s share of the vote. But closed-list systems may come at the cost of the neighborhood-level local representation and candidate choice that New Yorkers expect. Parties might not win enough seats to ensure that popular candidates in particular neighborhoods get elected to represent those constituencies, or party leadership may simply prioritize insiders instead of local figures.
Open-list PR systems, by contrast, grant voters influence over the selection of party candidates who are sent to the legislature. The party committee or some other internal mechanism selects the candidates who appear under the party’s label on the ballot (and, in some systems, the candidates’ order of appearance). Voters would then select their preferred candidate under the party heading. For example, in the version of open-list PR championed by Santucci, a single, non-ranked vote counts for both the candidate and the candidate’s party. Votes cast for any of a party’s candidates are summed up as the party vote, and seats are allocated to each party based on its proportion of the total vote. Individual candidates who receive the most votes from the party’s list are elected first, and then in descending order, until all the party’s seats are filled.
Eliminating the need for ranked-choice voting brings the benefit of simplicity, especially for first-time and low-information voters who may vote erroneously under RCV or not fill out all their choices. Open-list PR, for example, would use a ballot with several names under each party; the same names can appear under several parties, in line with fusion voting. A voter could, for instance, select a major-party candidate who also appears under a minor-party banner, as the illustration below demonstrates, with Benjamin Cardozo appearing on both the Democratic and Working Families line (Figure A1). Smaller parties might opt to field a common slate of candidates on a “joint list,” whereby votes for the candidates under either party would count toward the joint-party common slate.
Sample Open-List PR Ballot 
Another possibility is a mixed-member proportional (MMP) system, which allows for voters to elect representatives in single-member districts and for party committees to elect some members. Directly elected, district-level seats constitute only a portion of the total. The remaining fraction of “compensatory” seats is left to balance the composition of the body in accordance with each party’s share of the vote. MMP is, in some ways, a compromise between the current system of single-member district representation and PR. New Zealand and Germany utilize this system in at least some of their national elections.
With MMP, because only a certain proportion of council seats would be elected directly by voters in their electorate, existing district boundaries in NYC would necessarily have to expand, even if a 51-member council remains. One possibility might involve 40 directly elected seats and 11 party-elected seats, but obviously, there are many reasonable configurations.
MMP comes in two main varieties: two-vote and one-vote. In its more commonly used form, MMP gives voters two (non-ranked) votes, one for their preferred local candidate and another for a party. In each district, the candidate with the most votes wins via plurality. The party votes are used to allocate compensatory seats among the parties, which then select, from a closed party list, as many representatives as are necessary to match the party’s share of the vote. A single- vote option would function like the open-list PR proposal, counting for both candidate and party, with allocation of compensatory seats proceeding in the same way.
A single-vote variant of MMP would also work within New York’s fusion voting system. As with the open-list PR example above, a candidate’s name could appear under several party labels. If a voter selects a major-party candidate under a minor-party banner, for example, the candidate would receive the individual vote, and the minor party would receive the party vote. If minor-party candidates secure enough votes across all districts for the party to obtain representation in the council, the minor party would select a representative to occupy a compensatory seat from its closed list.
Finally, because party committees or caucuses would select a number of seats on the council, MMP might encourage greater competition within parties as well, though this is admittedly speculative. Insiders might, for example, battle for control of the party’s policy and messaging priorities and voter outreach efforts.
About the Author
John Ketcham is a fellow at the Manhattan Institute, working primarily on state & local policy and legal policy. He graduated magna cum laude from Harvard Law School in 2021 and earned a BS in management information systems from Fordham University as valedictorian. While in law school, John served as Treasurer of the Harvard Federalist Society and Managing Editor of the Harvard Journal of Law & Public Policy. Before joining the Institute as a collegiate fellow in 2020, he spent a summer at the Permanent Observer Mission of the Holy See to the United Nations. John was born, raised, and currently resides in Astoria, where he sits on the board of directors of St. Francis of Assisi Catholic Academy and serves as an officer of the Astoria Homeowners, Tenants & Business Civic Association and the Long Island City Lawyers Association.
The Manhattan Institute thanks the Klinsky Leadership Series for its support in the publication of this paper. Special thanks to David Schleicher, Matthew Germer, and Sean Parnell for reviewing and providing feedback on a draft of this report. Jack Santucci offered invaluable comments relating to New York City STV history and electoral systems, for which I am deeply grateful.
Photo: Michael M. Santiago/iStock