Stop Letting Candidates Label Themselves

On-ballot endorsements would give voters clearer signals about who candidates really are.

Courtesy Leon Bennett/Getty Images for CBS Television Stations

When Californians vote in next month’s open gubernatorial primary, their ballot will list 61 candidates in total. Each name will appear alongside the candidates’ party preference as well as a brief occupational label of their own choosing. Progressive billionaire Tom Steyer is a “Climate Advocate,” for example, while former Los Angeles mayor Antonio Villaraigosa is a “Housing Affordability Advocate.”

Such labels may help voters a little, but they also invite manipulation. If the state wants to give low-information voters a clearer signal about what candidates actually stand for, it should stop letting them write their own ballot descriptors in the first place. A better option would be to print endorsements on the ballot itself—whether from political parties, newspapers, civic groups, nonprofits, or others. Such a reform would be useful not just in California, but also in New York City, where most elections involve ballots that tell voters very little about how those candidates would govern in practice.

Under California’s “top-two” system, all candidates appear on a single ballot, regardless of party affiliation; the top two vote-getters advance to a general election runoff. The system emerged after the U.S. Supreme Court struck down California’s previous “blanket primary,” which allowed all voters to pick candidates for each party’s nominations. The court held that this approach violated the parties’ freedom of association. In response, the state sought a way to still help voters distinguish among dozens of candidates—hence, allowing candidates to offer a three-word occupational description (with some exceptions for officeholders with longer titles).

The basic idea behind California’s system makes sense—give low-information voters an immediate signal besides candidates’ “preferred” party labels. But in practice, candidates reach for the most flattering three-word description they can plausibly defend.

The problem is not unique to California. In New York City, where closed Democratic primaries decide most elections, a ballot full of Democrats tells voters essentially nothing about how candidates would govern.

These issues are especially acute in low-information municipal races, where many candidates have little name recognition beyond their own neighborhoods and local issues seldom fit neatly into national ideological categories. A voter’s view on bike lanes doesn’t necessarily predict his views on policing, property taxes, public education, or housing policy.

When immediate and reliable information is scarce, voters rely on whatever cues are available: name recognition, ballot order, the candidates’ perceived ethnicity and sex, and so on. Such guesswork erodes democratic accountability by delinking candidates’ electoral success from their campaign platforms and promises.

By allowing candidates to self-select both party preference and occupational designation, California and similar systems invite candidates to stretch the truth. Candidates must provide some factual basis for their declaration in their application to the Secretary of State, but whether that basis is adequate is often subjective.

Unsurprisingly, these labels have resulted in court battles. In one 2018 California Superior Court case, the court noted that “each election cycle those three little words generate significant litigation,” and said that candidates had become “creative” with their ballot designations.

There is a better way. In 2013, legal scholars Chris Elmendorf and David Schleicher proposed allowing top elected officials to designate their preferred candidates on the ballot as an “executive cue.” Their idea can be broadened into a fuller system of ballot-qualified endorsing organizations. Rather than rely on candidates’ self-description, states and cities should allow qualified organizations to place endorsements directly on the ballot. Over the years, I’ve written and testified about the desirability of adding such endorsements to New York City’s ballots.

As Elmendorf argued recently, on-ballot endorsements could also help recreate a healthier form of party-like competition in places where one party dominates. This is important because, as the authors noted in their original paper, it’s not possible to register as one party for national elections and another party for state and local ones. The Republican Party’s longstanding unpopularity in places like New York and Los Angeles renders it effectively incapable of serving as a political check.

In addition to parties, endorsements can serve as information cues to guide voter behavior. As Elmendorf suggests, ballot-qualified endorsing organizations could compete for voter recognition and then place their imprimatur next to one candidate in each race. That would give candidates a path to viability without depending so heavily on interest groups for money and campaign labor, while also offering ordinary voters a clearer way to distinguish candidates on ideological and policy grounds.

A teachers’ union endorsement, for better or worse, would tell a voter something important. So would an endorsement from a pro-housing group, local newspaper, or taxpayer association. Voters may agree or disagree with the endorser, but the signal is externally validated by a third party whose reputation is on the line.

The reform could also strengthen civic participation. If ballot endorsements carry real political weight, local organizations would have a stronger incentive to engage in local democracy. A neighborhood civic association, for example, could become a meaningful power broker by earning enough community trust to make its endorsement matter. That prospect could help spark the kind of local participation that counteracts the nationalization of our politics.

Such a system might also encourage factions within the same party to reorganize under new party or organizational labels. As Manhattan Institute adjunct fellow Jack Santucci has recently shown, New York City already functions as if it has several parties: the far left, the center left, and everyone to their right. Yet the ballot does little to help voters distinguish among them. Giving voters a choice among a Democratic Socialist, a Democrat, and a “Common Sense” center-right candidate would, ironically, be simpler than forcing them to decipher candidate differences within the city’s ranked-choice Democratic primary.

In nonpartisan or all-candidate systems like California’s, this system would strengthen party institutions rather than weaken them, giving parties a formal role in helping voters identify candidates. That could make parties buy into the reform instead of opposing it.

Under my proposed system for New York City elections, parties and other organizations would decide endorsees for each office through internal mechanisms. Other endorsers should qualify by collecting voter signatures, with thresholds scaled to the office. Organizations could also be required to have existed for a minimum period, perhaps one year, to prevent fly-by-night entities from gaming the ballot.

To avoid clutter, candidates could be limited to a fixed number of printed endorsements—say, three. Endorsers and candidates could use ranked choices to make their preferred selections, allowing election administrators to assign endorsements without repeated rounds of selection. Candidates should also be allowed to reject or not rank endorsements, preventing malicious or strategic attempts to saddle them with unpopular supporters.

On-ballot endorsements would not eliminate the need for voter guides, debates, local journalism, or civic education. Nor would they solve every problem with local elections. But they would give voters a clearer signal about the people asking for power. That would be a win for local democracy, wherever it happens to be.