What Could We Possibly Learn From a City Election 100 Years Ago?

Rose City Reform unpacks Portland’s 2024 election with Jack Santucci, who wrote a book about the history of Portland's new voting method.

Picture an election packed with candidates vying to reshape City Hall, their hopes lifted by a new voting system promising lower thresholds and multiple seats.

Yet, as elimination rounds progress, many find that with more competition on the ballot, even a lower bar can be surprisingly hard to reach.

If this sounds like Portland, think again—it’s Cincinnati in the 1920s, using the same ranked choice voting system Portland just revived.

Few people have studied Portland’s new voting method more extensively than political scientist Jack Santucci. His 2022 book More Parties or No Parties: The Politics of Electoral Reform in America, offers a deep dive into the history of Single Transferable Vote (STV) in the United States. Two years ago, I spoke with him about the significance of Portland reviving STV nearly a century after its use in early 20th-century American cities.

With Portland’s first STV election now in the books, I decided to check back in with Santucci to compare the results to historical trends. The parallels, he argues, are striking.

“People sometimes say, ‘What could you possibly learn about these systems from what happened almost 100 years ago?’ Well, the patterns are very similar,” says Santucci, a lecturer at the George Washington University and New York University in Washington, D.C. He highlights familiar dynamics: reforms drawing an unusually large candidate pool in the first election under the new system, few winners reaching the election threshold in the first round, and some winners falling short of the threshold entirely. What’s more, Portland’s results—where 11 of the 12 winners maintained their initial leads through multiple elimination rounds—closely mirror historical patterns.

Ironically, despite his expertise in STV, Santucci is skeptical of its long-term viability, pointing out that the system was repealed in nearly every U.S. city that adopted it. That, however, is a topic for another time. This conversation focuses on the parallels between Portland’s election and the STV elections of the past.

Here’s what he had to say.

This interview has been edited for clarity and brevity.

---

Based on the latest election results, it looks like no council winner cleared the threshold (25%+1) in the first round. While this could change in the final tally, how does it compare to historical trends?

Based on granular results from three or four U.S. cities that used STV for decades, it’s common for very few candidates to reach a quota [or threshold] in the first round. For example, I looked at all election data in Cincinnati, which had nine seats to fill from a citywide district. As you keep going through the years, you see a recurring pattern: one or two people achieve a quota in the first round—very rarely three. And the system always goes to elimination rounds.

(Useful context: Candidates don’t need to clear the threshold to win, but if they do, they win a seat)

Why don’t popular candidates get higher shares of first-choice votes?

With STV, there are more candidates to choose from, and the statistically likely outcome is not a candidate winning with a quota of first-choice votes. The likely outcome is winning on some number of elimination rounds. But the first round is a good indicator of support. It’s common to see, as you did in Portland, those leading in first choices eventually winning seats. I’m currently researching the prevalence of “come-from-behind” winners who overtake leading candidates through elimination rounds. Based on data from Cincinnati and New York City, we see that happen roughly 10% of the time.

We saw one winner fall short of the threshold. Was that common in jurisdictions that used STV?

Yes. What that tells us is that voters aren’t marking all their choices. As a general matter, as the number of candidates in a district rises, the ballot exhaustion rate is also going to rise. That means there are fewer ballots to go around in later rounds. Then, occasionally, we get a winner who doesn’t hit the quota.

If we look at historic elections in New York City, we see this happening a lot. New York had a fixed quota: 75,000 votes. But because voters didn’t rank choices to the extent that reformers expected, 81% of winners had less than a quota in the final round. The median final-round total for a winner was 63,736 votes.

(Useful context: An exhausted ballot is a ballot that can’t transfer to another candidate either because the voter has either only marked one choice or their lower preferences have already been eliminated or elected)

According to an analysis by FairVote, Portland’s exhaustion rates ranged between 11% to 16%. Is that consistent with historical trends?

It’s slightly higher than the 2% to 12% exhaustion rates I observed in my data for three cities’ initial STV elections.

How should we interpret voters’ decision not to rank all their choices?

We shouldn’t expect the default voter behavior to be ranking all choices. In political science, we talk about “undervoting” when a voter casts fewer votes than is permitted for any given race or office. If an electoral system permits five votes in a five-seat district, a ballot with only three votes is an undervote. But when voters are ranking ballots, the definition of undervoting is less clear. Does it mean not ranking all choices? If so, undervoting might look frighteningly widespread at first glance, but it’s really the norm because most voters don’t rank all their choices.

If we’re trying to assess the extent of voter confusion about the new system, a better metric, in my opinion, is to look at overvotes, where a voter makes a mistake and gives two or more candidates the same ranking. That’s a more serious problem because it’s an unambiguous sign of confusion, whereas ranking few choices might not be.

In two of our districts, voters had as many as 30 candidates to choose from. Based on historical data, should we expect those numbers to come down in the next election?

You probably won’t see as many candidates as you did in this first election. But it also depends on whether you get two-sided politics. If your city politics polarize and two well-defined sides emerge, that tends to deter candidate entry. The two sides send cues to voters saying, “Hey, this is a really tight competition, so bet on one of us.” If that’s not the dynamic in Portland and you get more of an uncommitted middle, then the size of your candidate pool might shrink less rapidly than we saw in some of the cases that I studied.

It also depends on whatever other exogenous events happen in politics. When I look at historical data on candidate entry, there’s always that first election with a lot of candidates—and then candidate entry starts to fall. But as you move through history, it picks up at intervals. I see spikes during the Great Depression, at the end of World War II, and during the Civil Rights Movement.

Some critics have said STV will make it harder to unseat incumbents because they don’t need to hold on to a majority of the vote. Is that fair?

STV is no more of an incumbent protection system than any other method in the United States. If you end up with two-sided politics, then you might see more incumbents holding on to their seats and that third swing seat in each district changing with the tide of public opinion. Is that incumbent protection? Maybe. I don’t think it’s fair to say STV leads to a permanent incumbent advantage.

As I digitized old STV returns, I noticed a trend. Somebody might run for office early in their career and not win. Then they’d come back the next time around and do a bit better, win a seat, have a really good term on the city council, hit peak popularity, and just keep racking up victories for the next two or three elections. Eventually, though, they would fade into the background. Their popularity declines over time as new faces come into politics.

In your book, you write about how the adoption of STV sparked coalitions competing for government control. You also highlight how some of these coalitions struggled to stay intact. What’s your definition of a coalition?

In STV systems, it’s useful to think of coalitions as pre-election organizations whose main purpose is to nominate a group of candidates and then cause them to win. Organizations that shaped the candidate pipeline from the get-go were historically most effective at gaining control of government. But the political science archetype of a coalition is a group of political interests that set aside their disagreements, or at least manage them internally, to gain control of government. The presence of some bargain is a defining feature.

In STV governments, successful coalitions weren’t always joined by ideology. In New York City, in 1938, the American Labor Party and the Republican Party were in coalition. Were they ideologically similar? No. But they were in coalition, controlling the mayor’s seat, housing policy, and the Board of Estimate [a powerful government body that’s since been retired]. That’s a pretty explicit bargain.

It will be fascinating to see what type of coalitions emerge in Portland, and how your neighborhoods react to those coalition deals.

U.S. cities that adopted STV and subsequent repeal dates. Note that Cambridge, MA, still uses the system over 80 years later. Source: Jack Santucci

---

What Shaped the 2024 Election?

Rose City Reform is hard at work exploring the key trends, historic firsts, and unique dynamics that defined this year’s election. From what united the winners to what set them apart, stay tuned for our in-depth report in December!