Bedfellows don’t get much stranger than one afternoon in Sacramento, when an ultraconservative congressman huddled with one of California’s top Democrat powerbrokers to save gerrymandering in the golden state.
It was an amber September day — after the summer scorch but before the winter fog. It was a day to be outside. But there I sat in a lobbyist’s office on L Street, overlooking the California State Capitol.
My first job out of college, I was press secretary for Rep. John Doolittle, a firebrand conservative from Northern California with an unfortunate last name for any line of work. My day job was writing red meat op-eds and newsletters to stoke his base. But Doolittle was a savvy backroom player, and the 19 GOP California House members chose him to protect their seats in the coming reapportionment.
Across the table from Doolittle was Michael Berman, the brother of legendary Rep. Howard Berman and the map-drawing genius behind the Waxman-Berman political machine. From 1970 to 2000, National Journal notes, the Waxman-Berman machine “dished out campaign cash, forged alliances, drew districts for friends (and themselves), and developed microtargeting techniques before a word for it even existed.”
But this was 1991, and a Republican governor stood in their way. Gov. Pete Wilson’s veto, if used, would allow the state Supreme Court to draw the new maps. To keep control, the Berman team needed enough GOP legislators to persuade Wilson to sign their bill or override the veto. These Republicans would be working with Democrats against their own governor to save their seats.
As bait, Berman dangled the prospect of custom districts shaped to fit their ambitions. “They’re not worried about gaining seats for Republicans if they can protect their own seats,” a Wilson aide later snarked to the San Francisco Chronicle.
In the lobby that day, GOP legislators flopped on leather chairs, thumbed magazines, told bad jokes, watched the clock, and awaited their turn in the map room. In the end, the plot fizzled, and the court redrew the maps.
Over the decades, I’ve often thought of that day in Sacramento. It was a pivotal moment in my political maturation. And, yes, cynicism. I began to suspect that elections were for the elected — not the electorate.
American cynicism about U.S. elections go back at least to the “hanging chad” chaos of Florida’s 2000 presidential election, settled by just 537 votes and a 5-4 U.S. Supreme Court decision.
Every close presidential election since has a similar tale.
Although 2020 was unprecedented — with a sitting president doggedly seeking to overturn election results and the violent events of January 6 — voter skepticism has been building for years. A 2018 poll found that two-thirds of Democrats believed that in 2016 Russia had “tampered with vote tallies in order to get Donald Trump elected President.” But, in 2020, the roles reversed. A recent NPR/Ipsos poll found that two-thirds of Republicans still believe that “voter fraud helped Joe Biden win the 2020 election.”
A Quinnipiac poll found 60% see our democracy in “danger of collapse,” and an NPR/Ipsos poll puts that fear factor at 70%.
How did we get into this mess? Can we get out?
Over the past month I interviewed political scientists, economists, reformers and activists across America, men and women who devoted years to these questions. Everywhere I looked, I found people fighting to expand ballot access, and others intent on reducing fraud. But I also spoke with experts who concluded that voter access is not at serious risk — and that voter fraud is largely a myth.
Sifting through the noise, I did stumble on one promising approach that some experts — including one Nobel laureate — think could measurably improve governance and reduce skepticism regarding our elections. And it could be done without the unrealistic and radical surgery that many proposed reforms require.
What if the primary problem with American elections is primaries? People are skeptical about general elections, because they feel their vote really doesn’t count. That’s the argument made by Nick Troiano, executive director of Unite America. “Ten percent of the voters elect 83% of Congress,” Troiano told me.
It’s simple math. Over 80% of congressional districts are safe for Democrats or Republicans, Troiano notes. In a safe seat, the dominant party nominee nearly always wins.
So the primary is the only election that matters.
But in 2020, a typical year, Troiano says only 10% of registered voters cast ballots in the partisan primaries that decided 83% of the Congressional seats.
“Primaries disenfranchise voters, they distort representation, and they fuel political division,” Troiano said.
What is true in safe congressional districts is also true statewide in states where one party dominates, notes Cal Jillson, a political science professor at Southern Methodist University in Dallas.
What comes next may feel like math, but bear with me.
In the GOP primary in Texas this month, Jillson notes, just 12% of registered voters voted in that GOP primary. In a hard-fought race for attorney general neither candidate won a majority, so Texas Republicans will vote again in a May runoff.
Jillson predicts that even fewer will vote then.
Because Texas is a safe GOP state, Jillson predicts that the Attorney General will be chosen by roughly 10% of the total electorate. Almost exactly the same ratio that chose 83% of Congress in 2020. And, as Jillson notes, low turnout typically means more extreme voters, which affects how candidates run and how winners govern.
Now consider Utah. In the 2020 GOP primary, Spencer Cox won 190,565 votes for governor out of 1.6 million registered voters. A heavily Republican state with very high turnout, 33% of registered voters voted in Utah’s GOP primary.
Cox won with 36%, and Utah does not use a runoff.
Add it all together, and just 12% of registered Utah voters voted for Cox in the decisive primary. In the perfunctory general election that followed, the losing Democratic candidate got nearly 443,000 votes — more than double Cox’s tally in the election that mattered: the primary.
“We are the world’s leading democracy, but our general elections are mostly meaningless,” says Katherine Gehl, a retired Wisconsin cheese entrepreneur, who like so many voters felt increasingly marginalized at the center.
She calls herself “politically homeless.”
Of course, some argue that not everyone has to vote for an election to be representative. And some defend the system by suggesting that many voters might choose not to participate in politically homogeneous states because they are OK with the likely outcome.
But the penny dropped for Gehl when she realized that American politics is an industry. With that insight, she teamed up with Harvard Business Professor Michael Porter to write “The Politics Industry: How Political Innovation Can Break Partisan Gridlock and Save Our Democracy.”
She also is the founder and chairman of the Institute for Political Innovation.
Gehl and Porter see American politics as a “protected duopoly.” Two dominant parties divide the customers — as my former boss Doolittle and his strange Democrat bedfellow Berman did that Sunday in Sacramento — to prevent competitors from entering the field. They look like competitors but are really collaborators.
The key to the duopoly is safe states and safe districts that choose candidates in sparsely attended partisan primaries. With such a framework, no matter how poorly the duopolists perform they still keep their customers.
Consider a well-known paradox. Congressional approval ratings have long hovered in the low teens or low 20s. But reelection rates are always well over 90%.
The paradox of low approval/high reelection is often blamed on gerrymandered districts, which does play a role, but it’s only part of the problem, and there are other ways around it.
“I only talk about things that are both powerful and achievable,” Gehl says. “Proportional representation, term limits and gerrymandering changes aren’t achievable. They all require amendments, or a different Supreme Court, or an act of Congress. Why do we spend our time on them?”
Gehl calls her proposed model “Final Five” voting, but her group actually began by promoting just “Final Four.” Alaska, their first success, adopted that model.
This year Alaska will use a Final Four nonpartisan primary. All candidates will appear on the same ballot, with party affiliations listed. Voters will choose one, and the four leading candidates will move on to the general election. There, voters will rank up to four choices. The lowest will be dropped, with votes reallocated upward till one gets a majority.
Gehl’s Final Five proposal is a close cousin within a larger reform family that in the U.S. is widely known as “ranked choice voting.” RCV is not really new. For over 100 years, Australia has asked voters to rank their preferences. In counting ballots, the weakest candidate is removed and ranked votes automatically shift upward until one candidate has a majority.
Similar models are now used in over 20 American cities, including New York and San Francisco. Variants are now used in Wales and Scotland and the U.S. states of Maine and Alaska, and ballot propositions to adopt a form of RCV are pending in Missouri and Nevada.
The term “ranked choice voting” is contested, as are the details as to which variants really work.
Eric Maskin is a Harvard economist and a Nobel Prize winner who sees ranked choice voting as a safety valve in general elections, a tool to keep major parties honest by allowing independents to run without risk of playing the spoiler. Maskin still laments the lack of an independent candidate in 2016 — when both major party candidates were relatively unpopular.
Ranked choice improves primaries, Maskin argues. In 2016, he wrote in The New York Times that if Republicans had used ranked choice in 2016 state primaries, results may have been different. “In the first 17 primaries Trump won, he got less than a majority,” Maskin told me. “He was in the 30s. Between 60 and 70 percent supported mainstream candidates.”
In 2020 Maskin supported a ballot proposition to create a version of ranked choice in Massachusetts. He blames COVID-19 for the proposition’s defeat, saying that the pandemic undercut voter education efforts.
“Roughly 70% of those who had heard of ranked choice supported it,” Maskin told me. “But those who knew little about it were opposed. And their numbers didn’t budge. We just could not get through to them. There wasn’t sufficient opportunity.”
Now, some doubt that the Massachusetts proposal would have done enough. Charlie Baker, the state’s popular moderate Republican governor, is not seeking reelection — many suspect because he could not survive a Trumpist GOP primary. And yet, because the 2020 ballot proposal would only advance one candidate per party to the general election, it may not have changed Baker’s dilemma.
Katherine Gehl doesn’t call her model “ranked choice voting.” Ranked ballots are a piece of the puzzle, she says, but they can be used in different ways. She chafes at loose terminology that dumps all such innovations — such as the 2020 Massachusetts ballot proposal — into a single basket.
Gehl says her preferred “Final Five” model is not about changing who wins, it is about changing how winners govern by making them actually represent their whole electorate, not the 10% who vote in the party primary.
With new customers, Gehl believes, the same legislators will think and act differently. Their incentives shift.
The expressive power of ranked choice could also improve voter IQ and temperament, she says. We often hear that politics is the art of compromise. Yet American voters are never given the chance to try it. Their candidate wins or loses. They may feel entitled if they win or alienated if they lose. But they never get to compromise — and perhaps it is no surprise that they often object when their representatives do.
In contrast, Gehl notes, the Final Five model allows voters to express deep concerns with a first choice they know will likely lose. No harm will be done, but a message will be sent. Their lower ranked choices then help them “learn by doing” that compromise is not surrender.
Gehl also argues that even losing ranked votes could allow submerged issues to more readily emerge, expressed by new parties — or movements and alternate candidates within existing parties. A cause or candidate who loses but exceeds expectations may help set the agenda, as others move in to serve those customers. Losing does not mean all is lost.
Gehl cites Ross Perot’s 1992 presidential run, which garnered 19%. President Bill Clinton, who won that election, subsequently bargained with a Republican Congress to tackle Perot’s signature issue, deficit reduction.
When Perot died in 2019, Clinton insider Paul Begala wrote in tribute: “I am not sure we would have ever balanced the budget without the pressure Perot and his voters brought to the issue.”
Encouraging future Ross Perots, Gehl argues, could draw out vital signals from voters. “Final Five provides almost real time polling,” Gehl says, “and allows expressions of issue concerns and salience. Even when the same person wins, we get a signal where the public is going.”
In fact, if Perot’s movement had fallen on the more fertile ground of a Final Five model and persisted, Gehl thinks the major parties might have detected and given a voice to Trump voters’ concerns prior to 2016.
Next up for the Final Five agenda are ballot initiatives this year in Nevada and Missouri, where grassroots activists are collecting signatures and clearing judicial challenges. In New York City, a different ranked choice pilot has already shown some interesting results. Eric Adams beat out a field of 13 to win the Democratic primary. Voters ranked up to five candidates, losing votes were reallocated upward, and Adams beat the second place finisher by 1% of the total. The general election was a formality.
Turnout for New York’s primary, however, was 29% higher than 2013. An exit poll of 1,662 voters showed large majorities understood ranked choice and found the ballot to be simple. These exit polls closely mirror similar findings by San Francisco State University in 2005, after San Francisco ran its first RCV election.
These innovations face plenty of skeptics. Doubters include party activists and, of course, office holders who benefit from secure states or safe districts.
In Nevada, a seasoned Democratic operative filed suit to prevent the Final Five initiative from reaching the ballot. Nevada is comfortably blue, and some Democrats prefer not to mess with a good thing.
Many political scientists also have doubts, and I spoke to several in researching this article.
“In times of crisis, the elite political class starts looking to structural changes to get us out of this mess,” Charles Stewart, a political science professor at the Massachusetts Institute of Technology, told me in an email. He added that, “If we are unable to get ourselves out of the current mess within the current structure of the electoral system we have, then RCV isn’t going to help us.”
One frequently quoted critic of ranked choice is Jason McDaniel, a professor at San Francisco State University. McDaniel’s primary concern is with nonpartisan city and local elections, where voters are set adrift with unfamiliar candidates and no party identifiers.
McDaniel objects to changes that “take out the party label and thus weaken the signals between voters and politicians, all based on a myth that voters are voting for the person, not the party.” But McDaniel doesn’t in principle object to a Final Five model, as long as it does not strip party identifiers.
McDaniel also worries that reforms are oversold. He pointed to the 2010 California reform — similar to that now used in Washington State — a simple nonpartisan primary without ranked choice. It accomplished none of its own stated objectives, he noted.
But the California/Washington State model is not “ranked choice” at all. It’s a normal one vote primary that advances only two candidates to the general election: a runoff, with none of the advantages of even rank choice voting. And it certainly is nothing like Katherine Gehl’s Final Five model.
McDaniel was one of two political scientists who actually said that those who feel politically homeless should just join the Democratic Party. The other was Charles Stewart at MIT. “Non-authoritarian Republicans need to be willing to cross the aisle. That can all be done now,” Stewart said via email.
Other experts said they worry that ongoing instability in voter confidence makes this a poor time to experiment.
But that objection brought Georgia to my mind.
The election crossfire in Georgia has been intense at least since Stacey Abrams lost a gubernatorial race in 2018. She subsequently argued for years that voter suppression cost her that election.
Then, after the 2020 election, defeated President Donald Trump spent weeks falsely alleging voting irregularities in Georgia, discouraging his own base from voting in the January runoff, likely costing the GOP control of the U.S. Senate.
Georgia Republicans later passed legislation attacking voter fraud concerns. That legislation created a backlash, in turn, alleging that the new voting rules were racially motivated. Major League Baseball stepped into the fray, moving its All-Star game from Atlanta to Denver.
Not a time to experiment? Chaos in states like Georgia and polls that show 60%-70% of Americans think our democracy is in crisis suggest that this might, in fact, be the very time for a calculated risk that could offer enormous upside for legitimacy and governance.