When Your Vote Doesn’t Matter, Try Switching Ballots

If your party is going to lose, you can at least have a say in how it loses.

A black-and-white photo of a voting booth in what looks like an ornate theater
David Dee Delgado / Getty

Pundits and voters of all stripes lament just how extreme, polarized, and ideological American politics has become. But such grievances rarely come with advice for how ordinary people can address this problem, other than by voting for their preferred political party’s candidates in general elections. Even that advice isn’t very helpful: Voters in many parts of the country do not have the chance to participate in close electoral contests.

Yet Democrats in Alabama and Republicans in New York, say, still have the power to secure better representation in Congress and strike a blow against political polarization. In places where electoral competition is lacking, primary elections by and large decide political outcomes. Voters in those places are accustomed to participating in their own party’s primaries. But often the opposite party’s primary is more competitive and more consequential. So why not strategically vote in the other party’s primary?

To be clear, we’re not talking about trying to throw the opposite party’s primary to the least electable candidate, which can easily backfire. Our proposal applies to races that either Democrats or Republicans know they will likely lose. In these circumstances, voters should try to influence how they will lose. Their ultimate goal should be to pull the other side toward their preferred ideological position.

Consider the recent Senate primary in Ohio. Democrats had an entirely uncompetitive race, with Representative Tim Ryan winning nearly 70 percent of the vote, handily defeating Morgan Harper, whose previous claim to fame was losing a Democratic congressional primary in the Columbus area in 2020. On the Republican side, however, polls showed a very close race with a crowded field, and early indications suggested that turnout would be relatively low. Ryan faces a tough general-election race; no Democrat not named Sherrod Brown has won a federal statewide election in Ohio in nearly a decade.

The most moderate candidate in the Republican field, state Senator and Cleveland Guardians part-owner Matt Dolan, who surged in the polls late in the race, came in third, trailing the Hillbilly Elegy author J. D. Vance, whom former President Donald Trump endorsed, by nearly 95,000 votes. At first, this seems like quite a large number of votes, and indeed it would have represented almost 40 percent of Dolan’s eventual vote total. But if just 20 percent of the more than 500,000 Democrats who voted in their primary had cast ballots for Dolan, a more moderate voice likely would have prevailed on the right, and Democrats would have increased their odds of being represented by a senator who shared at least some of their values. Instead, 500,000 Ohio Democrats cast a vote—more than 350,000 of them for Tim Ryan—in an election that wasn’t competitive at all.

Strategic, limited party-switching has some precedent. About 8,000 nonpartisan and Democratic voters reregistered as Republicans in Nebraska’s recent primary, likely in an attempt to prevent the Trump-endorsed candidate from winning the GOP gubernatorial contest. Or consider Mississippi in 2014. Incumbent GOP Senator Thad Cochran found himself forced into a runoff against the much more conservative Chris McDaniel. Cochran survived in part because he persuaded enough Black Democratic voters to turn out and support him. Conservatives cried foul, but what was the sin? Cochran had a reputation (at least for a Republican) for reaching out to Black voters, and whoever emerged from that primary would almost surely win the general election. By participating in the election that was very likely to determine the next senator from Mississippi, these voters helped ensure that their senator was the “least bad” option. Our reaction? More, please.

This calculus could just as easily apply to Republican voters in blue states such as Maryland, Massachusetts, and New York, or in liberal cities or congressional districts. The Republican presidential primary was effectively uncontested in 2020. It would have made more sense for Republican voters to try to prevent a progressive candidate like Bernie Sanders or Elizabeth Warren from becoming the Democratic standard-bearer than to validate Trump’s inevitable primary triumph. In Washington, D.C., Republican voters regularly participate in the city’s Democratic primary because they know that a Republican candidate is extremely unlikely to win a general election; the primary is where the election is decided.

Of course, this strategy has drawbacks. First, it doesn’t work in every state, because not all states have open primaries or Election Day registration (which allows voters to reregister with the opposing party for the general election with ease). Regardless, the hassles of reregistering as a Republican or Democrat might be too high to expect too many voters to do it.

Second, there is a risk that with moderate Democrats opting to vote in a Republican primary, the results of the Democratic primary might skew toward a more extreme nominee (and vice versa on the Republican side). But our advice is for voters to selectively intervene only when their own party’s primary is all but a foregone conclusion and when a member of the opposite party is favored to win in the general election. Voters might not know to do this on their own, especially when they like their own party as much as or more than they dislike the opposing party. So it is incumbent on political elites, civic leaders, and local and state-based journalists to offer guidance when the conditions are right.

The final drawback is perhaps the most serious, so let us reiterate: We are not encouraging voters to participate in the other party’s primary elections in an attempt to nominate someone unelectable. That practice, known in elections circles as “ratfucking,” has resulted in some notable successes. The most prominent was in the 2012 Senate election in Missouri, where then–Democratic Senator Claire McCaskill spent $1.7 million on the Republican primary in the hopes that the GOP nomination would go to Representative Todd Akin. Akin won the primary, and his campaign imploded shortly afterward, thanks to his infamous remark about “legitimate rape.”

But ratfucking is a risky endeavor. What if Akin’s campaign had not imploded? Missouri is Republican enough that a win in the general election would not have been out of the question, leaving the state with a far more radical senator than the other two potential Republican nominees. In the Republican gubernatorial primary in Pennsylvania just last week, Democrats aired TV advertisements as part of an intentional strategy to boost the chance that Doug Mastriano, a state senator who had sought to overturn the 2020 election results, would end up the Republican nominee. He did—and he might well win the general election. At the presidential level, many progressive commentators suggested in the early days of the 2016 Republican primary that Trump was the best GOP candidate—in the sense of the “easiest for Hillary Clinton to beat.” Oops. Going back further, Herbert Hoover worked behind the scenes at the 1932 Democratic convention to steer the nomination to Franklin D. Roosevelt. Oops again.

Of course, some voters don’t want less polarization; recent research suggests that the penalty for extremism in elections has declined over time. But most Republicans still prefer centrist Democrats to hard-core progressives, and most Democrats still prefer centrist Republicans to hard-core conservatives. If more voters saw that primary party-switching could work in their interest—under certain conditions—the result in the aggregate would be to help restore a more centrist leadership in Washington, and perhaps even make government functional again.