As the fall of House Majority Leader Eric Cantor and the rise of presidential candidate Donald Trump remind us, primary elections are becoming increasingly important contests at just the time that American voters are losing interest in them. Primary participation has fallen in many US states, often to historically low levels, with tens of millions of those who vote in general elections habitually abstaining in primaries, as Elaine Kamarck also notes in her article in this series.
These “November-only voters” miss the chance to participate in the elections that determine their general election choices. Can these voters be effectively mobilized, thus enlarging and broadening primary electorates, or are those who typically abstain from primaries a lost cause for primary elections?
Campaigns certainly seem to think that these occasional voters are hopeless, and this belief creates a vicious cycle. November voters who sat out primaries in the past look like unlikely voters in the databases that campaigns use to target their mobilization efforts. With limited budgets the consultants running primary campaigns keep their efforts focused on likely voters. They turn out their base and set aside the November-only voters. Without anyone mobilizing them, these voters once again sit on the sidelines while the consistent primary voters nominate the candidates for office.
Talking with candidates who ran statewide campaigns in California’s 2014 primary confirmed our suspicions about how this self-fulfilling prophecy works. One candidate told us that, “We targeted all Democrats and independents who were regular primary voters. The mailers were not going to people who don’t vote in primaries. Vote for vote, it is going to cost you more resources to mobilize one of those voters than it is to reach someone who is going to turn out.” Another candidate explained that, “[W]e targeted Republican voters: We sent out two direct mail pieces to Republican voters who turn out frequently in primaries.”
Turnout in California’s June 2014 primary was the lowest in state history, with only 25.2 percent of registered voters showing up at the polls. What about the November-only voters who turned out in the 2012 general election but had missed recent primaries? There are nearly 4 million of these voters in California. They were likely mostly ignored by candidates in the June 2014 primary because they were flagged by the databases as November-only. Only 9.3 percent of them turned out to vote.
Reaching November-Only Voters
How do you reverse this cycle and get November-only voters to vote in primaries? In a get-out-the-vote experiment in the 2014 primary, we worked with the non-partisan nonprofit group California Common Cause to contact a random sample of 150,000 of California’s 4 million November-only voters. We aimed to see if they could be mobilized effectively, providing a proof of concept to campaigns that this is a group they can reach and turn out. We also wanted to know whether all types of voters in this group would respond equally to the mobilization efforts, or whether there could be disparities in turnout that echo the overall disparities in who comes to the polls during primaries.
The full details of our research and what we found are in the article “Turning Out Unlikely Voters? A Field Experiment In The Top-Two Primary,” which is forthcoming in the journal Political Behavior. Here are the key findings:
- Most important and surprising, our randomized experiment showed that November-only voters are not a lost cause in a primary. In fact, they responded to the mobilization message more than voters usually do in get-out-the-vote efforts. Our targeted voters who received a single letter turned out at a rate that was 0.5 percentage points higher than other November-only voters. It represents a proportional increase in turnout of 5.4 percent compared to the baseline turnout for this type of voter. This group can deliver electoral bang for a campaign’s buck. The article by Alan Gerber and Greg Huber in this series also mentions experimental evidence in support of the argument that voters can be mobilized in primary elections.
- We also found that this effect was consistent across different types of voters. Voters were mobilized at comparable rates whether they were young or old, male or female, Democrat, Republican, independent, or members of minor parties. This demonstrates that targeted mobilization can broaden rather than narrow the electorate.
- The impact of our letters did vary by the political context of the congressional district to which we sent them. The impact of the letters was dampened in districts where there was lots of campaign spending already, indicating that large campaign spending in such districts is a substitute for our mobilization message. By contrast, the impact of our letter was magnified in districts with a closely contested primary election, suggesting that the buzz that tight contests generate can serve as a complement to our communication.
Our results lead to new questions about how to expand primary electorates that we’d like to explore in future research. A 0.5 percentage point increase in turnout from a single letter may be a nice proof of concept, but what if we went after these November-only voters the way that campaigns typically mobilize their base? Would multiple contacts through different methods bring a game-changing boost in turnout? Would all voters respond to such a get-out-the-vote campaign, or would differences emerge that could exacerbate inequalities in turnout?
All of these questions are worth investigating, but it is also worth keeping a focus on a recurrent puzzle in American politics: Why do so many of the citizens who turn out to vote in general elections abstain from primaries?
An important part of the answer to this question, our experiment suggests, is that these voters stay home because nobody asks them to turn out in a primary. Campaigns see these voters’ past turnout record and ignore them today, setting them up to be left out of future mobilization efforts. The way to turn this vicious cycle into a virtuous one is to recognize that November-only voters are not a lost cause in primaries and that mobilization now has payoff for turnout in future elections.
Our experimental findings should demonstrate to campaigns that their mobilization dollars will not be wasted on this untapped vein of voters. Spreading that message to campaign consultants and nonprofit get-out-the-vote groups could lead to new mobilization strategies that make primary electorates larger and more representative.