Introduction
Vote percentages of the Utah 2020 Gubernatorial Primary Election by county region. Regions were defined by the Rural Planning Group — a program of the Housing and Community Development Division, part of the Utah Department of Workforce Services.
The Republican Primary for the Utah Gubernatorial election raised questions about membership in political parties. Only registered Republican voters were allowed to participate in the Republican Primary— the election that would likely determine the next governor. The only way for voters not registered as Republicans to participate was to switch their party registration.
The ideologies of the four candidates who sought the Republican nomination complicated the situation. Two candidates leaned moderate, while the other two leaned right. Normally, candidates who have different ideological inclinations than the base of the party would not fare well in a primary election reserved for voters who identify with the party. Consequently, one of the candidates, former Governor Jon Huntsman, sought to enlarge the primary base by urging voters who are Democrats or unaffiliated to change their affiliation to the Republican party. An informational video created by his campaign described how to register as Republican to vote in the primary held on June 30th.
The encouragement of cross-party voting and new affiliations raised interesting questions about the composition of the primary electorate. Who switches affiliations? And once they switch their affiliations, do they vote? What impact, if any, did these types of voters have on the outcome of the primary?
Political Science and Cross-Party Voting
Cross-party voting has a long history. As states adopted voting rules, they considered the impact of having closed versus open primaries as a means by which the parties would select the candidates to place on the ballot. Because most policy makers are also politicians, they considered the strategic impact that different rules would have on the composition of the electorate.
A significant assumption they often made was that voters would behave “strategically” if they could. This means that voters would look at the candidates put up by the parties, consider what was best from a set of policy perspectives, and then choose accordingly.
Such strategic considerations would also involve assessing the candidates of the other party and considering what would be the best match-up for the general election. For example, voters can “raid” another party’s primary by voting for a weak candidate. Rush Limbaugh famously encouraged Republican voters to vote for Hilary Clinton in 2008 because he said the Republican nominee would fare better against her. Voters can also “hedge” their vote. Here voters from one party select the more moderate candidate from the other party to ensure that there will be a good candidate in the general election.
How do these strategic considerations apply to the case of the recent Republican Gubernatorial primary? There are many ways to think about why cross-party voting was encouraged, and what certain actors hoped it would achieve. For instance, it is expected that the winner of the Primary (Lt. Gov. Spencer Cox went on to win the primary) will go on to win the General Election for Governor. Consequently, some individuals did encourage non-GOP voters to register in order to vote against the candidates they saw as more radical in the primary. If a Republican is going to win the General Election, the argument goes, better that it be a moderate Republican such as a Cox or a Huntsman. This is a clear example of hedging.
But there was also “hedging” with a twist that occurred. Here, one of the Republican candidates, Huntsman, actually encouraged individuals from another party or who are not affiliated with the party to participate in the Republican primary. Possibly responding to polling that showed him as the most favorable candidate among unaffiliated voters, his campaign sought to get unaffiliated voters and Democrats to register as Republicans.
“So does the encouragement of “hedging” work? The answer is “yes” and “no.” ”
So does the encouragement of “hedging” work? The answer is “yes” and “no.” If party identification is, as most political scientists argue, a deep psychological attachment to a particular party, then persuading people to go against that attachment is hard to do. While there is a large number of unaffiliated and Democratic voters on the registration rolls, their psychological party identifications inhibit their viability as an electoral force in any one direction. The data seem to indicate that voters registered as unaffiliated still have some psychological attachment to a party, making them resistant to influence from opposing parties.
Other research affirms that cross-party voting rarely affects elections in substantial ways. In addition to the psychological attachments voters have, cross-party voting also assumes that voters pay enough attention to elections to vote strategically. However, most voters “...lack the political sophistication to vote strategically”. This is not to say that cross-party voting does not occur -- hedging certainly occurs, as does raiding – but they do not seem to occur in quantities large enough to affect the outcome of the election, unless the election is very close.
Practical Scenarios of Cross-Party Voting in the Gubernatorial Primary
So, what happens when we apply this form of reasoning to the recent election? Given known parameters of the Utah electorate — such as size and turnout — we can explain, or reverse engineer the outcome of the recent election to show how other forces at play in this election, as with other elections, overwhelmed the power that switcher Republicans might have had.
It is important to put switcher Republicans into context. Though there were indeed a large number of switchers in the recent Republican Primary, focusing too closely on raw numbers without understanding their relative size gives a false impression of their importance. According to Utah voter files, the population of registered Republican voters in the state of Utah was nearly 727,000 voters in June of this year. Of those, about 69,000 of them switched from other parties or from being unaffiliated since January, making up less than 10% of the eligible electorate in the primary election. In this case as the literature contends, this group did not swing the election for various reasons.
Where did the switchers come from before changing their registration to Republican? Analysis of voter files shows that 75% of them were unaffiliated voters, 20% of them were Democrats, and 5% were part of disparate other parties before switching.
Considering what we know about the switcher Republicans in the recent primary election, there are some probable explanations as to why this bloc did not sway the election as some supposed it would. Some explanations include previous party affiliation and sub-group turnout.
Switcher Republicans came from many different ideological camps before switching. A sizable minority were previously Democrats, but the vast majority were unaffiliated. Research shows that the party identification or ideology of unaffiliated voters is evenly split between liberals, true centrists, and conservatives, making it hard to predict their vote choice as a single group. In our polling, switchers who were previously Democrats expressed Huntsman as their clear favorite (83%), which might have been a boon to Huntsman if more Democrats switched to Republicans. Huntsman was also the preferred choice among a majority of previously independent voters (54%), but not by as large of a margin as 32% of them said they would vote for Cox. The split in preferences among switcher Republicans, combined with their relatively small size, made them a far less formidable force than many expected they would be.
In addition to party affiliation and preference among switchers, we must also consider election turnout. In order to sway the election, switcher Republicans would also have needed to turn out at significantly higher rates than previously registered Republicans. By looking at the data, it is clear that this did not occur. Over 520,000 votes were cast in this primary – an historic primary turnout of around 72%. In a simulation with an electorate of this size, where the votes of different groups are broken out the same way as our polling, and where turnout is 72% for all groups, projected results are consistent with actual election results, with Cox winning by a few thousand votes. Therefore, in order for switcher Republicans to sway the election for Huntsman, their turnout would need to be significantly higher than that of previously registered Republicans. Given how high turnout was across the board this election (due to high enthusiasm, high campaign spending, mail-in voting, and the closeness of the race), switcher Republicans would have had to turn out at rates nearing 100% to have a chance of swaying the election. The only other alternative is if the overall turnout were lower, which was not the case for this election. The combination of voter turnout, small size, and disparate candidate preferences among switcher Republican subgroups is responsible for their limited electoral power in this election.
Conclusion
Reacting to a very close election and multiple polls that showed Huntsman as the favorite among non-Republican voters, certain commentators became enthusiastic about a scenario where switcher Republicans effectively swayed the election. While ‘hedging’ and ‘raiding’ did occur here as occur in many elections, other forces were at play which, we have explained, overwhelmed the power of these voters. This election should be included in the contemporary literature of cross-party voting to show why and how it happens; and how it affects, or doesn’t affect election outcomes.