This is a guest post by Brigham Young University political scientists Christopher Karpowitz and Jeremy C. Pope.

The Republican Party recently changed its presidential nomination procedures, mainly by adjusting the calendar of caucus and primaries.  But what they didn’t do was reconsider whether to have caucus or primaries in the first place, despite the urging of some in the party.

Mitt Romney, for example, told the Boston Globe that he was “concerned that there’s an effort on the part of some to move toward caucuses or conventions,” something that is a mistake because “primaries are the place where you see whose message is connecting with the largest number of people.” Some Republicans in Virginia feel likewise: After replacing their primary system with a convention before the 2013 election, the more moderate Republican gubernatorial candidate, Bill Bolling, suspended his campaign, leading to the nomination of the more conservative Ken Cuccinelli II — who then went on to lose in the general election to Democratic candidate Terry McAuliffe.

Moderates in the Republican Party would like to move away from caucuses and toward primaries, believing that more extreme candidates are more likely to come from a caucus rather than a primary.  Our new research (gated here and ungated here) suggests that they have a point: Those who attend caucuses have a very different set of issue attitudes than those who take part in primaries.  Moreover, Americans tend to agree that caucuses are unfair and give an advantage to the wrong sort of interests.

It’s a standard piece of conventional wisdom that the voters who show up for caucuses are more ideologically extreme than those who vote in primaries.  However, the political science literature has not consistently confirmed that conclusion.  Our research does something relatively new, however: We took a large survey of Americans and determined who among them had actually participated in a primary or caucus in the 2008 presidential nomination, based on data from state voter files.  In other words, we did not rely solely on voters’ imperfect recollections of whether they voted in a primary or caucus (or not at all).

Even after accounting for many other factors, caucus attenders were more ideologically extreme than primary voters.  In terms of their willingness to take consistently conservative or liberal positions on the issues, caucus attendees look a lot more like members of Congress than they do average Republicans or Democrats.

The second part of our research sought to gauge how voters felt about primaries or caucuses in the abstract.  We posed this question to voters as part of a survey experiment.  Respondents to a 2008 survey were randomly assigned to read the basics of how either a primary or a caucus works.  We then asked them to think about the next time their party would nominate a candidate and evaluate the process they had just read about.

Those who had actually participated in the caucus system in the past were especially fond of it, but almost everyone else had concerns.  Relative to those who read about primaries, respondents who read about caucuses were less likely to say that the process would result in the “best decisions,” would be fair, and would be friendly to different points of view.  Respondents who read about caucuses were more likely to say that the process would advantage special interests.  This graph shows these results:

Many political scientists argue that reforming primaries is not likely to fundamentally change the types of candidates who are elected.  However, very few political scientists have compared primaries to their most common alternative, caucuses.  Caucuses attract not just a smaller group of voters, but a group that is the most committed and ideological.  Sen. Ted Cruz (R-Tex.) likes to say that voters “are hungry for principled, conservative fighters—because the threat to our liberties from Washington never has been greater.”  When he says that, he’s much more likely to be talking about a caucus-goer than he is a primary voter.  Because their issue preferences are different from those of primary voters, caucus electorates may be more likely to let through an extreme nominee who may be penalized in the general election — a possibility that is worth further scholarly examination.

Voters who perceive caucuses as unfair, less friendly to different points of view, and better for special interests may not be able to perfectly articulate what is wrong with caucuses, but their intuition that caucuses are not representative is supported by the data.

Is there any hope that parties may heed their views?  Our home state of Utah will be an excellent case to watch.  Utah parties select their nominees through a complicated process that can sometimes include primaries, but where many key decisions are made in party caucuses and conventions.  In 2010, a party convention helped Mike Lee defeat incumbent Sen. Robert Bennett, even before the primary was held — this despite the fact that Lee is further to the right on many issues than even the average Utah Republican and certainly more to the right than Bennett was.  But Lee fit the profile of convention delegates, who were more extreme ideologically and much more likely to be active supporters of the tea party than primary voters.

Right now a movement called Count My Vote is trying to change the Utah system from a caucus/convention system to a primary system and may get the issue on Utah’s ballot next fall, although the campaign is still in its early stages.  In our view, the key question is one of representativeness.  Will the parties choose electoral institutions preferred by the most ideologically committed participants, or will they choose institutions that attract a wider and more representative sample of voters?

[Correction: Due to an editorial mistake, the post originally said that Cuccinelli defeated Bolling at the convention.]