Working to find actionable solutions to the nation's key challenges.

QUESTION: On Wednesday, October 10 a joint report from the Bipartisan Policy Center and Center for the Study of the American Electorate found that the national primary turnout fell to the lowest level since presidential primaries proliferated in 1972.

Will voters turnout for the general election and is this indicative of increasing polarization in our politics?

View the full forum here.

By Edward B. Foley

The Bipartisan Policy Center has produced a report that invites questions about how we conduct primaries and why, in the context of a well-designed electoral system suited for the U.S. in this second decade of the twenty-first century. I’m not a statistician, and thus I will take as given the accuracy of its overarching statistical conclusion: from a national perspective, voter turnout was essentially at a “record low” during this year’s primary process. The question I wish to address is what electoral policy implications flow from this record low turnout.

The commentary accompanying the report draws its own wide-ranging policy conclusions from the data it presents, on matters as diverse as voter identification, voter registration, civics education, and the character of public culture. While I generally share the view that it is important to examine all aspects of a society’s electoral process in relationship to each other—that’s the “ecological approach” that my Moritz colleagues and I employed in our From Registration to Recounts—I’m not prepared to embrace today the report’s specific conclusions or approaches on the many issues it addresses. They certainly are items deserving of further attention, including in relationship to each other, and thus report is valuable in encouraging us to think further (and more rigorously) in this holistic way.

Still, for the purpose of this short “blog post” on the report, I want to focus on the role of primary elections as part of the overall electoral process.

Florida held its Republican presidential primary this year on January 31, the fourth earliest date in the process, when the race for the nomination was still very competitive. Yet, according to the report, only 12.46% of the eligible electorate voted in that primary, down from 15.42% in 2008. In New Hampshire, the nation’s first primary, which supposedly has that status because citizens there take especially seriously their civic duty to evaluate potential candidates for leadership of the free world, only 24.46% of the eligible electorate participated in that state’s Republican primary this year. On Super Tuesday, March 6, seven states held primaries, with the following percentage of voting-age population cast ballots in the Republican primary in each state:

  • Georgia (2.05%)
  • Massachusetts (2.89%)
  • Ohio (6.27%)
  • Oklahoma (4.12%)
  • Tennessee (3.36%)
  • Vermont (8.28%)
  • Virginia (16.66%)

Of course, these numbers reflect the fact that only part of the electorate is Republican, and on the Democratic side President Obama had no challenger for his party’s nomination.

But even if one looks back at 2008, a year of intense competition for the presidential nomination in both major parties, total turnout of the potential electorate for the primaries held on Super Tuesday (February 5 that year) were strikingly low:

  • Alabama (31.28%)
  • Arizona (24.45%)
  • Arkansas (26.16%)
  • California (36.40%)
  • Connecticut (20.30%)
  • Georgia (31.48%)
  • Illinois (33.69%)
  • Massachusetts (38.26%)
  • Missouri (32.35%)
  • New Jersey (29.27%)
  • Oklahoma (28.37%)
  • Tennessee (25.76%)
  • West Virginia (33.23%)

In none of these states, is the total percentage above 40%, and in many it’s below 30%. And this at a time when the nation was at the peak moment of its comparative evaluation of Barack Obama and Hillary Clinton, as well as John McCain in contrast to Mitt Romney and Mike Huckabee. Are these low percentages a cause for concern?

From one perspective, they are none of the general public’s business. If you take the view, as the U.S. Supreme Court sometimes seems to do, that a primary election is solely an internal party matter, then it is irrelevant who can and does show up for a particular party’s primary in any given year. The party could decide to choose its presidential nominee entirely by a “smoke-filled room” of party bosses, and offer that candidate to the general electorate in the fall. The only thing that would matter from a public perspective concerning the health of the democratic process would be turnout rates in that general election itself.

But that view of the role of primaries, in my judgment, is an unduly narrow one—especially in the context of presidential elections. It matters to the public as a whole whether the Democrats choose Barack Obama or Hillary Clinton, and it likewise matters whether the Republicans pick John McCain, Mitt Romney, or Mike Huckabee. Some voters in November might vote for Clinton but not Obama, or vice versa, and one Republican but not another. The choices made in the primary process constrain the options in November, and the public as a whole cares about those options when it comes to something as important as choosing a president. If only a relatively small fraction of the potential electorate, roughly one-quarter to one-third at best, is making the consequential initial choice, it suggests a malfunctioning of the democratic process as a whole—certainly not an optimally designed system whereby the total electorate decides how to winnow the field from many to several to one.

There are lots of reasons to think that our nation’s presidential election process, especially with respect to its primary elections, is irrational and deserves to be replaced with something better. The unseemly and chaotic way in which states attempt to jump to the front of the line, with Iowa and New Hampshire insisting that they are entitled always to go first, is a routine reminder each quadrennial cycle. And this year, with Iowa unable to announce accurately the winner of its caucus—and the potential consequence that Rick Santorum was deprived of momentum that he should have received going into New Hampshire—shows that the whole presidential primary system should be rethought from the ground up. Even if it is unrealistic to think that any major reform will be adopted immediately, it is worth continuing to shine a light on the bizarre nature of the current system and to compare it to something that would be much more appropriate.

In my view, there ought to be a ten-week schedule of primaries, with five states in each week, starting with the smallest states in the first week and ending with the largest in the last. Each week, there should be regional representation from different parts of the country (north, south, east, west), so that there might need to be some slight deviation from strict population size to assure the appropriate degree of regional diversity each week. But a schedule like this one would work:

  1. Wyoming, Vermont, Mississippi, Nebraska, West Virginia
  2. Alaska, Delaware, Kentucky, Kansas, New Mexico
  3. North Dakota, Rhode Island, Louisiana, Utah, Nevada
  4. South Dakota, New Hampshire, Minnesota, South Carolina, Arkansas
  5. Montana, Maine, Wisconsin, Alabama, Iowa
  6. Hawaii, Tennessee, Missouri, Maryland, Connecticut
  7. Idaho, Virginia, Massachusetts, Indiana, Oklahoma
  8. Oregon, Michigan, Pennsylvania, North Carolina, Colorado
  9. Washington, New Jersey, Ohio, Georgia, Arizona
  10. California, Texas, New York, Florida, Illinois

Now, if Vermont were willing to swap places with New Hampshire, and Nebraska with Iowa, I wouldn’t object.


The rules for voter participation in each state’s primary should be such that a voter can make a choice among candidates of different parties. To return to 2008 as an example, a voter in February should be able to consider among Obama, Clinton, McCain, Romney, and Huckabee as alternatives—regardless of the party affiliation of the candidates or the voters. Whether that criterion calls for a version of “top two” primary system in Washington and California (a system that the report seems dubious about), or some version of ranked-choice voting (comparable to the “instant runoff voting” in some municipalities around the country), the specific details can be worked out later. It would be necessary to be sensitive to the First Amendment concerns expressed by the U.S. Supreme Court in cases like California Democratic Party v. Jones, but it is also necessary that the Supreme Court be sensitive to the needs of the nation in designing an appropriately public winnowing process for presidential elections. When the field is narrowed to five main candidates, as it was by the beginning of February 2008, the public as a whole should have the opportunity to express its views—through the voting process—as between all five major candidates.


Adopting such a system would not guarantee that turnout rates for presidential primaries would match turnout rates for the general election in presidential years (and, of course, those rates are nothing to rave about). It would be necessary to reevaluate, and potentially tweak, the system if primary turnout rates remained dismally low. The overall goal of that tweaking process would be a two-stage electoral process—primary election followed by general election—in which the public as whole participated at appropriate rates at both stages, precisely because both are crucial to the selection of the nation’s president.

To be sure, the design of primary elections must consider offices other than the presidency. It must work for gubernatorial, U.S. Senate, and all other down-ballot races. But there is a reason why Americans pay so much attention to presidential elections in comparison to others: it is because the office is so disproportionately powerful—which is all the more reason why it is imperative to get the electoral process right for presidential elections. This new report is, alas, confirmation of the fact that, as a nation, we are so very far from getting that process right. Perhaps it can help jumpstart a conversation on how we might generate the political will to do better.

Edward B. Foley, the Chief Justice Thomas J. Moyer Professor for the Administration of Justice and the Rule of Law, is the Director of Election Law @ Moritz. He is also serves as a reporter for the American Law Institute’s Election Law Project.

Welcome to the Democracy Project Forum! This forum explores the Democracy Project’s work in government institutions, election administration, and leadership. The Democracy Project has invited some of the leading policy and political experts to respond to discussion topics, Democracy Project reports, and current events.

Any views on this forum do not necessarily represent the views of the Democracy Project, it’s Co-Chairs, or the Bipartisan Policy Center.