A Proposed Presidential Primary and Caucus Schedule

Now that the Presidential election is over it's time to fix our broken primary system. I propose a schedule of presidential primaries designed to put presidential candidates through a meaningful selection process and make voters in all states feel included in contested primary elections.

The system that has been in place is clearly at the breaking point. For years voters in states holding late primaries have felt disenfranchised, believing that the race was over before they had the opportunity to participate. In response to those concerns many states have begun to leap frog one another in an effort to increase their influence. February 5, 2008 was labeled "Super Duper Tuesday" by some with 24 states holding caucuses or primaries on that date.

This resulted in two serious problems. First, voters in states holding primaries after that date will rightly complained that their votes were going to be all but meaningless. On the Democratic side this proved to be premature, but the political facts underlying concern remain, and if not fixed, this is likely to be a perennial issue. More importantly, placing so many primaries on one day in February further increases the primacy of early fundraising, TV advertising, and staged events. It also decreases the influence of "retail politics," i.e. in-depth interactions with real voters. What is needed is a primary schedule that puts presidential candidates "through their paces" and makes voters in every state feel that their votes count.

Of course, we could hold all of the primaries on the same day. This would take care of the problem of late state disenfranchisement, but would be a grave mistake. In many years no candidate would earn more than 50% of the delegates leading to a series of brokered conventions. Candidates with less money and name recognition simply would not have the opportunity to make themselves known to voters. Our country is better served by a process that unfolds over time in which candidates are vetted in small face-to-face settings as well as debates, interviews with local and national media, the Internet, large rallies, and national addresses.

The National Association of Secretaries of State proposed a series of four (4) regional primaries with a different order in different election cycles (see nass.org). This proposal is flawed in several respects. First, although things will ultimately even out, voters in the fourth region (east, south, Midwest, or west depending on the year) will still be disenfranchised in any given election. The system also suffers from political bias. For example, years when the southern region went first might easily include (or exclude) different candidates then years when the New England states went first. A quick look at the map of results of the Democratic primary reveals that the luck of the draw would have significantly favored one candidate or another. The same would probably hold for the Republicans. Finally, with the exception of traditionally first Iowa and New Hampshire, regional primaries deprive voters of the opportunity to see candidates in smaller settings.

The schedule that I propose is founded on the following 6 principles:
(1) Make voters in every state count in all seriously contested Presidential primaries;
(2) Require candidates to engage in "retail politics" before relying on mass media;
(3) Reduce or eliminate any political or regional bias built into the schedule;
(4) Ensure that candidates face a diverse electorate at the beginning of the process and that voters from various ethnic groups are not under-represented in the early going;
(5) Make candidates alternate between liberal and conservative regions in the early primaries to avoid a distorted sense of momentum caused solely by the schedule;
(6) Provide candidates with multiple strategies for getting their messages out.

The basic logic behind the proposal is to put the smallest states first followed by increasingly larger states with a Super Tuesday at the very end – but before the race is over – electing about half of the delegates. The smaller-to-larger system is modified slightly to allow for some grouping by region and ensure that there is no major political bias between the early states and the Super Tuesday states at the end.

I used Electoral College votes as a measure of each states size, since they are part of the primary formula used by both major parties and it makes it easy to think about representation in the general election. I used each state's percentage of voters who voted for President Bush in 2004 as an index to conservatism. States with fewer that 47% of the voters backing Bush in 2007 can be considered relatively liberal, states where between 47% and 53% of the vote went to Bush can be considered moderate, and states with more that 53% supporting President Bush in 2004 can be considered conservative.

Following tradition, I suggest Iowa as the first state to hold a caucus and New Hampshire as the first state to hold a primary. I put them both on February 11. Week 2 starts on February 18 and has primaries or caucuses in Alaska, Hawaii, the District of Columbia, Delaware, and New Mexico. This diverse set of small states (plus D.C.) only account for about 3.3% of the population, but has relatively large numbers of African American voters in Washington, D.C., Asian voters in Hawaii, Hispanic voters in New Mexico, and Native American voters in Alaska. Liberals are over represented in D.C. while conservatives and libertarians dominate Alaska. A candidate might choose to concentrate on, or ignore, any of these states. Unlike the current schedule, it would be harder for the media elites to prematurely write off a candidate or prematurely declare someone the winner.

Weeks three, four, and five swing to the more conservative western states of Montana, North Dakota, South Dakota, and Wyoming followed by the more liberal New England states of Vermont, New Hampshire, Maine, Rhode Island, Connecticut, and Massachusetts and then the more conservative southern states of Arkansas, Kentucky, Missouri, and West Virginia. By this stage candidates have been judged by voters in 18 states with a conservative edge of only 1 state, yet less than 18% of the delegates have been selected.

By this point one would reasonably expect that most long-shot candidates have dropped out or been discredited. Yet unlike the 2008 schedule candidates in the second and third position could easily have strong showings somewhere to build upon. Although voters in these first states have access to the full range of candidates, voting later has advantages too. Are supporters of long-shot candidates always better off having their favorite candidates in the field? Not necessarily. If it isn't "in the cards" for candidate Smith to win more than 10% of the votes his or her supporters may be better off choosing among the two or three candidates who actually have a chance at winning the nomination. While political pundits will always be quick to declare a candidate finished or a race over, this schedule makes such predictions more perilous as candidates move from Wyoming to Vermont to Kentucky in a span of three weeks. Significantly, the relative importance of the first states is actually diminished in the proposed schedule because they are followed by more relatively small and medium sized states rather than a must-win early Super Tuesday.

Weeks 6 through 11 swing back and forth between the conservative west, moderate and liberal Great Lakes, conservative middle west, liberal mid Atlantic, conservative south, and a mix of liberal, moderate, and conservative western states. After 11 weeks voters in 38 states and D.C. will have cast ballots, yet only about half of the voters will have spoken. In Electoral College terms, those states represent 268 of 538 (49.8%) of the vote. Of course, many of the smaller states tend to be more conservative, and the Bush 2004 states net out as +7 after 11 weeks compared to +4 on Super Tuesday. However, this schedule stills ensures that the Super Tuesday states are approximately as conservative as the proceeding states. As an index to conservatism, we can multiply the percentage of Bush voters in 2004 by the number of Electoral College votes. This proposed schedule gives the early states a conservative index of 138.5 and the subsequent Super Tuesday states a conservative index of 137.3 – nearly identical scores.

I would give candidates two weeks to campaign before Super Tuesday. In the schedule I propose the primaries would end on May 12 (week 12 of voting) with the 12 states of California, Florida, Georgia, Indiana, Michigan, New York, North Carolina, Ohio, Pennsylvania, Tennessee, Texas, and Virginia participating in Super Tuesday. With about half of the delegates still up for grabs it is unlikely that Super Tuesday voters will feel their votes don't count. Of course, in many years primary candidates (particularly sitting presidents) are all but unopposed, and no schedule can guarantee that a primary race will remain viable from start to finish. Yet this schedule gives viable candidates many avenues to make themselves contenders going into Super Tuesday.

Any change to the primary system will pose significant political challenges. Choosing a President is the most important task of the American electorate. The process of electing candidates should be fair to all voters and allow us to scrutinize candidates through a rigorous and meaningful process. I believe this proposed Presidential primary and caucus schedule would achieve these goals and take a step toward returning our elections back to the voters.