Re-imagining and Improving the Presidential Primary Process
As Presidential Primary season has already begun, the same complaints that arise every primary season have also started. The primary calendar begins too early, starts with the same two states every year, fails to represent a diverse country, and is front-loaded so that states voting later consistently fail to influence the process. While there are systematic obstacles, including current state law and the political influence of Iowa and New Hampshire, that shouldn’t stop us from imagining what a better system that overcomes these problems might look like.
Scholars have looked at this issue before and come up with a variety of solutions. They historically have run into several of the same problems. These generally involve creating extraordinary traveling distances between voting states, building in biases against either large or small states, and creating a system impossible to succeed in for a candidate without major donor support. The most recent idea to receive attention, a National Primary, would have all three of these issues. Candidates would only focus on dense states and areas, travel across all regions of the country would be inefficient, and candidates with little prior name recognition or donor relationships would have no way of spreading their message broadly.
To create a system that solves for the problems in the status quo while avoiding these common problems, the best idea seems to be some type of rotational system. This ensures a variety of states can exert influence on the primary process. An ideal rotational system would involve grouping states that are, at a minimum, on the same side of the country to reduce travel, while balancing influence among smaller and larger states.
In my attempt to solve for these issues, I divided the continental 48 states into groups. I excluded Alaska and Hawaii from this system because there is no method to incorporate those states into an early primary rotation that’s financially feasible for any candidates beyond those who are extremely well-funded. First, I divided the 48 states into 2 groups based on PVI (Partisan Voting Index) to measure ideological lean, creating sets of the 24 most “Blue” State and the 24 most “Red States.” By ensuring Blue and Red States have evenly spread influence across the primary calendar, we can try to solve for some of the homogeneity that comes from having two relatively similar (Moderate, Older, Whiter) states voting first in the status quo.
After dividing into these 2 groups of 24, I divided each group of 24 into groups of 12 by population size. The result is 4 groups of 12: “Large Blue”, “Small Blue”, “Large Red”, and “Small Red” states. By having relatively small and large states grouped, we can ensure that smaller and larger states are distributed across the primary calendar and attempt to eliminate biases towards either smaller states or larger states.
Finally, I created 12 groups of 4 by selecting One State from the “Large Blue”, “Small Blue”, “Large Red”, and “Small Red” states that, while perhaps not close, are at least not on opposite sides of the country from each other. This method attempts to create diverse clusters of states based on partisan lean and size while making travel relatively manageable and ensuring small-dollar campaigns still have an opportunity to compete. As an example, I created the following hypothetical groups when dividing the 48 states in this way:
Taking the 12 groups of 4, a rotation starts where each group gets a chance to go first in the primary calendar. If necessary, we can decide which group starts by lottery, or allow the groups including Iowa and New Hampshire to be #1 and #2 as a concession for removing them from First-in-the-Nation status. I propose starting the primary calendar start the first Tuesday of March to fight back against the trend of Presidential elections starting earlier and earlier, though the basic system works with any Primary start date. On the first date, the “Small Blue” and “Small Red” states from the opening group vote. This keeps one state from dominating the process while ensuring candidates without a fundraising base have an opportunity to succeed early through retail politics. Then, on the 3rd and 4th Tuesdays in March, the 2 “Large” states from the group vote. This ensures each large state gets a voice as well by getting its own voting day and attention in the first month of the primary.
By the end of March, 1 of the 12 groups has had an opportunity to vote. By this point, the field should have winnowed significantly. In April, 3 more groups vote to incorporate a variety of different regions and states into the process. I spread these out across 2 “Super Tuesdays”. On the first Tuesday of April, all 4 states in the group 6 removed from the group that already voted (so if Group 1 voted, then Group 7 votes now, etc.) get a chance to vote. While competing in 4 states simultaneously should be manageable for any candidate that has made it this far, it still allows candidates that maybe weren’t suited for the first group of states that voted another chance to have their voice heard. Then, on the 3rd Tuesday of April, the Groups removed 3 and 9 away from the first group that voted (so if Group 1 went first, then Groups 4 and 10) vote. This allows another cluster of 8 states a chance to have early influence.
At this point, the primary field should be down to 2 or 3, if not settled. I don’t think there’s a need to micromanage voting dates at this point to ensure results that satisfy the criteria we want to see in a good primary process. I would simply say all remaining odd groups should vote at any point in May, while all remaining even groups vote in June, with odd groups and even groups flipping each cycle. Alaska, Hawaii, Washington D.C, and the American territories are able to vote anytime they wish after May 1 (perhaps even on the same day to maximize their overall influence if they wish). Here is what an example calendar using this system would look like:
This system, I believe, successfully solves for the issues in the current system and the National Primary. It breaks the current 2-state stronghold on the front of the Primary calendar and ensures every state gets a spot in the front 4 at some point. Because it revolves around rough geographic clusters, it forces candidate to become familiar with a variety of different policy issues (similar to how Iowa being first has encouraged a focus on ethanol) while limiting travel. Because both Red and Blue states are being moved early, the states voting first should ideally be a little more diverse than Iowa and New Hampshire. Finally, it allows a path for retail, less donor-heavy candidates to compete in early states while ensuring larger states are also represented early. To account for realignment and changing ideologies and populations in states, because a rotation lasts 48 years, it would be possible to reset the groups at the end to account for changes before beginning the next rotation.
There are systematic issues that prevent this from being easy to accomplish. Because Iowa and New Hampshire are currently crucial states for Presidential candidates, there is little incentive for any individual potential leader to challenge the status quo. Additionally, each State has its own laws that regulate the process. To implement this rotation system, it would require the National Parties to become involved and for each individual state to adjust accordingly. For the average state, though, this system is an improvement on the status quo. Every state would get to vote in the first 4 at some point, which is an improvement for 44 states. Every state would get to go in either the first 4 or on a Super Tuesday every 12 years, which is frequent enough to have influence. While it would require National Party buy-in and changes in delegate rules to enforce, the incentives exist for most states to make this system happen.
