Improving the (Republican) Primary
This is a continuation of Improving the (Democratic) Primary, which discusses how delegate allocation should change to be more conducive towards a general election strategy rather than introducing terms that are counter-intuitive towards it. Similarly, it is interesting to note that the Republican strategy is nearly identical. It awards states delegates in accordance with their electoral power and then provides bonuses based on party loyalty/accomplishments. Essentially, this amounts to both parties overvaluing safe states that neither change nor are at risk of changing. It is therefor relevant to consider the following formulation that is geared towards winning the general election:
Part Two: Primary Voting
Currently, neither party gives much tactical consideration to how states conduct their primaries. Republicans (and Democrats) make use of penalties (and bonuses) to encourage states to conduct timing in a particular way, but neither party considers the way in which states conduct their vote (i.e. Caucus vs Primary, Open vs Closed, etc).
The Washington Post and an independent group have done analysis to show how primaries pick the eventual (Republican) nominee more often than caucuses. In fact, since 1976, only seven states have flawless records, and they all use primaries.
However, what we should be concerned with is whether those selected candidates go on to become President. With that in mind, the success ratio for the Republican Primaries in 2016 is:
It would be fair to combine similar election types. Election years versus Democratic incumbents, election years following a Republican Presidency, or election years following a Democratic Presidency. Examining data back to 1980, the following axiom holds true: Primaries rule, caucuses drool.
It is interesting to note that primaries and caucuses also vary in the sense that they use closed and open systems. Closed systems expressly limit participants to party members, while open systems enable more voters to participate. A rarer form, a blanket primary, differs in that it enables voters to participate in both primaries.
It is intuitive why primaries might succeed in selecting Presidents more than caucuses since they are more reflective of a popular vote. It would, therefore, also make sense to assume that blanket primaries > open primaries > closed primaries in selecting Presidents. Let’s examine the success rates from the 2016 election to test this hypothesis:
It is evident, from the 2016 results, that open systems are superior to closed systems (I would be curious to find an analysis that shows whether this is true over time, but open/closed data is more elusive than primary/caucus data).
[As an aside, there were two states that used hybrid systems whereby unaffiliated voters can vote in the primary and are thereafter registered with that party. Both of those hybrid systems nominated the candidate who would become President.]
Now, let’s focus on diversity with regards to scheduling. Given that every previous metric we’ve considered has shown that the more that metric attempts to mimic the general election the more successful it is — it would follow that states that more closely resemble the typical (swing state) voter would be better at selecting Presidents.
Using data from the Census (on income and education) and the Kaiser Family Foundation (on race), as examples, this is what we find:
The general trend is that more affluent, educated, non-white states are better at selecting candidates that will succeed in the general election. However, these trends only represent the 2016 election and aren’t exactly reliable for determining scheduling.
What would be more prudent is to reward states that are successful with their selections with earlier primary slots. This affords them greater influence with regards to setting trends and identifying quality candidates in a broader field.
However, it should be taken into consideration that, at least in theory, states that are smaller in both population and geography make it more tenable for quality (grassroots) candidates, who may possess less wealth and/or access to organizational systems.
It is also worthwhile to maximize exposure candidates have at any given time by compartmentalizing primary schedules. Taking these concerns into account, the next major primary should look something like this:
This schedule was created such that the states are prioritized by how often they select Republican candidates that win in the general election (using data since 1980). It prioritizes smaller states before larger ones so as to create more equitable conditions at the onset. It also groups states by general location so as to minimize travel time and maximize exposure, with the exception of the beginning where a “carve out” mentality is still preserved. Such scheduling efforts promote candidates with the highest chances of success.
In summary, in order to improve the (Republican) primary process, we ought to move away from caucus systems and towards general primaries. We also ought to embrace open systems that are more inclusive rather than closed ones. Finally, we should endeavor to create schedules more conducive to selecting the highest quality candidates.
