Why Not Make the U.S. Primary System Fair?
A Proposal for a New and Fair System of Presidential Primaries and
Caucuses for the United States of America
The Current System Is Broken
Our current system of primaries and caucuses is a ridiculous system that would give Rube Goldberg a run
for his money.
• The current system gives highly unequal influence to different states and regions.
We now give three very small states -- and the same three very small states, Iowa, New Hampshire, and
South Carolina -- enormous and unwarranted influence on the choice of the next President in every
More generally, the majority of states, because their primaries and/or caucuses come late in the calendar,
now have very little influence on the parties' selection of their Presidential candidates.
• The current system is so unfair it is unstable.
Given the inequities of the current system, there is a huge and rational incentive for individual states that
now come late in the process to "game" the system and try to cut to the front of the line. Because of this
instability two major states, Michigan and Florida, were effectively unrepresented in the process this year.
However, political inertia aside*, these flaws are easily fixable.
* admittedly, a big proviso
The Fair Primary Proposal:
For the purposes of primaries and caucuses, the country will be apportioned into
Rationale: Regional primaries will allow the candidates, the media, and the nation to focus on issues of
regional importance. Short travel times would also allow candidates to visit every state in the run up to a
regional set of elections. Currently, because of geographically-incoherent election assemblages like
Super Tuesday, candidates are often logistically forced to focus on only a few of the largest states, and
ignore the smaller and geographically isolated states. In contrast, the regions proposed here are
apportioned into comparably large electoral-vote blocks, with contiguity, geography, and historical
association in mind (see details below).
To be clear, I use the term "regional primary" here and elsewhere, but this should be read to mean
"regional set of single-state primaries and caucuses." That is, the states would still organize and
administer their primaries and/or caucuses independently, in the manner of their choosing, and party
delegates would continue, of course, to be chosen by states. The only change under this proposed
system is that they would do so at the same time as others in their region.
The order of the regional primaries, and of the single-state primaries (see next
section), will be determined by lottery each election cycle.
Rationale: The main instability of the current system is due to the importance of a state's order in the
sequence of primary elections, with early states having far more influence. By assigning order randomly
by lottery, this will distribute political clout equitably over time. The lottery would be procedurally
transparent, supervised by an independent technical committee, and kept free from political influence.
The lottery should be held sufficiently early to allow preparation by parties and candidates; I would
suggest two years before the next election. This lottery, I believe, would have the additional benefit of
increasing political engagement, not only by fairly distributing political power in the Presidential selection
process among people in all states, but also by the news and excitement, if say, Hawaii, Vermont, or
Texas, were to be selected for one of the first three single state primaries. And as an event, for the
political junkies out there, it would rank up there with the NBA and NFL draft for excitement and
Although not strictly necessary for fairness in the long run, I believe two constraints to the random
ordering would be advisable: (1) The region that has the first primary in a given election cycle will not be
eligible for the first primary in the next election cycle, although it may have any other position. (2) The
region that has the last primary in a given election cycle will not be assigned the last primary in the next
election cycle, although it, too, may have any other position.
These two constraints simply act to limit the degree of "luck," either good or bad, that a region may have
in successive election cycles. The single-state primaries would be subject to a similar constraint (see next
Three states will be randomly selected by lottery to have the first, single-state,
Rationale: The proposed fair primary system will begin with three single-state primaries (or caucuses), as
is currently the case with Iowa, New Hampshire, and South Carolina. These single-state primaries serve
two purposes: (1) They allow candidates and parties with relatively little money a chance to be heard and
to compete in smaller, less-expensive venues, as well as allow all campaigns a chance to self-organize
and hone their message, and be vetted, before the larger regional primaries start. (2) The single-state
primaries also serve a civic purpose in allowing the entire country to focus in detail on three states, their
issues, their politics, their economy and their people. However, with our current system, while Iowa, New
Hampshire and South Carolina have enough ads and town hall meetings to make many of their residents
positively annoyed by the quadrennial onslaught, the majority of states, like Oregon, Hawaii, Colorado,
New Jersey, South Dakota, Alabama (among many others), get almost no attention at all. Many citizens in
these states would actually appreciate the attention, and the country would learn more about itself each
States selected for one of the single-state primaries would, of course, not participate in their regional
primaries that election cycle. (In the case of California, which because of its size is the only single-state
"region", its selection for a single-state primary would eliminate one regional primary.)
States that have one of the initial single-state primaries in a given election cycle will be disqualified for the
single-state lottery in the next election cycle, although they would participate normally in their regional
primary, of course. Again, although not strictly necessary for fairness in the long run, this constraint will
prevent short-term concentrations of "good luck" that some would perceive as "unfair" on the small
chance that, say, North Dakota were awarded the first primary two elections in a row (although this would
still be much less unfair than the current system, by a long shot).
The regional primaries would then follow the single-state primaries, with one
regional primary every two weeks, in the order determined for that election cycle by
Rationale: This is fairly straightforward. A biweekly spacing between regional primaries, or something very
close, is a schedule that would allow the completion of all the elections before the summer conventions,
without the same lulls and fits and starts of the current haphazard system.
Here's an example of how the system could work for the next, 2012 Presidential elections.
The result of this system would be a straightforward and fair system of selecting our
President -- the most powerful political office of our nation -- which would give equal
weight to every state, every region, and every citizen in the United States of
To me, this seems a worthy goal.
To the right is a proposed
regional primary map
The core of the proposed system:
• For each Presidential election cycle,
the order of the regional primaries
would be assigned by randomized
• Also, in every election cycle, three
states (from three different primary
regions) would be randomly selected
by lottery to have the first three,
single-state, primaries or caucuses.
Fulfilling the role (but sharing it fairly)
of Iowa, New Hampshire, and South
Carolina today (details below).
I know that there are people who will consider a rational
approach to correcting our political absurdities as a sort
of category error. But the proposed system is far more fair
and sensible than the current hodge-podge approach,
and I cling tenaciously to the concept that democracy
should aspire to fairness and reasonableness, despite
much evidence to the contrary.
This is Elliptical Research Contribution 2008.2.
I welcome comments on this proposal at the email: email@example.com
Copyright © 2008 Timothy D. Kailing