This is primarily a blog about physics, astronomy, and science writing, but every so often, I like to dip my toe into discussions about politics or economics. I am not a political scientist or an economist, so you can take my opinion at face value. Besides, there are some good precedents for astronomers using their analytic skills to think about political questions.
I know the rate of astrophysics posts have gone down in recent months, but since I’m in the midst of writing my next book (“The Universe in the Rearview Mirror“), most of my science writing effort is going in that direction.
That said, I have a quickie today, and that is: Why is the primary calendar set up like it is. That is, consider the first three states in the primary calendar this year:
- Iowa (D +9.5%)
- New Hampshire (D +9.6%)
- South Carolina (R +9.0%)
Where the numbers in the parentheses are the margins of victory for each state in the 2008 cycle. It is clear to most observers that the field gets winnowed very early in the process so that states like Pennsylvania (where I live) get virtually no say in the process whatsoever.
Of course, it makes sense that there isn’t a single, one-day national primary. It would be ruinous financially to candidates from both major parties. At least in principle, having a slow winnowing of the process makes some sort of sense. But the order, at present, is ridiculous.
Certain states should almost certainly be ignored in the primary. California, for example, will almost certainly go Democrat, and has done so in every election since 1992. Obama won it by 24% in the 2008 election. Texas, on the other hand, will almost certainly go Republican, and has done so in every election since Reagan won it in 1980. It’s worth noting that both Iowa and New Hampshire have gone to both Democrats and Republicans in recent elections, but that they’re only worth a combined 11 electoral votes. South Carolina (worth eight) last went Democrat in 1976, and has gone Republican ever since.
My point is that there are a relatively small number of swing states, and that both parties should have it in their best interest to court the actual preference of those states from the outset. I propose that it makes sense to have primaries in order of the gap in the previous presidential race. In this case, it would be:
- Missouri (R +0.1%)
- North Carolina (D +0.3%)
- Indiana (D +1.0%)
Or, perhaps even better, the order could be determined by the difference in margin from the national average. In 2008, Obama won by 7.3%. A Republican would have to improve his numbers by that margin (on average) to win the race. So it would make more sense for the Republicans to run their primaries in states where Obama won by very close to the national average:
- Virginia (D +6.3%)
- Colorado (D +9.0%)
- Iowa (D+ 9.5%)
Or this could be tweaked still further, including a weighting for more compact states or something like that.
The point is discussion is simply that if both parties want to nominate a candidate who has the maximum probability of winning under the current rules (winner take all within a state) while minimizing their expenditures during the primary, the current system seems very ill-suited.