The Early States Should All Vote at the Same Time

Why do we let Iowa and New Hampshire decide our President? Why let any state decide? Surely all states should have an equal say, right?

Well, a one-day national primary would be tough. Every candidate would have to run a nation-wide campaign without the support and resources they’d have in a general election. And there wouldn’t be much chance to narrow down the field – there were 12 Republican candidates on the ballot in Iowa this year, and the winner only got 28% of the vote. That latter problem could, I suppose, be resolved with some sort of ranked voting system. But early primaries give parties a chance to react to new information. Votes, donors, and endorsements will move away from candidates who disappoint in early states, especially states they’ve heavily campaigned in. If Chris Christie can’t crack 8% in New Hampshire, his supporters will realize it’s time to look elsewhere.

Having “early state” primaries is not, in itself, such an awful idea. But are the right states voting early? In 2016, Iowa, New Hampshire, South Carolina, and Nevada are the only states allowed to vote before March. Do they deserve that privilege?

These are the properties you want in an early primary or caucus state:

  • Small population
  • Small land area
  • Representative of the nation
  • Representative of a region

In less populous states, candidates can meet more voters face-to-face. In geographically smaller states, voters are easier and less expensive to reach; traveling to them is easier, radio and TV ads can cover ground more efficiently, and each yard sign will attract more eyeballs. The smaller and less populous a state, the easier it is to get known without having to rely on media attention. Thus, you can get a sense of someone’s appeal among voters in that state. But for that information to be relevant, you want the state to be representative of the nation, or at least of a specific region.

 

Which States Should Vote First?

To find the “ideal” early states, I crudely added up these properties. First, I took the rank of all states from least populous to most populous (so Wyoming is 1, California is 50, and so on). Then I did the same thing for land area. Then I took the state’s Cook PVI, not accounting for party (so Texas and Massachusetts are both 10). The higher a state’s PVI, the more dominated it is by a single party – and, therefore, the less likely that it’s representative of the nation.

I added all three numbers. The states with the lowest total numbers are the smallest, least populous, least partisan states.

What about the fourth property, regional representation? Well, if we have four early states, they should each come from a different Census region:

And let’s ignore states that border another region, since they’re probably not representative. (Who still considers Delaware Southern?) Also, let’s only look at continental states (sorry, Alaska and Hawaii).

So according to my system, the ideal early states are the four states with the lowest number in each region, excluding states that border another region.

And the winners are… drumroll… Iowa, New Hampshire, South Carolina, and Nevada.

Yes, really. No, I wasn’t trying to get that result, it just… happened. It turns out those states work just fine!

Midwest Northeast South West
IA (50)

WI (59)

MN (69)

MI (74)

 

NH (18)

RI (20)

VT (26)

ME (27)

CT (32)

MA (52)

NY (79)

SC (47)

MS (48)

VA (54)

LA (56)

AL (64)

NC (67)

FL (75)

GA (79)

 

NV (62)

ID (70)

OR (71)

WA (74)

UT (81)

AZ (89)

CA (107)

 

A Proposal

The problem isn’t that the wrong states vote early. It’s that they vote one at a time.

If someone’s strongest in the South or the West, but weakest in the North and Midwest, they start out at a disadvantage. By the time South Carolina comes around, two states are already lost, and the media might have moved on. This makes no sense. It undercuts the whole purpose of early primaries and caucuses – to provide data on who can win, and where. Instead we’re getting limited, distorted data. And nobody wants to alienate Iowa or New Hampshire, which gives certain groups – such as corn farmers – disproportionate political power.

Iowa, New Hampshire, South Carolina, and Nevada should all hold their primaries and caucuses on the same day.

https://trivialanalysis.files.wordpress.com/2016/02/earlystates1.png?w=960

(map source)

The winnowing process would be both faster and more accurate. Iowa and New Hampshire might get ignored entirely by some candidates. And why not? The Iowa Caucus tells us barely anything, except who is most liked by Iowans. There is no reason it should have priority over the other three states.

 

Hypothetical: A Better Schedule

So the first four states have voted. Now what?

Well, why not go down the list? After week 1, the field will be narrower. Week 2 could be sort of a re-do of week 1 with fewer candidates – and with fresh information on how each region might vote. The next-best states in each region, according to my method, are Wisconsin, Rhode Island, Mississippi, and Idaho. I like how that works out; all of those states are quite different from the first batch. Wisconsin, a populous swing state, would be a big prize to win so early on. But someone with an urban base might bet on Rhode Island from the beginning.

https://trivialanalysis.files.wordpress.com/2016/02/earlystates2.png?w=960

My “no border states” rule cuts out a few cultural regions. The mid-Atlantic, Appalachia, the Southwest, and the Great Plains states haven’t gotten a say. For week 3, let New Jersey, West Virginia, New Mexico, and North Dakota vote. They’re all high-ranking states from those areas, and they’re all from different census regions. A candidate who hasn’t won a state yet might hope for a last stand in one or two of these.

https://trivialanalysis.files.wordpress.com/2016/02/earlystates3.png?w=960

Then what? I sort of like the next 3 states on the list: Minnesota, Vermont, Virginia, and Oregon. Now the West Coast is finally represented; and in Virginia, you have another major swing state.

https://trivialanalysis.files.wordpress.com/2016/02/earlystates4.png?w=960

And that would round out the first month of voting. All parts of the country have been represented. Anyone who hasn’t won a state by now doesn’t stand a chance. By the second month of voting, only the most viable candidates will remain.

https://trivialanalysis.files.wordpress.com/2016/02/earlystates5.png?w=960

After the first month, states with fewer delegates should get more priority. That way, smaller states would matter because they’d vote early, and late states would matter because they’d yield a lot of delegates. Also, nobody would accumulate too many delegates early on, so the party would have time to react to new information (such as a scandal for the front-runner). California, Texas, Florida, and New York would all vote on the final day of primaries.

If four states voted every week, it would take about 13 weeks – 3 months – to hold primaries and caucuses in all 50 states (not counting D.C. and the territories). Voting could start in mid-March and last through mid-June. Hold the first debates in January. Campaigns could start as late as December.

 

Something Needs to Change

That’s all just fun, idle speculation. But why not speculate? The way we pick Presidents is seriously flawed. Better ideas, even implausible ones, ought to get tossed around.

The 2016 primaries last from February 1st to June 7th – more than 4 months. (And that’s an improvement from 2012, when primaries went on for 5 and a half months.) In February, only one state votes every week. Then, suddenly, most states vote in March. The states that vote in April, May, and June probably won’t matter, even though that’s more than half the schedule.

Our current nomination system lasts too long and gives too much power to Iowa and New Hampshire. It is designed to prevent a contested convention at all costs. As a result, most of the campaign happens in the eight months before voting begins – and then the actual voting only matters for a few weeks. This strikes me as totally backwards.

There’s no single, obvious solution to this. I like Jay Cost’s proposal to have an early convention that selects five candidates. And a ranked voting system would probably be superior in multi-candidate races like the 2016 Republican contest.

Unfortunately, such ambitious reforms seem to be a long way off. In the meantime, the current system still has plenty of room for improvement. Changing the nonsensical schedule should be the parties’ top priority.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s