Electoral Math
Reality-Based^TM Political Numbers from Nicholas Beaudrot

Sleater-Kinney, Live at the Showbox, December 14th | Home | Bumper stickers on the back of a Toyota Camry, Seattle, WA

The States that Howard Dean's support: an analysis

Howard Dean's campaign website has a place where you can  pledge your support to Howard Dean in the upcoming primary, complete with a list of the number of pledges for each state. The top of the list isn't terribly surprising:

[note: statistics are as of December 16, 2003. The pledge count on Dean's website is updated regularly.]

Gee. Five of the top six states in total pledges are the most populous states in the country. The only outlier is the tech-heavy, environment-friendly state of Washington. (I would say "yeah Washington", but I think the Dean support here lines up with the strange populist vibe I get from the place. There's a reason I'm voting for Clark in Washington's primary caucus). So there aren't any surprises on this list besides Washington. We can try and get a more accurate picture of Dean's support by calculating the rate at which voters are pledging to vote for him, rather than simply looking at the raw number of pledges. As a quick and dirty measure, I took the 2002 census population estimates for persons of voting age, and divided the number of pledges into that number. For example, in Vermont there are an (estimated) 476,930 persons of voting age, and 1022 pledges to vote for Dean, or roughly 467 voters for every pledge. Here are the top 10 states on that list (a lower number means more support for Dean):

Looking at this list, it's hard to find a common thread for the states that have heavy support for Dean. We can probably safely ignore the level of support in Vermont, DC, and New Hampshire, as there is a heavy bias in these areas. Clearly New Hampshirites take pride in their primary season kickoff, so they are probably paying more attention to politics than the rest of America. It's unlikely, therefore, that their support for Dean is in any way comparable to those states where people are paying less attention to the pre-primary season. Likewise, Vermont is probably going bonkers at the prospect of putting one of their own in the White House (even if he is from New York), and the DC number probably includes many folks who pay attention to politics for a living. The rest of the list contains six states that went to Gore in the 2000 election. Two were blowouts (Massachusetts & Connecticut), two were moderatly close (Washington & Maine), and two were very close (New Mexico & Oregon). The only state on this list that did go to the GOP was North Dakota. All the states had more than a 3% vote for Nader, though North Dakota's was only marginally over 3%. Noticeably, there are zero states on the list from either the Midwest or the South, both of which are seen as regions where any candidate must succeed in order to win the presidency.

So now we've got a better picture of where Dean has support, but we haven't found states where Dean's support might help him win the general election. He'll need to pick up states that Bush won in 2000 in order to take the White House. For the sake of argument, let's assume that the 2000 presidential election results are likely to repeat [this is reasonable, to a first approximation. Between 1992 and 1996, only six states change voting habits.]. So here's the list of states Dean has strong support, but which the Democrats lost in the 2000 election.

The top three states of NH, ND, and VA combined give Dean the election; however it seems reasonable that this list overrepresents the amount of support for Dean. Let's throw out New Hampshire and Virginia again. In addition, we might want to consider slightly discounting North Dakota, Arizona, and Oklahoma as well, since these are "Super Tuesday" states where candidates have already started campaigning as well. In any event, winning either Arizona, Virginia, or Missouri alone would tip the election in favor of the Democrats. Winning Colorado would result in an exact tie (here you can find procedures in the event of an electoral tie).

We've so far got a decent list of places where Dean has the most support, but this doesn't necessarily mean that Dean stands a good chance of winning those states. For instance, Montana could be populated by a small number of people who find Dean very appealing and a large number who are certain to vote against him. Since there's no way to measure the number of Dean haters just on this pledge, we might be overestimating Dean's chances to win that state. The next step, then, is to look at states that were a close call in 2000. Let's take the list above and add a few more columns: the Bush-Gore-Nader breakdown from the 2000 election (the Federal Election Commission has these numbers), an extra column that combines the vote for Gore and Nader, and the difference between Bush and the combined power of Gore and Nader. We'll assume that Gore was right, and that everyone who voted for Nader would, in a two-party election between a Democrat and a Republican (instead of a Republicrat and a Democrican), vote for the Democrat.

Suddenly, the flyover states of North Dakota, Montana, and Oklahoma aren't looking so promising. But the list of states where Dean has support is still quite informative, particularly because of what states aren't on the list. How far down the list do we have to go to find states in the Midwest and the South, where Democrats supposedly must have strength in order to win? Here are the states in those regions that Bush won in 2000.

[note: Nader did not appear on the ballot in North Carolina.]

This is a somewhat shocking list. The Democrats have very little ability to pick up any ground in the Midwest--Ohio is really their only real shot--and no chance at picking up electoral votes in the South, except Florida. For all the hype surrounding "NASCAR dads", and Dean's infamous appeal to "the guy with the confederate flag on his pickup truck" [can someone find me a direct quote, please? I don't want to misquote the man], Dean has significantly more appeal in mountain states than he does in the South. In reality, this is a continuing trend; Republicans have been replacing Southern Democrats in Congress since the Lyndon Johnson administration. At this point, much of the South is simply too Republican to be swayed by a Democrat who happens to have an accent. We've reached a split in the south where Gore couldn't even carry his home state in 2000. Considering Bush won many southern states by margins of over 10%, it really looks like the idea of spending campaign dollars in the South is just money poured down the drain, as Bush will steamroll his opponent everywhere below the Ohio River.

There's an important corollary to this point as well. The margin of victory between Bush and Gore in 2000 was so large that any candidate, even a Southern Democrat like Edwards or Clark, stands a good chance of losing every single state in the South.

I should offer a quick caveat here. It's important to note that the presidential candidates haven't spent much time campaigning in the South at this point in the election cycle. It's possible that as the candidates begin to spend effort attracting Southern voters, Dean's support in those areas will rise. But for now, even in the early primary state of South Carolina, Dean is having trouble mustering support --SC ranks 43rd in total support for Dean at this point.

Let's put it all together. Suppose it's July 4, 2004, and I'm Joe Trippi for a day. Where should I concentrate my resources?

1. Ohio
2. New Hampshire
3. Missouri
4. Florida
5. Colorado
6. Arizona
7. Pennsylvania
8. Michigan
9. Minnesota
10. Iowa
11. Wisconsin  

The last five states went to Gore in 2000, but were quite close in the popular vote. Considering Bush is likely still receiving a boost from a positive view of his handling of the War on Iraq and the War on Terrorism, it's possible that the Democrats will need to campaign hard in those states just to keep them in the upcoming election

If the pledge results accurately reflects Dean's popularity in various states, and if the Democrat/Republican split hasn't changed radically since 2000 in any particular state, and if Dean ends up as the Democratic nominee (and those are all very large ifs), then punting South to Republicans is in fact a reasonable strategy. The Democratic coalition that started breaking down when Strom Thurmond left the Democratic party may in fact be totally dead. I guess there really are two Americas.

Now if I could only turn this into a color-coded map, I could probably parlay this into a consulting gig with the Dean for America. :)

| | technorati

Home | Mail  | RSS

Last updated by Nicholas Beaudrot on 11:32 13 February 2005