Wargaming the Electoral College

A fun project, but one which took longer than I expected.

Using Excel 97’s mapping feature, I took the state-by-state polling results and made a map of the 2000 election results. New sets of cells were then created so that polling results from next November would be reflected state-by-state.

How’d I do that? Simple — too simple. Gore received about 47.8% of the vote in 2000. If you assume that the Democratic candidate this year gets 48% of the vote, then divide 47.8 into 48 to get a multiplier of 1.004. Apply the multiplier to the total votes cast for Gore in 2000, and you get a simple prediction for whoever the Democratic candidate is in 2004.

If Dean (or unDean) gets less than Gore did in the popular vote, then the multiplier becomes a divider.

As I said, it’s too simple — the spreadsheet assumes voting patterns will remain essentially unchanged over a period of 4 years. Given the fact that each party is going to start with 40-42% of the voters, I’m going to go with that assumption. However, things like turnout — in case the election looks like a blowout one way or the other — could change things fairly dramatically.

But why let facts get in the way of good speculation?

First, here’s the map the Bush-lovers want to see. Assume a 52% vote for Bush — which isn’t unlikely, given his current poll numbers:

Bush 52 Dem 47 Green 1.jpg

In this case, I’ve given the Green candidate 1% of the vote, since I doubt Nader will be able to create the kind of excitement he got last time around. Also notice that at 47%, the Democrats cease to exist in the Midwest, with the lone (although EC-rich) exception of Illinois. Even Pennsylvania, Oregon, and Michigan tip to the Republican column.

So how can the Democratic candidate win? It’s going to take the Upper Midwest, and New Mexico, and Florida, as this 50D, 49R 1G map shows:

Bush 49 Dem 50 Green 1.jpg

Even then, it’s a close race.

In all fairness, I honestly don’t know what Nevada is going to do, nor do I know how to simulate it. The state has grown immensely, even since just 2000, and the voter mix is hard to predict. But I’d bet the influx of people has made the state more competitive for the Democrats, rather than less. But also note that in none of these scenarios would Nevada tip the scales by its lonesome.

Most interesting, I thought, was the next map. Let’s say something goes wrong. Really wrong. So wrong that a sitting President gets only 47% of the vote in a two-way race. What happens then? This:

Bush 47 Dem 52 Green 1.jpg

You’ll note that when the Republican candidate loses 47/52, he loses by 138 electoral votes. But look above again, and you’ll see that when the Democratic candidate loses by the same figures, he loses by 180. In other words, without Ross Perot around, the Democrats remain electorally challenged.

Finally, for you red meat conservatives out there, I plugged in the numbers from Bush I vs Dukakis. Here’s how Bush II would fare with daddy’s popularity:

Dukakis Numbers.jpg

Well, Bush II doesn’t get as big a win as Bush I. 370 e-votes to 426. Why? Because the Democrats have a lock on California — and right now, that looks to be their only saving grace.