Forecasting the Electoral College

by John Sides on March 27, 2008 · 3 comments

in Campaigns and elections


Two recent analyses use current state-level polling data to forecast the Electoral College vote, assuming, of course, no systematic changes in support for the Democratic and Republican nominees.

The first is by Robert Erikson and Karl Sigman, and was posted over at They take a 50-state SurveyUSA poll and simulate the outcome of a McCain-Obama or McCain-Clinton race; the simulations take into account sampling error in the polls. Their findings:

…our simulations yield a 88% chance of Obama beating McCain (with 306 Electoral College votes on average versus 233 for McCain), and a 74% chance of Hillary beating McCain (with 285 Electoral College votes on average versus 253 for McCain). About one percent of our simulated outcomes were Electoral College ties. (We ignored within-state variation in Maine and Nebraska, which divide their electoral votes by district.)

The second is by Josh Putman over at Frontloading HQ. (That is his map above.) His data are different: the latest polls posted on Real Clear Politics. His results mirror Erikson and Sigman’s in that Obama is more likely to beat McCain than Clinton. However, he estimates that Clinton would actually lose to McCain:

In the McCain-Clinton contest, the solid and leaning categories give McCain a 235-179 electoral college vote advantage with 124 electoral votes falling in “toss up” territory. If you allocate those states’ votes to the candidate with the leading average, McCain wins by a 90 electoral college vote margin, 314-224…

And he finds that Obama’s victory would be narrower:

Factoring in the toss ups, Obama has a 199-174 lead over McCain with 165 electoral votes to close to call. Again, if those electoral votes are allocated to the candidate leading in the average of post-Super Tuesday state polls, Obama claims victory by a 273-265 margin.

Obviously, much could change between now and November, but these are still interesting exercises in ascertaining the current state of play, if only for cocktail party conversation.


Mike March 28, 2008 at 10:09 am

Texas and South Carolina= tossups!?

Seth Masket March 28, 2008 at 2:30 pm

Yeah, same with Massachusetts and New Jersey. I don’t buy it.

Josh Putnam March 29, 2008 at 9:20 am

What, no one wanted to bring up the Dakotas and Nebraska?

I think John’s last statement is instructive. I did this as a quick and dirty look at what the electoral college outcome would be if an average of the latest state polls reflected what the November vote outcome would be.

In the end, I think this underestimates Democratic support simply because the Democratic nominee isn’t known yet (not to mention the forecast doesn’t account for factors like the economy or presidential popularity). However, it is interesting that, even in the face of that fact, some of these states are as close as they are.

What this analysis demonstrates is that 1) Clinton’s electoral college argument to the superdelegates may not be a route her campaign wants to take and 2) Obama seems to bring more states into play while the Clinton map (see Frontloading HQ) largely resembles the map from four years ago (the battlegrounds are mostly the same).

Comments on this entry are closed.

Previous post:

Next post: