Predicting the election outcome months ahead of time: discussion and link to revised paper with Kari Lock

There are two aspects of a presidential election that can be predicted: the national popular vote and the relative positions of the states. The national popular vote can be forecasted months ahead of time given the economy and other predictors. for example using Doug Hibbs’s model:
hibbs6.png
.

(As I wrote a few months ago, “the incumbent party sometimes loses but they never have gotten really slaughtered. In periods of low economic growth, the incumbent party can lose, but a 53-47 margin would be typical; you wouldn’t expect the challenger to get much more than that.”)

The relative positions of the states don’t actually change much from election to election:

2004_2008_actual.png

You can do slightly better by using polls. As Matthew Yglesias puts it, “the large number of public polls on something like a presidential election makes the outcomes quite easy to forecast based on crude measures. What’s more, even absent polling, Presidential election outcomes seem to be pretty predictable based on nothing more than macroeconomic variables.”

Actually, even the February polls turn out to be pretty good–when combined with previous election results–to pin down the relative positions of the states.

Bayesian combination of state polls and election forecasts

Here’s the revised version of my article with Kari Lock [updated more recently] in which we forecast the election using Hibbs for the national popular vote, and a weighted average of last election (corrected for incumbency) and the February polls to get the relative positions of the states.

Lots fo fun stuff there, including this prediction (based on February Clinton-McCain and Obama-McCain polls) of which states Clinton or Obama were expected to win in November:

kari.png

3 thoughts on “Predicting the election outcome months ahead of time: discussion and link to revised paper with Kari Lock

  1. Just glancing at the top chart, it seems like the incumbent party benefits less from income growth if the party has held the presidency for 2 or more terms. Johnson in '64, Reagan in '84, Nixon in '72, and Eisenhower in '56 did very well running for reelection after taking the presidency from the other party 4 years earlier. Humphrey in '68, Stevenson in '52, and Ford in '76 had a difficult time holding their party's control of the White House for more than 2 (or more than 5, in Stevenson's case) consecutive terms even in times of >1% income growth.

  2. Your paper is pretty interesting. I have one major criticism. It appears (from figure 4(b) in the linked paper) that you introduced a significant amount of error into your predictions by overstating a candidate's 'home state advantage'. Your 2008 predictions for the four home states of the 2004 and 2008 candidates (TX, IL, MA, and AZ) are all off. The Texas prediction is especially egregious because, like Indiana, it's in the wrong quadrant (you picked the wrong winner) and your prediction was way off. I think there is a 'home state advantage', but it doesn't work the way you modeled it (+6%). Bush had very little advantage in Texas (where I live) compared to any other Republican. On the other hand, there's probably no way to model what happened in Hawaii, where Obama got a huge 'favorite son' boost even though his connection appeared tenuous at best.

    Overall, I would describe your predictions as the campaign playing field. It gives a way to judge the effectiveness of the respective campaign organizations. Obama clearly picked up Indiana by fielding a better operation than McCain. He pushed North Carolina into Dem column when it was a tossup. He won Virginia by more than predicted, held Florida despite some intangible advantages for McCain.

    Missouri is the only close state that McCain carried, although Obama should be disappointed by his margin in Ohio. Obama poured a huge amount of resources into Ohio at the end of the campaign and the result was still closer than it should have been.

Comments are closed.