The ‘Bread and Peace’ Model of election forecasting

In a comment to my post on what really matters in predicting presidential election outcomes, where I said that political scientists tend to look at things like the change in disposable income, reader Vincenzo kindly provided a link to an essay by political scientist Douglas Hibbs where he spells out the model more precisely. Hibbs has what he calls a ‘Bread and Peace’ Model, which looks at what influences the percentage share of the two-party vote received by the incumbent party’s candidate.

Hibbs says that “Aggregate votes for president in postwar elections are well explained by just two fundamental determinants: (1) weighted-average growth of per capita real disposable personal income over the term, and (2) cumulative US military fatalities owing to unprovoked, hostile deployments of American armed forces in foreign conflicts. No other objectively measured, exogenous factor systematically affects postwar aggregate votes for president.”

Hibbs gives the results of his analysis in the form of a graph that shows the relationship, with the blue line the regression curve and the location of the dots on the vertical axis the percentage of votes received by the incumbent party, as a share of just the two-party vote (i.e., votes for any third parties are not included). Above 50% means that the incumbent party candidate was elected since the electoral votes tend to go with the popular vote, except in the 2000 election. Given the combined contribution of the above two factors at any given time, one can read off the expected share of the two party vote by the candidate of the incumbent party. In 2012, that would be Obama.

According to this model, Obama’s chances for re-election look pretty bad at this point, with him predicted to get only 45.5% of the vote. Of course, this prediction is based on the data as of the end of 2011. But even if the fatalities of the current wars continue at their relatively low rate through election day, for him to have a good chance of re-election Obama needs per capita real income growth to “average out at an annualized rate of more than 4 percent per quarter until the end of the term”, and such a rapid recovery of the economy seems highly unlikely.

Note that the blue line correctly ‘predicts’ incumbent party wins in in 1956, 1964, 1972, 1984, 1988, 1996, and 2004. It also correctly predicted incumbent party losses in 1952, 1960, 1968, 1976, 1980, 1992, and 2008. The one failure was in 2000 where according to this model Al Gore should have won easily with close to 55% of the two-party vote, whereas he got just barely over 50% but lost in the electoral college in the famously disputed election.

(Hibbs also has a ‘Bread and Incumbency’ model that predicts the outcomes of congressional elections, which is determined by “(1) the number of House seats won by the in-party at the previous midterm election, which registers the impact of institutional advantages enjoyed by incumbents in the US single-member district, constituency service-oriented legislative system, and (2) weighted-average growth of per capita real disposable personal income over the congressional term.” According to that the current probability of the Democrats regaining a majority of the House of Representatives is zero. They are actually likely to lose 5 seats.)

It would seem that the 2012 election was the Republicans for the taking as long as they nominated some bland centrist candidate who did not make any waves and just let the economic tide take him safely into harbor. Mitt Romney may have seemed the perfect choice for this with the only fear being that he might not be interesting enough to galvanize the voters. But as is often the case with the best laid plans, this too has ganged agley, with a militant party base, fed and watered by candidates like Michele Bachmann, Rick Perry, Herman Cain, Rick Santorum, and Newt Gingrich not content to let the party go through a pro-forma primary campaign but throwing it into turmoil.

Whether this turbulent primary is sufficient to overturn the Bread and Peace model prediction is doubtful since the model does seem pretty robust and Obama is in a pretty deep hole. Hibbs does provide Democrats with a glimmer of hope, saying “Yet remember that every election is affected to some degree by idiosyncratic factors, which at times are important enough to overwhelm the persistent influence of fundamentals. Indeed idiosyncratic events contribute a lot of the fun to political affairs and their unexpected appearance and impact from one election to the next are why many of us follow election cycle developments so carefully in the media.”

But that is a tenuous hope. Before an election, political pundits tend to think that the current election is different from the ones that came before. While it is true that the current Republican race for the nomination is indeed the craziest that I have seen in my lifetime, that does not mean it is crazy enough to defy the predictions of the statistical model. Such things are usually judged after the fact, when an election outcome is different from the model’s predictions.