Election prognostication, 1 of 2

One of the more interesting books I've read is The Keys to the White House: A Surefire Guide to Predicting the Next President by Allan Lichtman. He and his co-author crunch a number of variables to ascertain which ones most reliably determined who would get more popular votes in a Presidential election. The thesis he derived from this is that such elections are a referendum on the party in power. He posited 13 keys and if 6 or more turn against the party in power, it will lose the popular vote. This system correlates 100% with Presidential elections from 1860 to 2004. Bear in mind that it predicts popular vote and thus didn't pick the actual winner for 1876 (the controversial Hayes-Tilden election, the Gilded Age counterpart to the 2000 race), 1884 and 2000.

The keys are divided into four groups:
The first four deal with the political strength of the party in power in the White House.

• Does the party in the White House have more seats in the House after the midterm than it did in the last midterm?
• Is there a serious contest for the nomination (serious contest being defined as the winner has less than 2/3s of the delegate vote). Interestingly, such a contest in the challenging party doesn't correlate with election results.
• Is the sitting president running?
• Is there a significant third-party campaign? Significant being defined as getting at least 5% of the vote.

The next seven deal with the incumbent party's performance.

• Is the economy in recession during the campaign? This is more a question of perception of a recession than whether the economic agencies in question declare one. Indeed, such a declaration would not likely take place until after the election.
• Per-capita growth in the current term against the average of the prior two terms.
• Did the incumbent administration make major policy changes in the current term?
• Was there significant social unrest? This has to be on the scale of the Red Scare of the 1920s or the riots of 1968.
• Is the administration tainted by major scandal (think Teapot Dome or Watergate).
• Did the administration suffer a major foreign policy failure?
• Did the administration obtain a major foreign policy success?

The last two deal with the personality of the candidates.

• Is the incumbent party candidate charismatic or a national hero?
• Is the challenger either or both of these?

The standard here is quite high. For example, Grant and Eisenhower led victorious armies and obtained this advantage in elections occurring less than a decade afterwards. JFK and both Roosevelts had considerable charisma, as did Reagan.

Allan Lichtman did to an analysis of the keys as they could apply to 2008 in October 2007. In my next post, I'll discuss, and update, his analysis with my own take.