Sunday, October 18, 2015

Forecasting the 2016 election economy: social issues + the long arc of history


 - by New Deal democrat

This week housing permits will be reported.  That will give us most of the long leading indicators for Q3, and that means we can begin meaningfully talk about whether things like GDP, employment, and personal income will still be positive in Q3 2016.  Since these are some of the better-rated economic indicators found by Nate Silver as predictive of Presidential elections, it also means we can begin to talk meaningfully about whether the incumbent Democrats or the opposition Republicans are likely to win the 2016 election.

But this is a good time to note that, to paraphrase Bill Clinton's famous slogan, "it's not *just* the economy, stupid!"  There are social and moral issues, and unless you think the Vietnam War, the Civil Rights Acts, Watergate, Willie Horton, the Lewinsky affair, and the Swiftboating of Kerry were irrelevant, I submit that they have had an effect on election outcomes.

If the sole determinant of an election were whether or not the economy is in a recession in the 3rd or 4th quarter of the election year, then Dewey would have won in 1948, Stevenson in 1952, Humphrey in 1968, Ford in 1976, Bush in 1992, and Gore in 2000.

Even adjusting to trend, almost all economic indicators weren't just positive, but *very* positive in 1968 and 2000. Humphrey and Gore shouldn't have just won, but won in a landslide.

So clearly the economy isn't the sole determinant of winning the Presidential election.

Probably the best analysis of the relative import of social vs. economic issues was contained in Thomas Schaller's "Whistling Past Dixie," the tome which forecast Obama's winning electoral coalition. Schaller wrote that, in Dixie, social issues are gatekeepers.  Candidates' views on the economy aren't even considered until they pass through a social "screen" of acceptability.  Schaller was writing from the Democratic point of view, but I suspect a similar dynamic plays out on the left.  If there were a candidate with Bernie Sanders' economic populist views, but right wing views on abortion, gun control, and gay rights, would progressives really be interested?  In 2012 GOP Gov. Mike Huckabee of Arkansas was pretty much that, and I recall zero interest in his candidacy by the left.

The bottom line is, I believe voters do in fact vote their pocketbooks, among those candidates whose views on moral or social issues they find basically acceptable.  Thus much effort is expended in campaigns to make the opposing candidate simply morally unacceptable.

Prof. Douglas Hibbs' "Bread and Peace" model takes at least one moral issue - war - and integrates it into his analysis, which makes use of the number of military casualties during a President's term. This method does at least partially explain Humphrey's loss, and subtracted at least a little from Gore's total (Serbia) and George W. Bush's and McCain's (Iraq and Afphanistan).

It is well beyond my competence to try to incorporate other "moral issue" into an election model.  It is simply well worth it to remember that such issues are important.  If the economy is going well, the opposition will focus on them.  If the economy is doing poorly, there is barely a need.

A second limitation of the model can be filed under "what have you done for me lately?" There is a line of thought that, after two terms of one party, there is an increase in voters restless for change, regardless of what the economy is like.  That certainly helps explain 1968 and 2000, but there aren't enough data points to enable reliable comparison.

That being said, it is certainly interesting that the concept of parties only lasting in power typically through two presidential terms is a relatively recent phenomenon.  In the long arc of US history, it is fair to consider the US as being a rising economic power from its inception through the 1940s, at its apogee in the 1950s and 1960s, and relatively speaking at least being a waning power from at least 1974 on.  Now let's compare against the frequency with which control of the White House shifted from 1800 to 1952:

Democratic-Republicans: 1800-28 (28 years)
Jacksonian Democrats: 1828-40 (12 years)
1840-60: multiple changes of power
1860-1884: Republicans (24 years)
1884-96: 3 changes of power
1896-1912: Republicans (16 years)
1912-20: Democrats 
1920-32 Republicans (12 years)
1932-52: Democrats: (20 years)

Since 1952, however, control of the White House has changed every 8 years with the exception of 1976-92, a single Democratic term followed by 12 years of GOP control.

While the US economy was in its ascendancy, voters were willing to trust a single governing party with control for long periods of time, switching when economic progress stalled.  Only the 20 years of turmoil leading up to the Civil War stand out as different.  Since the US economy reached its relative apogee, voters have behaved as if they felt the parties had no solution to the relative decline in their fortunes.

Although we probably can't measure it in any accurate statistical way, over the long term US voters are behaving as we would expect them to if they were voting on the waxing and waning of their economic well-being.