Mitchell’s Musings 12-26-16: Wrap Up for 2016

21 Dec 2016 6:52 PM | Daniel Mitchell (Administrator)

Mitchell’s Musings 12-26-16: Wrap Up for 2016

Daniel J.B. Mitchell

This musing will be the last for 2016. As we did last year, we will discontinue musing during the coming winter quarter due to teaching obligations. Unless something that requires a special musing occurs, we will resume in April 2017. So here is some wrapping up for the year related to the presidential election.

First, in last week’s musing (of December 19, 2016), we asked who might play the role that George Shultz – with his academic/research background - did during the Nixon administration. Nixon had fantasies about cabals that were tinkering with official statistics to make him look bad. He also wanted to manipulate data in his favor. But Shultz played a calming role.[1] We know that President-elect Donald Trump doesn’t take kindly to anything that contradicts his views of how things are or should be. But, as of last week, there were no academics in major roles that might protect official data from harm when the figures do not display supportive results. We did note that the one academic who prominently supported “Trumponomics” was Peter Navarro, a professor at UC-Irvine, but that he had not been appointed to any official position. According to the latest news, however, he has been appointed to oversee a new trade office in the White House.[2] So let’s hope that Navarro serves as a voice for honest official data production.

Second, in a musing of three weeks ago, we suggested that when it came to explaining Trump’s narrow win Electoral College win, any simple unicausal story – e.g., it was all angry displaced white males – was inadequate. So let’s elaborate a little more on that notion. Suppose you want to tell a story that is some variant of “it’s the economy, stupid.” Economic models of election outcomes tend to focus on where the economy was moving before the election. In Appendix A to this musing, we show the percent change in nonfarm employment by state, divided up between “Clintonland” – the states which Hillary Clinton won – and “Trumpland” – the states that went to Donald Trump. Was it the case the Clintonland was prosperous and Trumpland was depressed?

There was in fact little difference between Trumpland and Clintonland. According to the latest (preliminary) data, employment in the former rose by 1.5% over November 2015-November 2016. In Clintonland, the rate over that period was 1.6%. Moreover, both areas showed a mix of results with some states in each doing notably better or worse than others. Trumpland did contain some states adversely affected by low energy prices, states that experienced employment declines. (West Virginia – with its much-discussed depressed coal industry – actually grew in jobs at a 1.4% rate.)

The three states that went for Trump that weren’t “supposed to,” and in which Green Party candidate Jill Stein pushed for recounts, showed mixed results. Michigan actually outpaced the national average. Pennsylvania was very sluggish (but not negative); Wisconsin had slower than average job growth. The Ray Fair economic model predicted early on that Trump would win – but that model referred to the popular vote (not the electoral vote) so it was wrong.[3] However, we can simply take that prediction to mean simply that a generic Republican should have beaten (in popular votes) a generic Democrat. Trump didn’t win the popular vote and squeaked by narrowly in certain swing states – presumably because of his high negatives. Clinton did better than “expected” in popular votes, but apparently did not overcome her own high negatives in states that mattered.

The chart below has some further information that is part of the 2016 story. It shows November-to-November annual changes in nonfarm employment going back to 1940. (The November month selection matches our Appendix’s state-level data and, in any case, an official estimate for December 2016 is not available at this writing.) When you look at the chart, you should focus on the line at 2.5% growth. All recessions until 2000, regardless of their cause, showed significant “snap backs” of rapid growth (above 2.5%) at some point in the post-recession expansions. But the two recessions after 2000 never attained 2.5%. So there were two downturns and then an extended period of sluggishness, even though the unemployment rate came down to something like a full employment number eventually.



Given the macro picture, a generic Republican was predicted to win by a purely economic model rather handily against a non-incumbent in the popular vote. And that win didn’t happen. Moreover, lots of unusual events occurred: an FBI announcement, Russian hacking, recordings of nasty language, a hostile takeover of a major political party, a more extended primary season than expected for the other party, etc. Each event tilted a few votes one way or another. That is the story of the 2016 presidential election.

By April 2017, when these musings resume, we’ll know more about the consequences.


Appendix A: Employment Change in Clintonland and Trumpland


% Nonfarm Employment change

Nov 2015-Nov 2016

Clinton States









District of Columbia
















New Hampshire


New Jersey


New Mexico


New York




Rhode Island












[1] See also the musing of two weeks ago (December 12, 2016):  


[3] Fair comments on his error: Why such a large error? While this is not possible to test, most people would probably say that it is due to Trump's personality. Had the Republicans nominated a more main stream candidate, they may have done much better---much closer to what the equation was predicting. The prediction from the equation from the beginning in November 2014 was that the Republicans were heavily favored. The election was theirs to lose because of the economy and the duration effect, and they almost lost it! See  

Employment Policy Research Network (A member-driven project of the Labor and Employment Relations Association)

121 Labor and Employment Relations Bldg.


121 LER Building

504 East Armory Ave.

Champaign, IL 61820


The EPRN began with generous grants from the Rockefeller, Russell Sage, and Ewing Marion Kauffman Foundations


Powered by Wild Apricot. Try our all-in-one platform for easy membership management