What the Jobs Report Says About the Election
Could the latest jobs report really have an effect on the outcome of the presidential election? Weirdly, it might -- and it's looking good for Hillary Clinton.
The conventional wisdom has been that Donald Trump gains from a bad report and Hillary Clinton from a good one. That's because Trump has ridden a wave of dissatisfaction to the Republican nomination, while the Democrat Clinton can argue that she will build on the successes of the Barack Obama administration.
But really? Would voters actually care about a report that is known to be sort of a random number generator? The initially published monthly change in payrolls is merely a ballpark estimate, subject to revisions that can exceed 100,000 jobs. Could anyone be swayed by a single data point that may or may not reflect the true state of anything?
Maybe. Consider the correlation between two numbers -- the unrevised monthly change in payrolls reported just before each of the last nine elections 1 (expressed as a percentage), and the share of electoral votes that went to the incumbent party's candidate. If the relationship is strong -- that is, if a good number tends to help candidates like Clinton -- the dots in the scatter plot below should more or less form a diagonal line from bottom left to top right:
Wow, they sure do (with the exception of one outlier -- the 1980 race in which Ronald Reagan beat Jimmy Carter). What's a bit comforting is that longer-term job growth appears to have predictive power as well, so it's possible that the pre-election report works only to the extent that it does reflect reality. Here's a plot comparing job growth 2 over the entire year before the last nine elections to the share of electoral votes won by the incumbent party's candidate:
So what does this mean for Hillary and Donald? Well, the October report was pretty good: The Labor Department estimated that nonfarm employers added 161,000 jobs in October, a one-month growth rate of 0.11 percent. Judging from the way the points line up in the first scatter plot 3 , that would give Clinton 304 electoral votes -- a good deal more than the 270 needed to win the election. The longer-term trend points in the same direction: In the year through October, payrolls grew by an estimated 2.4 million jobs, or 1.7 percent. That's roughly equivalent to 297 electoral votes for Clinton -- again a comfortable lead.
To be sure, there's plenty of reason to believe that this election could be different. Trump isn't your typical candidate, and various measures -- such as prime-age employment -- suggest more slack in the labor market than job growth alone does. But if you want to know what the headline number tells us about the result, there it is.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
I use only the last nine elections due to data limitations: The BLS doesn't publish unrevised initial payroll growth estimates for years before 1979.
Using revised data.
Excluding the outlier.
To contact the author of this story:
Mark Whitehouse at email@example.com
To contact the editor responsible for this story:
James Greiff at firstname.lastname@example.org