Number games.

Photographer: INDRANIL MUKHERJEE/AFP/Getty Images

Which Swing States Are Doing Great, and Which Aren't

Justin Fox is a Bloomberg View columnist. He was the editorial director of Harvard Business Review and wrote for Time, Fortune and American Banker. He is the author of “The Myth of the Rational Market.”
Read More.
a | A

The most eventful Republican National Convention in a long, long while is over. Soon we'll learn whether all the controversy and apocalyptic gloom helped or hurt Donald Trump's standing in the polls. In the meantime, though, there are some other numbers to chew on that might have relevance to November's election -- the state employment figures for June, out this morning from the Bureau of Labor Statistics.

The newly released data shows that of the 10 states most likely to be battlegrounds in the November presidential election, six have lower unemployment rates than the nation's 4.9 percent average, and one (North Carolina) is tied with it.

When you examine employment growth since the last presidential election (which seemed like as good a starting point as any), the swing states don't look quite so good. Only four of the 10 have been adding jobs at a faster rate than the country as a whole.

Will any of this matter in November? It might. Economists have been building models that predict how people will vote on the basis of economic conditions for decades. Sometimes they work, sometimes they don't. The best-known of these models, maintained by Yale University's Ray Fair, predicted that Mitt Romney would win the national popular vote in 2012 (news flash: he didn't). It's currently predicting victory for Donald Trump.

In the Fair model, faster gross domestic product growth and lower inflation translate into better odds for the party currently in the White House. While inflation is currently quite low, GDP growth -- which counts for more in the equation -- has been really, really slow compared with past decades. That leads me to several thoughts:

  • Maybe slower growth is the new normal -- a product in part of the aging of America and concomitant slowing in the growth of the working-age population -- and Fair will need to ratchet down the growth expectations inherent in his model.
  • Maybe slower growth is the new normal, but voters have taken a while to realize it. Now they have and they're really cranky about it.
  • Maybe the recent disconnect between weak GDP growth and strong employment growth (caused by extremely low productivity growth) is of electorate significance. At least over the short term, people care more about jobs than GDP growth, don't they?
  • Maybe, in this extra-strange election year, economic factors won't matter all that much.

Still, they must matter some -- which brings me back to economic conditions in those battleground states. State GDP data comes out with too long a lag to be of much use in election prediction, and state inflation data generally isn't available.  That leaves just the jobs numbers, which show that a lot of the swing states seem to be doing neither much better nor much worse than the national economy. Most of the states with worse-than-average job growth, for example, have better-than-average unemployment rates. But there are some interesting outliers.

Standing out on the upside is Colorado, which has very low unemployment and rapid job growth. Two states with a lot of electoral votes, Florida and North Carolina, are also doing better than national average in both unemployment and employment growth. Both had it pretty tough in the 2000s -- Florida because of the real estate bust and North Carolina because of big manufacturing job losses -- but lately things have been looking better. Which could help Hillary Clinton there.

Meanwhile, the two swing states where both unemployment and employment growth are worse than the national average are Ohio and Pennsylvania. Ohio's unemployment rate of 5 percent is statistically indistinguishable from the national rate of 4.9 percent (that is, it's within the margin of error), though, and its job growth doesn't lag all that far behind the nation's either. It's a state that's doing OK. Pennsylvania is another matter. As Trump put it, mystifyingly, in his strange interview with the New York Times this week:

I have statisticians, and I know, like if I went to Pennsylvania, I say, “Give me the statistics on what is going on with respect to manufacturing.” Numbers -- 45, 55, 65, I have states that are so bad.

Mr. Trump, I , too, have the statistics on what is going on with respect to manufacturing in Pennsylvania! And when I searched the monthly manufacturing employment numbers since 1990, I did indeed find a lot of 55s and 65s scattered among the data. Interestingly, I saw just one 45 pop up -- in July 2008, when manufacturing employment in the state was 645,400. The overall trajectory since 1990 looks like this:

That's actually pretty bad, although the worst damage appears to have been done in the early 2000s, and the number of manufacturing workers stopped declining in 2009. Still, while lots of other states saw manufacturing jobs rebound somewhat after 2009, Pennsylvania didn't. The state has experienced a very slow economic recovery overall since the Great Recession.

There was a small bright spot: the Marcellus Shale fracking boom, which propelled Pennsylvania into second-place nationwide after Texas in natural-gas production and created about 15,000 new jobs directly (and probably twice that many indirectly) in 2010 and 2011. But low prices for gas have since turned that boom into a bust, and most of those new jobs have disappeared since the beginning of 2015. Only seven states -- all with significant oil, gas or coal resources and all but one firmly in the red column in the last two presidential elections -- have seen slower overall job growth than Pennsylvania since 2012.

Recent presidential elections have been close in Pennsylvania, but the state hasn't actually gone for the Republican nominee since George H.W. Bush in 1988. The state's economy seems to have been doing what it can lately to give the GOP a boost, though.

  1. This is a judgment call, of course. Politico includes 11 states in its Battleground States Polling Average, but it's not at all clear why Michigan is there given that the state hasn't gone for a Republican in the presidential election since 1988 and Clinton is well ahead of Trump in most polls there.

  2. The fourth quarter 2015 state GDP numbers were released in June. As for inflation rates, at least one state calculates its own, but the Bureau of Labor Statistics just releases figures for regions and some metropolitan areas.

  3. They are, in order, with the worst job growth first, Wyoming, West Virginia, Alaska, North Dakota, Louisiana, Oklahoma and New Mexico.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Justin Fox at

To contact the editor responsible for this story:
Susan Warren at