, , , ,

On the off chance that the American political and chattering classes ever conclude that they might have responsibilities other than waging the Russiagate wars, some new Labor Department data might usefully remind them of the economic pressures faced by so many of their compatriots that resulted in the 2016 presidential vote producing such a stunning – and anti-establishment – upset. Because there are no signs that the situation has improved meaningfully since election day.

The figures describe trends in weekly wages in the United States county-by-county for the 344 largest American counties and how they changed between 2015 and 2016. They’re awfully revealing when viewed collectively, but they become especially illuminating when overlaid on the election results, and add to the case the fragile and/or falling living standards were major contributors to Donald Trump’s victory.

The headline wage development was that average weekly pay in the United States overall fell 1.5 percent between the fourth quarter of 2015 and the fourth quarter of 2016. As the Department put it:

This is one of only eight declines in the history of the series, which dates back to 1978. The 1.5 percent decline in average weekly wages was the largest decline since fourth quarter, 2011, when average weekly wages decreased by 1.7 percent. The most recent decline occurred in 2016, when the U.S. average weekly wage decreased 0.6 percent over the year.”

Even worse, this drop-off isn’t adjusted for inflation. Although by conventional measures, the cost of living hasn’t been rising much lately, it’s still been rising. So in real terms, weekly wages decreased still more. And it’s hardly a stretch to suppose that this development had more than a little something to do with so many voters’ willingness to roll the dice and put a political outsider and novice in the world’s most powerful position.

The correlation between economic distress and pro-Trump votes comes through much more clearly when broken down by county. But before examining these statistics, two big context-setting points need to be made.

First, according to the Associated Press, Mr. Trump won an astonishing 2,626 out of 3,113 total U.S. counties. His Democratic Party opponent, Hillary Clinton, won only 487. So candidate Trump triumphed in more than 84 percent of counties.

Second, however, as is not the case with, say, House of Representatives districts, the size of counties varies enormously. Just one example: As of 2006, the population of Los Angeles County (won by Clinton) was larger (10.29 million) than that of 42 states. And indeed, Clinton won the popular vote precisely because she won so many of those gigantic urban counties.

Still, looking at the 20 large counties where weekly wages fell and rose by the greatest percentages from late 2015 to late 2016 shows that the former heavily went for Mr. Trump and the latter for Clinton – a pattern similar to that demonstrated by state-level growth numbers I looked at recently. (The county-by-county vote total can be found on this New York Times interactive graphic.)

The specific numbers? Seven of the ten counties where weekly wages worsened the most went for the president in 2016; only three voted for Clinton. Of the ten second worst wage performers, six were won by Mr. Trump, three opted for Clinton, and election data was unavailable for one county (Anchorage, Alaska).

But in my view, the correlation between economic performance and Trump and Clinton voting shows up most strikingly when examining the winning counties. After all, Clinton won very few counties to begin with. In fact, on a percentage basis, her results in the worst performing large counties were better than her county-by-county results overall.

This was also the case with the large counties where wages rose the fastest on a percentage basis. But what stands out to me is the difference between the share of these counties the Democratic candidate won, and the share of total counties she won.

The Democratic candidate won six of the ten large counties where wages performed best, and eleven of the 17 “next ten” performers (there were several ties). In other words, Clinton captured nearly 63 percent of these winners, versus fewer than 16 percent of total counties large and small.

These data don’t conclusively resolve the debate over whether Mr. Trump’s appeal (and Clinton’s failure) was rooted in economic distress or other issues (racism, sexism, server-gate).  We’d need data going back farther than that. But they do underscore the continuing gap between the apparent priorities of official Washington and its various satellites on the one hand, and much of the public on the other. And they raise the question of whether, in 2018 and 2020, the voters will respond by throwing the rascals in both major parties out once again.