When you hear reports on the unemployment rate it’s usually a single number. For example, in October that number was 4.9 percent. But that single number is the national average, and can conceal a wide range at the state and local level.

For instance, in September South Dakota and New Hampshire had the lowest rates in the country—2.9 percent—while six states (Nevada, Mississippi, West Virginia, Louisiana, New Mexico, and Alaska) all had rates that were twice that number.

Not surprisingly, states with unemployment rates that are rising tend to vote differently in national election than states where people are having an easier time finding jobs. As Anna Louie Sussman of the Wall Street Journal notes, of the 17 states where unemployment rose over the past 12 months, Trump won 13 while Clinton only won four.

Take the swing state of Pennsylvania, which, until this year, hadn’t voted Republican since 1988. It saw its unemployment rate rise from 4.8% in October 2015 to 5.8% in October 2016. Mr. Trump found a e audience there with voters who’ve seen manufacturing and other jobs vanish and a rise in immigration, two issues he returned to repeatedly over the course of his campaign.

What about the states where unemployment fell over the year? In those, the two candidates fared about evenly, with Mrs. Clinton winning 16 to Mr. Trump’s 17.

While the unemployed were not the sole reason for Trump’s victory, the numbers do indicate that regional increases in unemployment (especially in Rust Belt states) caused many voters to take a chance on him.

Whether Trump will do much for the long-term unemployed remains to be seen. But both political parties need to start paying more attention to the plight of the jobless.