Republican presidential nominee Donald Trump speaks during a campaign stop last week in Johnstown, Pa. (Justin Merriman/Getty Images)

Johnstown, Pa., is the site of two great tragedies perpetrated by industrial titans.


The first tragedy occurred more than a century ago. An artificial pond used for recreation by Pittsburgh's steel barons over-topped the dam holding it back in May 1889. A wall of water washed through the hilly area around Johnstown and plowed into the city. The flood and an ensuing fire in the debris killed more than 2,200 people.

The better-known tragedy is more recent: Johnstown was decimated by the collapse of the American steel industry. Overproduction in the 1970s led to a collapse in the market, quickly shutting some steel mills in a swath across western Pennsylvania and Ohio and causing more to stumble over the ensuing years. The day after Christmas 1983, Bethlehem Steel cut 2,500 jobs in the city, a number equivalent to about 6 percent of the total population. In 1977, the per capita income in Johnstown was 88 percent of the national figure. A decade later, it was 74 percent.

Johnstown is the sort of place where Donald Trump should do well. Blue collar. Economically frustrated. It's in a state that's generally considered a battleground, and so Trump showed up there on Friday afternoon.

But, in reality, it's hard to see why.

As it stands, Hillary Clinton leads in Pennsylvania by more than six points, according to the latest RealClearPolitics average of head-to-head polls in the state. In theory, that sounds like it's fairly close. In reality, and in comparison with other states, it isn't.


There are currently 10 states where the contest is closer than it is in Pennsylvania, according to RealClearPolitics, states that together add up to 160 electoral votes. (They're the states whose margins fit within the gray box, above.) There are three states within three points where Trump could win 44 points — two of which, Arizona and Georgia, being ones that he would be expected to hold since Mitt Romney won them in 2012. (And John McCain won them in 2008, and George W. Bush won them in 2000 and 2004.) Yes, he needs to win some states that Romney lost, by why not then focus on Florida, which has substantially more electoral votes and is currently a closer race? Why not head back to North Carolina, which President Obama won and then lost?

But, really, why not go to Ohio?

Johnstown is a bit east of Pittsburgh, in the western part of Pennsylvania. It's about a two-hour drive to Youngstown, Ohio, the northeastern city in that state that perhaps best exemplifies the same economic struggle that Trump clearly hoped to tap into in Johnstown. Why stop by Pennsylvania, where you trail by more than Clinton trails in Texas (Texas!), instead of shoring up support in Ohio's Mahoning County, where your message would be expected to resonate just as well?

Cambria County, where Johnstown is located, is already Republican territory anyway. For years, it was solidly blue-collar blue, like so much of the Rust Belt. But that has switched, with the county backing Romney by 16 points in 2012. Four years earlier, the result was about even, while the state went overwhelmingly for Obama.


Mahoning County, home of Youngstown, is the sort of place that Trump has repeatedly argued he can swing. It's still heavily Democratic, in part because the city is more heavily nonwhite. (Mahoning County is 77 percent white; Cambria County is 93 percent white.) But it's also bigger: About 136,000 people live in the Pennsylvania county compared with 232,000 in Ohio. Move a few thousand of the voters in Mahoning County to head to the polls, and it could make the difference in the state. CNN's recent poll in Ohio didn't break out numbers by city, but Trump fared better in every part of the state except Cleveland and Columbus.

So: Why?

One theory is that Trump stopped in Pennsylvania because he needs to expand the map if he's going to be elected. But again: Why not focus on Florida, which Obama won four years ago? If national polls shift back toward Trump, Pennsylvania will be closer, sure, but so will a lot of states, including Nevada, Wisconsin and North Carolina. If Pennsylvania is in play in two weeks, it means that the map has changed significantly and the momentum is so heavily with Trump that he can look elsewhere to tip a state into his column.

Going to Pennsylvania might (might) move the state a few fractions of a percentage point to Trump. In Ohio, though, that fraction of a percentage point is more likely to be the difference between victory and defeat. Rallies like the one Trump held can be good to boost turnout, but getting a few thousand more Republicans to the polls in Johnstown and nearby areas isn't going to win a state where Clinton leads by a margin that seems likely to translate into a few hundred thousand votes. If the goal was to get spillover media attention in Ohio — why not just rally in Ohio?

Perhaps Trump has internal polling that shows him closer in the state. Possible! Is it possible that Trump's internal turnout estimate (the number used to estimate what voters will make up the electorate) are more accurate than all of the pollsters who have looked at Pennsylvania? It is. Is it likely? Not particularly — especially given Trump's antipathy to polling over the course of the campaign.

Maybe Trump decided to go to Johnstown because of what it represents. Johnstown starred as Charlestown, the beaten-down home to a minor league hockey team, in the Paul Newman movie “Slapshot.” It was the run-down town in “All the Right Moves.” The picturesque hills that floodwaters washed against in 1889 and the grimy, rusty factories down by the river helped make it the Hollywood archetype of the small-beaten-down-by-the-changing-economy. Maybe that's why Trump the showman figured it was a good place to hold a rally.

Youngstown, of course, is the focus of a Springsteen song. That's much less Trump's style.