Yet another piece of conventional wisdom undone by Donald Trump’s electoral upset has been the Democratic advantage in the political ground game, particularly in cities.
The advantages of population density combined with urban demographics had long made Rust Belt cities such as Philadelphia, Pittsburgh, Cleveland, Detroit, Milwaukee, and Madison strongholds for Democrats. And for decades, Democratic voters in those cities were enough of a counterweight to offset Republican rural voters in Pennsylvania, Michigan, and Wisconsin.
But as my colleagues Linda Poon and George Joseph write, Trump rammed right through the Democrats’ “blue wall” in this part of the United States. The president-elect pulled this off by boosting turnout in smaller and mid-sized former industrial cities, which ate away at the margins produced by those larger cities.
With some data provided by Joshua Darr from an earlier FiveThirtyEight post, we mapped both of the campaign’s field offices in the Rust Belt. On the map below, the locations of campaign field offices in early October hint at how the Trump campaign saw a new way to find votes: small and mid-sized cities. The blue dots represent locations or clusters of Democratic field offices and the red dots represent Republican field offices.
Trump vs. Clinton Field Office Locations
Note that this is a snapshot of the field offices that existed in early October for both Clinton and Trump. But to really see the difference in Trump’s strategy, it’s helpful to compare it with Mitt Romney’s presence in Pennsylvania in 2012:
Romney Campaign Pennsylvania Field Offices in 2012
The above map shows the location of Romney’s field offices in Pennsylvania in 2012. His campaign did have offices in mid-sized cities such as Harrisburg, York, Lancaster, Allentown, Wilkes-Barre, and Scranton, but for the most part they clustered in and around Philadelphia and Pittsburgh, where suburban voters were believed to be the crucial swing vote.
Trump Campaign Pennsylvania Field Offices in 2016
The second map, above, shows Trump’s much wider net in the state, and highlights what the Clinton campaign might have missed about the potential of those smaller and mid-sized cities.
Trump effectively flipped counties with mid-sized cities. In Erie County, the margin of votes went from 17 percent for Obama over Romney to a difference of only 2 percent between Clinton and Trump. Luzerne County, which includes the city of Wilkes-Barre, went from a 5 percent margin of victory for Obama over Romney to a nearly 20 percent advantage to Trump.
Trump’s margin also increased in counties that include the smaller cities of York and Reading, while Democratic leads in Lancaster and Harrisburg dwindled. Obama’s 2012 lead of 27 percent in Lackawanna County, which includes Joe Biden’s hometown of Scranton, fell to just 3 percent for Clinton versus Trump. Those margins accrued to give Clinton thousands of fewer votes than Obama and gave Trump thousands more votes than Romney. When precinct totals are released next week, we’ll be able to see more precisely how much the vote in those cities changed.
The strategy was clearly a success for Trump, who ultimately carried states that had reliably voted for Democratic presidential candidates since at least 1988. The pattern also highlights a cultural disconnect between these small and mid-sized cities and their big-city counterparts.
CityLab spoke with Richey Piiparinen, director of the Center for Population Dynamics at Cleveland State University, about the extent to which the demographics of this region helped propel Trump to the White House.
CityLab: What did everyone who was surprised by the election result miss about cities in the Rust Belt?
Piiparinen: Deindustrialization started 1960 and 1970, so it's tough to say that it was the backlash from deindustrialization that brought Trump to the presidency. I know that's the most common narrative out there and to some extent it's true, but when you look at the fact that half of the people didn't vote who are registered in this country, and then the vote is split 25 percent apiece between the two candidates, you get a sense there's a few things going on. One was the 'change' candidate and people wanted change. Trump represented that.
Now, that does tie into the economic backlash against neoliberalism, or deindustrialization. In Cleveland and Detroit, the deleveraging for manufacturing was recent, 2000. So the displaced worker frustration is still fresh in these parts. And then in corollary with deindustrialization there is the ramifications when a place like Cleveland moves forward into the knowledge economy, with many skilled and prepared, and many not. So we see this bifurcation that has been going on in knowledge economy centers like Boston, which deleveraged from industry in the 1960s, and Pittsburgh did it in the '90s.
Finally there is a racial component, or 'whitelash,' and that is partly an effect of eight years of a black president. You see this playing out in the exurban turnouts in Philly, Cleveland, and Detroit, where Republican support increased. So it was really the confluence of the patchwork of white votes that he got for varying reasons that pushed him over the edge, and then the lack of turnout for Hillary.
Republicans expanded their base in these states from rural areas all the way up to mid-sized cities. To what extent are these kinds of cities simply not experiencing the growth and expansion associated with the global, knowledge-based economy?
The question is, do we want young creatives to concentrate in a few places on the coasts? This is Enrico Moretti's idea: let’s get all the young creatives in the most productive parts of the country. Let's figure out affordability so we can produce in San Fran, we can produce in New York City, we can produce in Chicago, and if we don’t, then overall growth will stagnate.
But we need to take into account how political power is doled out in America. Because if progressive votes are clustered geographically in a few places, then the power of the vote is lessened. A thousand votes in Ohio have ten times the impact of a thousand in Brooklyn. And so when you have such a divided nation, you have a loss of representation and an increased likelihood of an extremely reactionary American policy going forward that's not going to do well for economic productivity from a policy stand point.
Where do you think the national conversation needs to go to bridge this disconnect between positive global trends, and what people in these smaller and mid-size communities feel in their daily lives?
There's the realistic look at what's going on macro-economically, and then there's the nostalgic look. Unless we look at that clearly, then we're going to end up with Donald Trumps.
In the late 1800s in the agrarian economy, 80 percent of the country was employed in agriculture. Today that's 2 percent. Does that mean we don't make food, or we don't eat? No. It means that economic era matured and became automated and became its most efficient.
Next is the industrial era. The proportion of America employed in mining, construction, and manufacturing peaked in 1960 around 40 percent. Today that's 20 percent, with manufacturing alone at around 8 percent. It will likely be five or so percent going forward. Now that doesn't mean we don't make things. Manufacturing output is near all-time highs. It means the industrial economy is maturing, like the agricultural economy before it.
So, yes, the problem is the disconnect between the “two Americas” in relation to dislocated workers. What do we do with the labor that we once needed not too long ago? That's the big problem and there's no easy answer. And of course in places like the Rust Belt the glut of labor is still pretty acute. Cleveland's population peaked in 1950. It did so because it had a macroeconomic reason to, but Cleveland will never have that macroeconomic reason to support a large population of factory workers en masse again. Our macroeconomic reason now is in life science research and that doesn't fill up factories, and it doesn't fill up the suburban homes or inner city neighborhoods, and it doesn't provide upward mobility at scale.
And it requires education and training, things that shift slowly rather than in the immediate aftermath of job losses.
And that's why Obama's plan to make community college free is right on. Because you're either going to train people for the new economy and spur innovation and knowledge, or you're not going to do that.
You have to keep in mind America's comparative advantage right now is to produce knowledge. It's no longer to produce trinkets. Rwanda is receiving low-skilled factory work from China now. China's going through the transition from industry to services that we've been doing since 1950s and 1960s.
So many of the factors of openness and diversity that make cities vibrant places economically seem to cluster in large cities. When you have smaller or mid-sized cities that don’t manage to attract those qualities, does it wall out those kinds of positive gains?
Our political system doesn't match up with the economic idea that you can just have a “spiky world” of productivity that can carry the day for the country. Because there's problems with having a spiky world, affordability problems but there's also political problems. Political power is controlled and disseminated differently than economic growth, and for a while each have existed on separate rails. Middle America is still powerfully a political battleground, but less so an economic powerhouse. The void in between produced President Donald Trump.