In counties where manufacturing and other export-intensive industries are most important to their economies, President-elect Donald Trump nearly ran the table in the 2016 election.
A new analysis from Emsi shows that almost without exception in key battleground states such as Michigan and Wisconsin the counties with the highest concentration of manufacturing, agriculture, and mining and oil and gas extraction employment voted for Trump.
Only two counties in the six Midwest/Rust Belt states analyzed had more than twice the national share of jobs in these three sectors and voted for Hillary Clinton. Those were Green County, Wisconsin, and Muskegon County, Michigan.
(See our interactive graphic here at Emsi Labs).
The same counties who overwhelmingly voted for Trump have seen an increased dependence the last 15 years on factory jobs and ag and mining employment. Every county where the concentration of these jobs increased substantially from 2001-2016—meaning they’ve become even more reliant on manufacturing, ag, and mining—swung in favor of Trump over Hillary Clinton.
Overall, 437 of the 489 counties in the six states analyzed (Michigan, Ohio, Minnesota, Wisconsin, Indiana, and Pennsylvania) voted for Trump. Many of the 52 counties in these states where Clinton won are more urban and have a more diverse mix of large-employing industries.
Manufacturing employment in the U.S. has declined 25% since 2001. That’s more than four million lost jobs in a sector that typically brings new income into regions through the products its exports.
Trump has focused on restoring America’s manufacturing vitality and keep jobs in the U.S. He and Vice President-elect Mike Pence made waves recently by intervening to keep 800 jobs at Carrier Corporation in Indiana.
In Upper Midwest counties that voted for Trump, manufacturing, ag, and mining employment declined drastically during the Great Recession and have bounced back at a faster rate since the downturn than in counties that voted for Clinton. Two counties stand out on both sides of the spectrum.
Elkhart County, Indiana, where Trump prevailed, is one of the largest and most dense manufacturing areas in the country. Around half of the payroll jobs in Elkhart County are in manufacturing, and the concentration of factory jobs there has grown at a much faster rate than the nation. Meanwhile, Philadelphia County, Pennsylvania, which went for Clinton, has lost nearly half of its manufacturing jobs, going from 40,000 in 2001 to 21,000 in 2016. But manufacturing makes up just 3% of payroll employment in Philadelphia County (nationally, the share is 8.2%) and has continued to shed jobs post-recession.
About the analysis
Emsi looked at employment data and election results for counties in six Rust Belt and Upper Midwest states that played a key role in the president election: Michigan, Ohio, Minnesota, Wisconsin, Indiana, and Pennsylvania. For each county, Emsi tracked the change in the concentration or density of jobs in crop and animal production (agriculture), manufacturing, and mining, quarrying, and oil and gas extraction. The analysis was based on location quotient, a standard metric to determine how concentrated a single industry or cluster of industries are in a particular region compared to the national average.
Read more about our analysis in the Washington Post.