Offshoring divides US incumbent support along education lines
Offshoring increases incumbent support in graduate-heavy US counties but reduces it in manufacturing-heavy counties. A Bank of Canada working paper published October 16, 2026, finds these effects largely cancel out overall.
Education separates electoral winners
Anna Maria Mayda, Lindsay Oldenski and Walter Steingress study US presidential, House and Senate elections from 2000 to 2020 using 74,657 observations.
Their baseline finds no statistically significant aggregate effect on incumbent-party vote shares.
Doubling offshoring exposure lowers incumbent support by 2.52 percentage points at the fifth percentile of college education but raises it by 1.83 points at the 95th percentile.
Presidential effects range from -8.27 to 5.11 points.
Counties with more offshorable jobs or manufacturing workers generally reduce incumbent support; the manufacturing interaction is not statistically significant for House elections.
A 1 percent rise in exposure increases county employment by 0.08 percent overall, with gains concentrated in more educated counties.
Tracking exposure across local economies
The authors combine US Bureau of Economic Analysis foreign-affiliate employment with election data, allocating industry changes using commuting-zone employment shares from 2000.
Their preferred measure covers the top 50 percent of industries ranked by production sold back to the US.
Controls include imports, immigration and labor demand; export and trade-agreement checks preserve the main patterns.
Alternative exposure measures produce positive aggregate effects.
The data exclude outsourcing to unaffiliated foreign suppliers.
No uniform electoral verdict
The authors link electoral responses to uneven domestic employment effects, rather than a uniform rejection of offshoring.
Instrumental-variable checks support the education result but provide mixed evidence for occupation and manufacturing effects.
The findings are not official Bank of Canada positions.
Source: Offshoring and Voting Patterns in the US
IN: