This is the abstract of a new paper titled "Local Economic and Political Effects of Trade Deals: Evidence from NAFTA", which argues that many "white, less educated voters left the Democratic Party" due to NAFTA:
Why have white, less-educated voters left the Democratic Party? We highlight the role of the 1994 North American Free Trade Agreement (NAFTA). In event-study analysis, we demonstrate that counties whose 1990 employment depended on industries vulnerable to NAFTA suffered large and persistent employment losses after its implementation. Voters in these counties (and protectionist voters regardless of geography) turned away from the party of President Clinton, who promoted the agreement. This shift is larger for whites (especially men and those without a college degree) and social conservatives, suggesting that racial identity and social-issue positions mediate reactions to economic policies.
I don't have access to the paper at that link, but an earlier ungated version is here. The abstract of the earlier version says:
Why have white, less educated voters left the Democratic Party over the past few decades? Scholars have proposed ethnocentrism, social issues and deindustrialization as potential answers. We highlight the role played by the 1994 North American Free Trade Agreement (NAFTA). In event-study analysis, we demonstrate that counties whose 1990 employment depended on industries vulnerable to NAFTA suffered large and persistent employment losses relative to other counties. These losses begin in the mid-1990s and are only modestly offset by transfer programs. While exposed counties historically voted Democratic, in the mid-1990s they turn away from the party of the president (Bill Clinton) who ushered in the agreement and by 2000 vote majority Republican in House elections. Employing a variety of micro-data sources, including 1992-1994 respondent-level panel data, we show that protectionist views predict movement toward the GOP in the years that NAFTA is debated and implemented. This shift among protectionist respondents is larger for whites (especially men and those without a college degree) and those with conservative social views, suggesting an interactive effect whereby racial identity and social issue positions mediate reactions to economic policies.
One problem with this story is that Republicans were more responsible for NAFTA than were Democrats. NAFTA was mainly negotiated by the Bush administration. Then when Clinton was elected, he got behind it and added labor and environmental provisions to the deal, with the aim of making it more acceptable to unions and progressive groups (and causing some free traders to abandon their support). While Clinton and Al Gore were the public face of NAFTA passage, it's clear that Republican politicians were more supportive than Democrats were. In the House, the Republicans voted 132-43 for NAFTA, whereas Democrats voted 156-102 against. And in the Senate, the Republicans voted 34-10 for NAFTA, whereas Democrats voted 28-27 against.
So why would the passage of NAFTA steer voters towards Republicans? The Republican presidential candidates after Clinton's time in office were trade liberalizing people such as George W. Bush and John McCain. Why would the previously Democratic voters annoyed with Clinton's NAFTA support shift to a Republican party led by these free traders? Why wouldn't the passage of NAFTA steer voters towards protectionist Democrats, of which there were and are plenty? At the local level, in the districts that were affected by increased trade with Mexico, I suspect that the Democratic politicians were just as strongly, if not more so, against NAFTA than their Republican opponents were. I'd love to see some case studies of this though: What were the candidates in these districts saying about trade and NAFTA?
Perhaps the authors would say that Clinton's prominence in promoting NAFTA's passage made people think the Democrats were more supportive of free trade, even though this was clearly not the case. Or maybe they would just say something more general like "voter ignorance." Regardless, I think any analysis of the impact of NAFTA on voting patterns of affected districts needs to grapple with this issue a bit.
Obviously, there's no way to evaluate the counterfactual in which there is no NAFTA, to see how voting patterns would have evolved in its absence. But my instinct has always been (and polling tends to confirm this), that trade policy doesn't have much of an impact on voting, and it's the other factors mentioned that are a better explanation of the shift in party preferences. It's my sense that the Democrats became more socially liberal in important ways during the period in question, and that's probably a key factor in the shift in political preferences of social conservatives.
