The Roots of Right-Wing Populism: Donald Trump in 2016

1 A widely read summary by a distinguished economist is Krugman (Citation2018): “Study after study has found that racial resentment, not economic distress, drove Trump voters.”

2 The literature is gigantic; Ferguson et al. (Citation2018) analyze the various streams in more detail than we can here.

3 See, e.g., the NBC/Wall Street Journal poll of July, 2016, available on the web at https://www.scribd.com/document/318507755/NBC-Wall-Street-Journal-National-Poll

4 See the text of Trump’s announcement speech, available at the UCSB American Presidency Project archive, at http://www.presidency.ucsb.edu/ws/index.php?pid=110306

5 Here again, the literature is very large. See, e.g. (Ferguson and Rogers Citation1986) for the earlier years; for later years, cf., the papers cited supra, Note 1.

6 We spent considerable time comparing various other approaches; see the discussion in (Ferguson et al. Citation2018).

7 For race, the scores run as follows. In Citation2008 the overall weighted mean was 3.46; for Republicans, 3.84; for Democrats, 3.18. In 2012, the corresponding scores were 3.52, 3.99, 3.11. In 2016, 3.19; 3.78; 2.63. The GOP primary mean in 2016 was 3.84; for Trump primary voters, 3.99. For sexism, the scores run as follows. The overall weighted mean for Citation2008 was 2.54; for Republicans, 2.80, for Democrats, 2.38; in 2012, the corresponding numbers were 2.48; 2.71; 2.28; for 2016, 2.34; 2.88; 2.03; the GOP primary weighted mean was 2.85; for Trump voters, 2.74.

8 The racial resentment and modern sexism coefficients are roughly equivalent to attitudes about the ACA, and at least twice as big as the other coefficients.

9 The (Autor et al. Citation2016) paper has since been expanded. Our reference is to the earlier one and not the most recent version, which raises additional issues this article cannot take up. For other countries, see, for example (Alabrese et al. Citation2019); (Dipple, Heblich, and Gold Citation2015).

10 Much of our congressional district data ultimately derives from the U.S. Census Bureau’s American Community Survey. But we drew from other datasets as well, such as that produced by the Social Science Research Council’s Measure of America, available at http://www.measureofamerica.org/congressional-districts-2015/. We also used Policy Map’s compilations for the 115th Congress. Data on presidential voting by congressional district came from the set constructed by the Daily Kos, at https://www.dailykos.com/stories/2013/07/09/1220127/-Daily-Kos-Elections-2012-election-results-by-congressional-and-legislative-districts. We used the Daily Kos’ compilation of data on religious affiliations, at https://www.dailykos.com/stories/2018/1/7/1728838/-The- Daily-Kos-Elections-guide-to-the-nation-s-religious-populations-by-congressional-district. Also helpful was a set of economic and social data issued by the Economic Innovation Group of Washington, D.C. We used their 2017 study, which drew on data for earlier years. Their recent studies are available at https://eig.org/dci.