Abstract
In Identity Crisis, John Sides and Lynn Vavreck, joined by Michael Tesler (STV), offer a sequel to The Gamble, their analysis of the 2012 US presidential campaign. As in that study, Identity Crisis presents a data-rich, thorough, and wide-ranging quantitative analysis of the contentious nomination contests to the scandal-plagued general election campaign.
It is impossible in a brief review to do justice to the book’s nine chapters, appendices, and 80 figures and tables covering an exhaustive array of subjects in this wild campaign, but three main points can be distilled. First, observing “little evidence of any increase in ‘voter anger’ leading into the election year,” (p. 31) STV conclude voter anger and “a desire for change” (p. 155) had little impact on the election. Second, based on the Index of Consumer Confidence and GDP data, economic conditions in 2016 are portrayed as favorable for the in-party Democrats. Less rosy readings of the economy are attributed to partisan misperceptions. Finally, they contend the Trump campaign activated a sense of “white identity,” making racially charged, group identity politics central to the election. Americans have become more polarized, according to the authors, not by political ideologies but by attitudes about race and ethnicity and “what it means to be an American.”
Although a great deal of the extensive campaign analysis is quite well done, the three main points unfortunately do not hold up so well under closer scrutiny. First, the authors dismiss “anger” as having a major role in the election because Americans were no angrier than they had been in 2012. But whether anger increased from one election to the next is irrelevant. Anger unaddressed remains potent and many conservative Republicans in particular were very angry with the system, with their opposition, and even with their own party. Exit polls during the primaries indicated about 40% of Republicans were “angry” with the federal government and another 50% were “dissatisfied.” More than half of Republican primary voters felt “betrayed” by their own party’s leaders. The reason: most conservative Republicans (81% according to Pew) saw themselves as usually being on the losing side in policy conflicts. In contrast, only 44% of liberal Democrats were so frustrated. Contrary to STV, Republican anger contributed to Trump’s success (as a fighter) in the nomination contest and combined with polarization in the closing weeks of the campaign to bring home many Republicans uneasy about their party’s candidate.
Second, contrary to STV, economic conditions were not good in 2016 and did not favor the in-party Democrats. Although fewer voters in 2016 compared to 2012 were displeased with economic conditions (62% in 2016 compared to 76% in 2012 in the exit polls), large majorities in both elections saw the economy in negative terms. Because blame for the economy’s weakness in 2012 was directed at his predecessor, Obama had avoided the political fallout. This was not the case in 2016 (though, as a successor candidate, Clinton avoided some blame).
Two reasons account for STV’s misreading of the economy. First, they rely heavily on the change in (rather than level of) the more future-oriented (rather than present-oriented) Index of Consumer Sentiment. Things may have been a bit better, but were still far from good as both the exit polls and the Fed’s monetary policy attested. Second, their regression analysis of the historical relationship between GDP growth and the in-party vote (p. 158) did not correct in its estimation for the attenuating effect of the extraordinarily poor economy of 1980. Once 1980 is taken into account Democrats fared better than expected in the election based on the sluggish economy they had presided over.
STV also overstate partisan rationalization of economic impressions. Although partisanship colors individual assessments of the economy, as it does for everything else, the electorate overall responds to real economic conditions. Exit poll data from 2016 and 2018 make the point. In 2016, a Democratic plurality electorate assessed Obama’s economy negatively (36% positive to 62% negative). In the 2018 midterm, an electorate even more favorable to the Democrats evaluated Trump’s economy quite positively (68% positive to 31% negative). The difference in perceptions was not about partisanship. The difference was the economy (real GDP) was growing at a rate of only 1.6% in 2016 and around 3% in 2018. The economy in 2016 was relatively weak and the average voter accurately perceived this.
Third, there are good reasons to doubt that identity politics, and not ideological polarization, was central to the election. Are we to believe the two previous elections with candidates of different races were less racially charged than in 2016 with two same-race candidates? Not so long ago, many bemoaned the deterioration of social group connections in America. Now they are the driving force of our politics? If 2016 was so racially charged, why did black turnout actually decline (p. 180) and why did black and Latino support for the Democratic candidate decline by 4 to 5 percentage points (according to the exit polls)?
Why was the importance of identity politics exaggerated? First on the list is the harshness of Trump’s rhetoric, though his invectives were spread liberally beyond identity politics and he made noteworthy overtures to minority voters (“what do you have to lose?”). The racial inequality (aka racial resentment) survey measure used in the analysis may also have been misleading. It is of questionable validity and likely susceptible to both social acceptability and changing contexts effects (e.g. following Obama). But the main reason that STV exaggerate the role of identity politics is their dismissal of ideology as the basis for the public’s polarization. As with others before them, STV’s mistake is following Converse’s unfortunate conflation of ideology with sophisticated political thinking. The two are not the same. One can easily be ideological without thinking deeply about politics. It is done all the time.
Contrary to STV, Americans are ideologically quite polarized. Identity politics or group differences are a component and an outward manifestation of ideological polarization. Once ideology is taken into account, as Abramowitz and Saunders have documented, actual socio-demographics have little or no effect on partisanship.
Identity politics (particularly the immigration issue) did play a role in 2016, but as a “culture war” component of intense ideological polarization and not the driving force. As STV saw it, the public divided over the question of what it means to be an American. They may be right, but the division might be deeper. The difference may be over whether “what it means to be an American” is an important question - a fundamental clash between nationalism on the right and multiculturalism on the left. In that sense, the Identity Crisis title may be right on target.