Knowledge Production Intervenes in the Electorate
RECENT RESEARCH by faculty cluster members demands that we consider the ways in which knowledge produced about elections and voters can itself influence voting behavior. That is, we should recognize enterprises such as election polling, forecasting, and analysis not merely as representations of the electorate, but also as interventions in it. As these forms of knowledge circulate, they shape the way voters understand their own political concerns, efficacy, and community. This can push them not only to one candidate or party over another, but also away from electoral participation altogether.
Faculty cluster member Taeku Lee has been a leader in critical research on the influence of polling on political participation and representation. In his book Mobilizing Public Opinion, he recounts the history of social surveys’ rise in the 1930s, and their transformation of the public understanding of “public opinion.” This history is one of a shift from “public opinion” conceived as an abstract and speculative matter to one subject to the authority of scientific inquiry and reducible to a grid of survey data.3
The result, Lee argues, is that pollsters and survey research centers have an undue say in determining what count as significant—and even legitimate—issues of public opinion and debate. Lee carefully makes the case that polling can propagate norms of mass opinion that limit citizens’ political imaginations of what can and should be. In presenting themselves as “mirrors of society,” opinion surveys in turn delegitimize those claims (and claimants) that lie outside those norms— contributing to their marginalization, alienation, and exclusion.
Furthermore, these marginalizing effects are distributed unevenly across socio-demographic groups. Lee points out that polls focused on the voting electorate—or what pollsters would call “high-propensity voters”—create representations of public opinion that are biased along lines of race and gender.4 So rather than a passive “mirror,” opinion polling mediates how—and indeed whether— constituents can locate themselves in electoral politics. It is an intervening institution with excessive power, argues Lee, to direct as well as disaffect democratic participation.
If opinion polling constitutes a story about what and who are politically consequential through what it asks and whose views it pursues, post-2016 surveys have done a disservice to the effort to realize a more inclusive electorate. Though overall participation was up slightly in 2016, millions who voted in 2012 did not cast ballots, along with tens of millions more eligible voters. Yet despite innumerable studies and profile pieces on Trump voters, we know remarkably little about these non-voters. Too often this group is cast ipso facto as “disengaged”—a gloss that rationalizes excluding them and their perspectives. When polling fails to examine those who did not vote, it conveys that they need not be treated as factors in future campaign strategies. In this way, its intervention is to create bases for their continuing disenfranchisement.
In Lee’s view, the undue influence of polling in politics is best countered by a reinvigoration of the public sphere. Central to this is renewing civic debate and discussion among well-informed citizens—what Lee considers the true marker of “public opinion.”5 Experimental field studies have shown that, when given opportunities to deliberate in small groups across moderate ideological differences, individuals can indeed reconcile their views and construct richer, socially legitimate expressions of public opinion.6 Widely shared “disagreement-curiosity” and openness to persuasion among the studies’ subjects are welcome findings for those looking to turn back the partisan polarization and animus prevalent in the US today. Yet that goal will undoubtedly require sustained commitments across sectors, including research, civil society, philanthropy, and government. The final section of this brief—on political discourse, deception, and distrust—offers some insights into the obstacles to reaching it.
Much of how politicians, analysts, and everyday citizens alike think about voter groups is organized around a grid of demographic categories, with race/ ethnicity (and often gender) at the center. These categories tend to dominate how election results are narrativized, as was certainly the case for the 2016 presidential election.7 On one hand, this is logical. Beyond their general social import, some familiar socio-demographic categories have in recent years become incredibly strong predictors of partisan voting.
Yet there is also reason for researchers, strategists, and funders to practice discretion about how they reinforce these categories. The work of faculty cluster member G. Cristina Mora is instructive. Mora’s research reminds us that demographic categories are neither inevitable nor passive reflections of social reality; rather, they are contingent constructs that hold significant sway in the formation of identities.8 Combining historical and social-science methods, Mora denaturalizes a group category that is today taken for granted in most understandings of US diversity: Hispanic. Mora shows how the panethnic “Hispanic” identity came together through a dynamic interaction among activists and advocates, government officials, and mass media. Despite that these actors were not motivated by a shared purpose, their common promotion and repetition of the term ensured the rise of Hispanic as a social fact.
As Mora further demonstrates, nor even did these actors have a shared definition of “Hispanic.” But in this ambiguity, she explains, was a key to the category’s success. The relatively fluid boundaries around the meaning of “Hispanic” allowed for different actors to invest the category with different interpretations, and thereby facilitated broader Hispanic identification. The patterns of its uptake also influenced what it would come to mean, showing how subjective identities and schemas for classifying people shape one another reciprocally.9
Mora’s analysis of “the making of Hispanics” is germane to the current political moment in light of the rise of a similarly ambiguous and inconsistently defined category: the white working class. This term— together with its initialism, “WWC”—is not entirely new of course, but its prevalence has increased dramatically over the past two years. Since at least March 2016, when Trump settled in as the GOP frontrunner, there has been a burgeoning cottage industry of polling, focus groups, and other studies producing knowledge on this nominal group. Knowing what we do about the role of such knowledge in forming social identities, researchers and funders should ask whether their investigations are actually propping up and intensifying the salience of the category they study. It is unlikely that this is the intent behind the considerable financial resources that have bolstered the “WWC” knowledge industry, but Mora’s account reminds us that their motivations may not matter to the outcome.
Knowledge that reinforces the idea of a discrete “white working class” threatens the building of an inclusive electorate for two reasons. The first is that it is fundamentally anchored in the construct of whiteness. As I have noted elsewhere, analyses of “the white working class” in the context of the 2016 elections vacillate between using income, educational attainment, non-urban residence, and myriad cultural factors as proxies for the amorphous notion of “class.”10 The common denominator is white identification; absent it, there would be no cogent identity of which to speak.
Historical and social research tell us that not all ethnoracial identities are exclusionary, but whiteness tends to be. Faculty cluster member Chris Zepeda-Millán and colleagues demonstrate, for example, that political mobilization around the Latino group identity increases participants’ sense of commonality with other marginalized groups.11 Meanwhile, recent findings from Taeku Lee and colleagues in the Voter Study Group show that strong in-group white identification closely correlates, among other things, with descentbased and religiously and linguistically exclusionary ideas about what it means to be American.12 For this and other reasons, researchers should pay close attention to the impact of changes to the question about race on the 2020 US Census. In its new formulation, the question will introduce a write-in area for white-identifying respondents to further specify their identities (see Figure 1). Unlike other changes proposed for the 2020 Census, this addition will appear without having been extensively field tested, and its purpose and effects are unknown.13
The second threat posed by knowledge that strengthens white working-class identification lies in the identity’s current articulation with far-right ethnonationalism. It is up for debate whether the interests of those represented as “the WWC” are served by mass deportation, corporate tax cuts, slashed social spending, and tariffs. But what is certain is that the WWC knowledge industry—birthed in response to Trump’s rise—feeds the narrative that “the white working class” has a linked fate with Trump, and that “it” is a natural constituency for his nativism. It also potentially offers in “WWC” a social positioning from which fringe actors and ideas can consolidate, and be launched into the mainstream, under the cover of a venerable label. This, in fact, is already happening.14
Forecasting the Electorate
Finally, based on her expertise in category construction, Mora has also had an important voice in scholarly debates on predicting the future composition of the electorate. More specifically, these debates concern how to forecast the share of the country, or the voting-eligible population, that will identify as something other than “white” by a given year. Such forecasts rely on Census Bureau data, and turn largely on models for estimating the number of persons born to one parent who is non-Hispanic white and another who is not.
Mora and her co-author Michael RodríguezMuñiz remind us that such (future) persons’ selfidentification patterns are not inevitable.15 Even more than other types of ethno-racial quantification, forecasts rest on fraught debates about how to assign who to which category.16 Given this, Mora and Rodríguez-Muñiz argue that we must recognize ethno-racial forecasts as themselves political and social interventions. That is, their projections into the future are tools of politics in the present. Among other things, by treating ethno-racial identities as inherited, forecasting inaccurately depicts these identities as given, timeless, and existing outside of experiences of injustice and political consciousness.17 Such a misrepresentation can end up feeding nativism and other exclusionary, “us-and-them” essentialisms. Over-emphasis on identity as inherited can also complicate solidarity efforts crucial to bridging difference for a broader, inclusive civic identity.
- 3. Taeku Lee, Mobilizing Public Opinion: Black Insurgency and Racial Attitudes in the Civil Rights Era, The University of Chicago Press, 2002. See Chapter 3.
- 4. Ibid., p. 90
- 5. Sina Odugbemi and Taeku Lee, “Taking Direct Accountability Seriously,” in Sina Odugbemi and Taeku Lee, eds., Accountability through Public Opinion: From Inertia to Public Action, The World Bank, 2011.
- 6. Kevin M. Esterling, Archon Fung, and Taeku Lee, “How Much Disagreement is Good for Democratic Deliberation?” Political Communication 32(4): 529-551 (2015); and Kevin M. Esterling, Archon Fung, and Taeku Lee, “Modeling Persuasion within Small Groups, with an Application to a Deliberative Field Experiment on U.S. Fiscal Policy,” working paper on file with the author.
- 7. Clark, “What Didn’t Happen?”
- 8. G. Cristina Mora, Making Hispanics: How Activists, Bureaucrats and Media Constructed a New American, The University of Chicago Press, 2014.
- 9. Even so, we should not downplay the potency of certain institutional events—such as the US Census Bureau’s choice to collect data on people as “Hispanic origin”—in cementing a category’s social significance.
- 10. Furthermore, many analyses conflate these, cover up their own conflations, and thereby contribute to reifying and naturalizing the “WWC” they purport only to investigate. Clark, “What Didn’t Happen?”
- 11. Michael Jones-Correa, Sophia J. Wallace, and Chris Zepeda-Millán, “The Impact of Large-Scale Collective Action on Latino Perceptions of Commonality and Competition with African Americans,” Social Science Quarterly 97(2): 458-475 (2016).
- 12. Democracy Fund Voter Study Group, Insights from the 2016 VOTER Survey, at www.voterstudygroup.org/publications/2016- elections.
- 13. By way of contrast, the proposal to add a separate “Middle Eastern or North African” category to the 2020 Census form had been pushed by a broad network of advocacy groups and thoroughly researched over many years. Nonetheless, it was recently scuttled.
- 14. Consider, for example, the man who founded an organization to stage heavily armed protests against the “Islamicization of America” outside mosques in Texas. Profiled in The Washington Post, he applauded the media and public attention to “working-class whites,” which he felt made him part of something both powerful, and importantly, not “fringe.” “It’s not like I’m Joe Blow anymore,” he said. “I have a name, and people would listen.” That name—the one that made him feel like “more than a man with a Facebook account, a passion and a gun,” as the Post reporter put it—is “white working class.” Robert Samuels, “A showdown over Sharia,” The Washington Post, September 22, 2017.
- 15. G. Cristina Mora and Michael Rodríguez-Muñiz, “Latinos, Race, and the American Future: A Response to Richard Alba’s ‘The Likely Persistence of a White Majority’,” New Labor Forum 26(2): 40-46 (2017).
- 16. Faculty cluster member Michael Omi has written a number of widely cited critical analyses of the socio-historical contingency of ethno-racial categories in US official statistics. See for example, Michael Omi, “Racial Identity and the State: The Dilemmas of Classification,” Law and Inequality 15(1): 7-23 (1997).
- 17. In contrast, Taeku Lee proposes that social scientists consider an alternative system of ethno-racial self-identification wherein survey respondents may allocate “points” across a range of identity categories. This would make projects to measure identity better reflect identity’s conceptualization in social theory as multiple and fluid. Taeku Lee, “Between Social Theory and Social Science Practice: Toward a New Approach to the Survey Measurement of ‘Race’,” in Rawi Abdelal, Yoshiko M. Herrera, Alastair Iain Johnston, and Rose McDermott, eds., Measuring Identity: A Guide for Social Scientists, Cambridge University Press, 2009.