Implicit (and explicit) racial attitudes barely changed during Barack Obama’s presidential campaign and early presidency

https://doi.org/10.1016/j.jesp.2009.12.003Get rights and content

Abstract

As a high-status, omnipresent Black exemplar, Barack Obama and his rise to the presidency of the United States may have induced a cultural shift in implicit racial attitudes, much like controlled exposures to positive Black and negative White exemplars have done in the laboratory (Dasgupta & Greenwald, 2001). With a very large, heterogeneous sample collected daily for 2.5 years prior to, during and after the 2008 election season (N = 479,405), we observed very little evidence of systematic change in implicit and explicit racial attitudes overall, within subgroups, or for particular notable dates. Malleability of racial attitudes – implicit or explicit – may be conditional on more features than the mere presence of high-status counter-stereotypic exemplars.

Introduction

Since the election of Barack Obama as 44th president of the United States of America, academics, pundits, and ordinary citizens have pondered a variety of social and political questions about the implications of electing the first Black president. His successful candidacy and election was surely a product, in part, of societal changes in racial attitudes. But, was Obama’s rise itself a mechanism for change? Has the election of Barack Obama affected the citizenry’s attitudes toward African Americans more generally? This article examines whether implicit and explicit racial attitudes changed during the emergence, election and early presidency of Barack Obama.

Social attitudes evolve over time. Perhaps the most dramatic example is the seismic shift in American society of explicit attitudes and stereotypes about African Americans. The “Princeton Trilogy” studies, for example, document a remarkable decline in endorsing stereotypes of African Americans as superstitious and lazy from more than 70% endorsing in 1933 down to 10% or less in 2001 (Madon et al., 2001). The impact of the civil rights movement is imprinted in the pervasive egalitarian beliefs of today’s citizenry, beliefs that have little relation to concepts like “separate but equal” that were popular and daily practice in decades not long past.

Despite this dramatic evidence of change, it is also clear that the stereotypes of yesteryear have not left the public (un)consciousness. Whereas Americans in the 21st century largely decline to endorse stereotypes about African Americans, they have little trouble describing them (Devine, 1989, Devine and Elliot, 1995, Madon et al., 2001). Likewise, the persistence of cultural stereotypes is implicated in the pervasive encoding of implicit associations of African Americans with bad and White Americans with good more than the reverse, even among people who hold egalitarian explicit beliefs and attitudes (Nosek, Smyth et al., 2007). At the same time, evidence suggests that these associations are sensitive to the situation and amenable to change, perhaps particularly in response to high-status exemplars or situational circumstances that counter the prevailing stereotypes (Dasgupta & Greenwald, 2001; Lowery, Hardin, & Sinclair, 2001; Mitchell et al., 2003, Sinclair et al., 2005, Wittenbrink et al., 2001).

Dasgupta and Greenwald (2001) found that exposure to positive Black exemplars such as Martin Luther King, Jr. and negative White exemplars such as Jeffery Dahmer decreased implicit preferences for White Americans compared to African Americans on the Implicit Association Test (IAT; Greenwald et al., 1998, Nosek et al., 2007). As a prominent and highly accessible exemplar during the 2008 presidential campaign and post-election, Barack Obama may serve as a naturalistic version of the Dasgupta and Greenwald (2001) exposure paradigm. Obama is a persistent media presence, ran for and presently holds the most powerful, high-status job on the planet, and contradicts historical racial stereotypes about African Americans.

These factors have produced anecdotal examples and empirical demonstrations of an “Obama effect” on racial attitudes and stereotypes. For example, after finding no implicit preference for White Americans compared to Black Americans in two university samples collected during the campaign season (and relatively weak bias in a third), Plant et al. (2009) concluded that “extensive exposure to Obama resulted in a drop in implicit bias” (p. 961). This research provocatively suggests that a single, prominent exemplar could be sufficient to eliminate a widespread implicit bias that emerges early in childhood (Baron & Banaji, 2006) and exists across every age, gender, racial and national group that has ever been investigated (Nosek, Smyth et al., 2007 – with the exception of African Americans themselves, on average).

As Plant et al. point out, there are reasons to be cautious about their conclusion. The apparent change was obtained in comparatively small, specific samples – undergraduates at University of Wisconsin–Madison and Florida State University. Also, the link between exposure to Obama and change in implicit bias was indirect; they observed a within-sample correlation between accessibility of Obama and implicit race bias. Further, related investigations of an Obama effect in decreasing stereotype threat effects among African Americans reveal mixed evidence for his impact (Aronson et al., 2009, Marx et al., 2009). Finally, other research suggests that the malleability effect due to exposure to high-status exemplars (Dasgupta & Greenwald, 2001) may be weaker than suggested by the existing literature. Joy-Gaba and Nosek (in press) replicated the Dasgupta and Greenwald exposure paradigm with large samples and observed a very weak malleability effect (average d = .08). They also found suggestive evidence that exposure to positive Black exemplars alone may not be sufficient to shift implicit race bias. Making race accessible and even exposing participants to negative White exemplars may also be necessary.

Despite all these cautions, it is difficult to imagine a naturalistic exemplar exposure paradigm with a higher profile than the election of Barack Obama as the first Black US President. As such, we investigated whether implicit and explicit racial attitudes shifted from the period of time prior to the candidacy of Barack Obama to after he was inaugurated President of the United States.

We leveraged a very large sample of visitors to the Project Implicit website (https://implicit.harvard.edu/) who volunteered to complete an IAT and self-report measures of their attitudes toward White versus African Americans. The site enjoyed sizable traffic throughout the data collection (daily median = 396, STD = 730), and participants found the site through a wide variety of means – such as assignment for class or work, recommendation of friend, links from other sites and media coverage. Although the sample is very heterogeneous and had consistent demographic characteristics over time, it is not representative of the US population.

Section snippets

Participants

Participants were visitors to the Project Implicit main demonstration site who selected the “race IAT” between September 28, 2006 and May 11, 2009 for a total of 957 days of data collection. The start date was four months before Obama’s announcement of his candidacy for president, and the end date was almost four months after inauguration and the day before data analysis for this project began.

Main effects

Across the whole sample, participants showed a preference for White Americans over Black Americans both implicitly (IAT D = .34, STD = .45, t[479403] = 524.41, p < .0001, d = .76) and explicitly (M = 0.33, STD = 1.12, t[479403] = 204.13, p < .0001, d = .29). The IAT and self-reported racial attitudes were positively correlated, r = .33, p < .0001.

Variation in racial attitudes over time

Implicit: To test whether the average magnitude of implicit racial attitudes changed over time we regressed session date on IAT score (see Step 1 in Table 1). The results

Discussion

We explored the possibility that the pervasive implicit and explicit preference for White people compared to Black people declined during Barack Obama’s political rise to power and found that, essentially, it did not. The lack of change was minimally moderated by a variety of demographic variables, and we found no evidence that particular time periods or events were associated with shifts in implicit racial attitudes. Finally, explicit preferences for White Americans over Black Americans may

Acknowledgments

We thank Ashby Plant and Trish Devine for comments on a previous version of this article, and Alexandra Waldron and Matthew King for assisting with Lexis–Nexis searches. Nosek is an unpaid officer of Project Implicit, Inc., a nonprofit organization that includes in its mission “To develop and deliver methods for investigating and applying phenomena of implicit social cognition, including especially phenomena of implicit bias based on age, race, gender or other factors.”

References (22)

  • J. Aronson et al.

    The Obama effect: An experimental test

    Journal of Experimental Social Psychology

    (2009)
  • D.M. Marx et al.

    The “Obama effect”: How a salient role model reduces race-based performance differences

    Journal of Experimental Social Psychology

    (2009)
  • Bar-Anan, Y., De Houwer, J., & Nosek, B. A. (2009). The role of contingency memory, intentional processes and number of...
  • Bar-Anan, Y., & Nosek, B. A. (2008). Transitive relations cause indirect association formation between concepts....
  • A.S. Baron et al.

    The development of implicit attitudes: Evidence of race evaluations from ages 6, 10 and adulthood

    Psychological Science

    (2006)
  • N. Dasgupta et al.

    On the malleability of automatic attitudes: Combating automatic prejudice with images of admired and disliked individuals

    Journal of Personality and Social Psychology

    (2001)
  • P.G. Devine

    Stereotypes and prejudice: Their automatic and controlled components

    Journal of Personality and Social Psychology

    (1989)
  • P.G. Devine et al.

    Are racial stereotypes really fading? The Princeton trilogy revisited

    Personality and Social Psychology Bulletin

    (1995)
  • A.G. Greenwald et al.

    Measuring individual differences in implicit cognition: The implicit association test

    Journal of Personality and Social Psychology

    (1998)
  • A.G. Greenwald et al.

    Understanding and using the implicit association test: I. An improved scoring algorithm

    Journal of Personality and Social Psychology

    (2003)
  • A.G. Greenwald et al.

    Race attitude measures predicted vote in the 2008 US Presidential Election

    Analyses of Social Issues and Public Policy

    (2009)
  • Cited by (82)

    • A 3D framework of implicit attitude change

      2023, Trends in Cognitive Sciences
    • Effects of Mass Shootings on Mental Illness Stigma in the United States

      2023, Personality and Social Psychology Bulletin
    View all citing articles on Scopus
    View full text