The Politics of Skin Color: Skin Color as a Politicized Identity for African Americans
Download data and study materials from OSF
University of Michigan
Sample size: 1045
Field period: 11/14/2018-03/04/2019
Heterogeneity in skin color is linked to vast differences in life experiences, both within and across racial groups. For example, darker-skinned Black people have lower incomes, less education, poorer health, and receive harsher criminal sentencing than lighter-skinned Blacks. While political science research often examines racial groups as uniform collectives, skin color may be a meaningful political identity beyond race. Indeed, there is growing evidence of skin color disparities in political leadership and perceptions of Black politicians by both Black and White voters. Building on an interdisciplinary literature on skin color and colorism, as well as my own survey and qualitative data demonstrating the existence of skin color as an identity, I ask the following research questions: What are the conditions under which skin color identity can be activated to influence public opinion and political behavior? How does one’s level of skin color identity—and the inclusion vs. exclusion of actionable remedies to inequalities—influence policy support? This research has implications for scholars, policymakers, and political candidates alike.
(H1a) When viewing a message regarding skin color inequities and remedies, the interaction of skin color and skin color identity will result in those with high levels of skin tone identity—but at opposite ends of the color spectrum—to exhibit the largest differences in support for political remedies to color-based inequities (e.g., dark- vs. light-skinned high-identifiers). Alternatively, (H1b) those with high levels of skin tone identity—regardless of skin tone—may be most willing to acknowledge colorism and thus most supportive of political remedies to color-based inequities relative to low-identifiers, who will be the least supportive (i.e., high- vs. low-identifiers).
This study uses a 2x2 experimental design, plus a (neutral) control condition. The experiment varies whether information is presented about race-based or color-based inequalities, and whether only information is presented or a combination of information and potential remedies are included.
The key dependent variables are related to policies, perceptions of discrimination, and behavioral outcome measures.
Summary of Results
I find that the way skin color and identity interact are most powerful for dependent variables focusing on skin color. Neither treatments nor the two primary identity groups (skin color or race) were associated with movement on traditional policies like race-based affirmative action or the government taking steps to reduce income inequality. Similarly, none of the subgroups had a heightened propensity to take action following reading a story about inequities based on either race or skin color in society. Where large treatment effects were observed was with respect to items that invoked skin tone explicitly. Here, movement was primarily seen amongst those in the color treatment relative to the control and among strongly identified dark-skinned individuals specifically. Dark-skinned high identifiers were consistently the most willing to recognize color-based disparities across a myriad of issues. These effects stood as an especially stark contrast to individuals who only weakly identified with their color, regardless of what their skin tone was. This suggests, then, that skin color identity has the potential to be politically important—but primarily to highly-identified people with dark skin. Strongly identified light- and dark-skinned individuals appear to respond quite differently to the same information about skin color. This suggests that in contrast to studying other identities, even those of an intersectional nature, the tightly interwoven nature of race and skin color identity results in very nuanced relationships between the two.
Yadon, Nicole. 2020. "The Politics of Skin Color." Dissertation Manuscript. University of Michigan.