What Counts as Racism? How Conceptual Disagreement Fuels Social and Political Division

Download data and study materials from OSF

Principal investigators:

Eric Groenendyk

University of Memphis

Email: grnendyk@memphis.edu

Homepage: https://sites.google.com/site/ericgroenendyk

Antoine Banks

University of Maryland

Email: abanks12@umd.edu

Homepage: https://gvpt.umd.edu/facultyprofile/banks/antoine


Sample size: 2030

Field period: 06/23/2021-02/04/2022

Abstract
Americans have long been divided over matters of race, and tensions are again escalating. This project examines an understudied contributor to this division: disagreement over what Americans count as racism. Over time, scholarly conceptualizations of racism have evolved, but has the public kept up? When individuals refuse to apply the label of racism, to what degree does it reflect denial of injustice versus resistance to or misunderstanding of what is meant by the term racism? Through an experiment, we distinguish between those who deny racial injustice—making them likely to oppose reform—versus those who merely show reluctance to use the word racism—making them potential allies in a coalition to combat injustice. Participants were asked to read scenarios designed to exemplify various conceptualizations of racism (explicit racism, subtle racism, institutional racism, and colorblind racism). Those randomly assigned to the baseline condition were asked to decide whether or not each scenario constitutes an instance of racism, whether racism is still a significant problem in the United States, and two questions regarding policies to address racism. In the treatment condition, participants were asked the same questions with a substitute for the word racism (either racial insensitivity (individual acts) or racial injustice (societal and institutional scenarios)).
Hypotheses

H1: Across the scenarios written to exemplify the four conceptualizations of racism, we predict greater willingness to answer in the affirmative when the affirmative response option is labeled “racial insensitivity” or “racial injustice” compared to when it is labeled “racism.”

H2: We expect labeling effects to be larger in cases where scenarios are designed to exemplify newer conceptualizations of racism (subtle racism, institutional racism, and colorblind racism) compared to older conceptualizations of racism (explicit racism).

H3: We expect more people to agree that “racial injustice” is still a significant problem in the United States compared to “racism.”

H4: We expect higher levels of support for policies when they are said to be designed to address “racial injustice” or “racial insensitivity” compared to when they are said to be designed to address “racism.”

Experimental Manipulations
This project involves a simple question wording experiment. Participants randomly assigned to the baseline condition were asked to answer a series of questions regarding a) whether various scenarios constitute instances of racism, b) whether racism is still a significant problem in the United States, and c) whether they support or oppose policies designed to address racism. Those randomly assigned to the treatment condition were asked the same questions, but the word racism was substituted for alternative language. When questions pertain to individual acts, racism was substituted with racial insensitivity. When question pertain to society or institutions, racism was substituted with racial injustice.
Outcomes

Conceptualization Measures: Participants read eight scenarios and decided whether or not each one constituted an instance of racism (racial insensitivity/ racial injustice). Each of the four conceptualizations of racism was captured using two scenarios, so responses to these scenarios were combined to create a single measure for each conceptualization, yielding four variables. Each variable was rescaled to run from 0 to 1.

Problem Measure: Participants indicated whether or not they believed racism (racial insensitivity/racial injustices) is still a significant problem in the United States.

Policy Measures: Participants indicated the degree to which they supported or opposed two policies designed to address racism (racial insensitivity/ racial injustice).

Summary of Results

Participants were significantly more likely (p = .004) to indicate that scenarios designed to represent the concept of colorblind racism constituted racial injustice (.378) compared to racism (.323). However, they were not more likely (p = .535) to indicate that scenarios designed to represent the concept of institutional racism constituted racial injustice (.643) compared to racism (.631).

Participants were also significantly more likely (p <.0001) to indicate that the scenarios designed to represent the concept of subtle racism constituted racial insensitivity (.461) compared to racism (.355). However, they were not more likely (p =.720) to indicate that scenarios designed to represent the concept of explicit racism constituted racial insensitivity (.895) compared to racism (.891).

Contrary to expectations, participants appear to have been slightly less likely to acknowledge that racial injustice (.742) is still a significant problem in the United States compared to racism (.775), but the effect is only marginally significant (p=.084). The experimental treatment had no significant effect on either of the two policy measures: racial sensitivity training to address racism (racial insensitivity) among government employees (racism = .669; racially insensitivity =.670, p = .939) or government action to address racism (racial injustice) (racism =.678; racial injustice = .674, p =.786). Note: All tests are two-tailed, and all variables are scaled to run from 0 to 1.