The Effects of Gain and Loss Frames on Perceptions of Racial Inequality

Abstract

Previous content analytic research has examined the extent to which the media frame racial disadvantage in terms of Black losses and gains and White losses and gains, finding that news reports are by far most likely to frame disadvantage in terms of what Blacks are more likely (than Whites) to lose. This study is an empirical test of the effects of racial gain and loss framing. Results reveal loss frames amplified perceptions that the issue was important and due to systematic, institutional causes. No main effects of race were found, but race did interact with the frame manipulation to influence perceived importance and symbolic racism. Further, regression models showed the influence of perceptions of importance, causal attributions, and symbolic racism in predicting support for two proposed remedies to alleviate the inequality.

KEYTERMS:

• equivalency framing
• policy
• race
• racial inequality

Notes

Notes: Condition means are raw means by group for each listed variable. F statistics are from ANOVA analysis of the effects of Frame (gain or loss) and Race (Black or White) on listed outcome variables. Total N = 113.

#p = .1; *p < .05.

Notes: Values are unstandardized regression coefficients (and standard errors) from hierarchical linear regression. Values are coefficients after entry of each step, as labeled. R² change indicates incremental R² by listed step.

^a Race manipulation coded as 0 = Black, 1 = White.

^b Frame manipulation coded as 0 = Loss, 1 = Gain.

*p < .05; **p < .01.

Notes: Values are unstandardized regression coefficients (and standard errors) from hierarchical linear regression. Values are coefficients after entry of each step, as labeled. R² change indicates incremental R² by listed step.

^a Race manipulation coded as 0 = Black, 1 = White.

^b Frame manipulation coded as 0 = Loss, 1 = Gain.

*p < .05; **p < .01; ***p < .001.

We confirmed the unidimensionality of each novel scale (perceived issue importance, academic ability, and classroom effort) by checking the alpha if item deleted and using principal axis factoring with promax rotation to confirm that there was a single factor with eigenvalue greater than 1 and loadings above .60.