In the aftermath of the United States’ 2020 presidential election, state legislatures have introduced and passed an unprecedented number of restrictive voting bills. While past research has looked at the state-level drivers of restrictive voting legislation, this project explores what factors predict which legislators within states push for these laws. Specifically, I ask whether district-level characteristics predict when lawmakers use bill sponsorship to send messages about their positions beyond those sent by simple roll-call votes. I use theories of geographical threat and racial resentment to predict where sponsorship of these bills is most likely. My results tie these theoretical expectations to observed legislative activity: the whitest state legislative districts in the least-white states were the most likely to be represented by lawmakers who sponsored restrictive bills, as were districts with the most racially resentful white residents. I conclude that, despite lawmakers justifying these restrictive laws by claiming that fraud is a major problem, race and racism are inherently tied to the introduction and passage of these bills. This raises important questions about commitments to multiracial democracy.

Most research studying minority representation concludes that minorities enjoy better representation when they constitute a larger share of a constituency, but only through the partisanship and race/ethnicity of the representative. Other research finds that minorities receive worse representation when they constitute a larger share of a constituency. We argue that minority composition will have an independent effect on representation, but that this effect will differ depending on the representative's partisanship. We apply our theory to Latino composition and state legislative voting on immigration policy and find that Latino composition has no effect on voting among Democratic legislators, who are less likely to vote in a restrictive direction on immigration than Republicans regardless of the Latino composition in their district. However, Republicans are more likely to vote to restrict immigration as Latinos comprise a larger share of their district. Our findings suggest that scholars should consider the moderating effect of legislator partisanship when examining minority composition and representation.

Research has shown that white Americans oppose welfare spending in part due to their racial prejudices. Yet, this conventional wisdom ignores the importance of local geography in determining whether whites are likely to view welfare in racialized terms. This article demonstrates that the effect of prejudice on whites’ welfare preferences depends on the salience of welfare's racialized image in a given geographic context. I present a novel application of the racial threat hypothesis—conceptualizing both prejudice and place as multidimensional—to argue that the racial geography of an area amplifies the effect of traditional racial stereotypes on welfare preferences, whereas economic geography amplifies the effect of symbolically racist attitudes. I test these propositions using geocoded data from the years following the 1996 U.S. welfare policy reforms. My analysis reveals that racial attitudes are more predictive of welfare preferences not simply where the stigmas of welfare are salient due to a large black population or high poverty rate, but more specifically where residential segregation makes black or poor households more visible from the viewpoint of whites. These findings highlight the subjectivity of the white perspective, and call for more scholarship theorizing whites’ agency in seeing racialized issues as threatening to white interests.