Research
My research employs multiple methodologies including innovative survey experiments and survey analysis of the American National Election Study to test how multiple identities influence political attitudes and behaviors. Please contact me for drafts of working papers.
"Facing a Black Woman: the Irrational Response to Underperforming White Male Incumbents"
Does gender matter in an environment of racialized politics? Existing research has largely treated race and gender as independent identities, and generally ignored their interaction. Using an innovative survey experiment, this paper tests how race, gender, and qualifications together affect Whites’ support for candidates (measured as vote choice and campaign contributions), and how racial priming changes the effect of identity. The results support the double bind theory, whereby Black women are particularly disadvantaged. Informing respondents that crime has risen leads them to reward the incumbent with larger campaign donations when the challenger is a Black woman than when the challenger is a man—either Black or White—indicating gendered racial opposition to Black women’s candidacies. Surprisingly, highly qualified Black male challengers are actually advantaged relative to similarly qualified White male challengers when respondents are informed that crime is on the rise. A second survey experiment indicates that this counterintuitive finding may result from associations of Black men with violence, leading to stereotypic assumptions that Black male candidates are better equipped to handle crime policy. The findings suggest a need to better understand how so-called “racial” issues such as crime, elicit specific gendered effects as well, and vice versa.
"Conditions of Black-Brown Political Coalitions"
In this paper, I leverage random assignment in a survey of 1,000 Latino adults to test the effects of race/ethnicity, out-group ties, and racial threat on the boundaries of Black Latino electoral cooperation. Existing research on Black and Latino political relations is notable for its contrasting findings. Some scholars contend that there is a great deal of political cooperation between the two groups, while others suggest that conflict is often likely. Most of the previous literature on Black-Latino cooperation examines elite behavior, while relatively few scholars have explored whether there is an attitudinal basis for elite behavior among the African-American and Latino mass publics. This provides an important contribution to the literature by specifying under what conditions we are likely to observe political cooperation among Black and Latino Americans. I develop and test a theory that the threat of anti-Latino discrimination broadens Latinos’ conception of the in-group beyond the ethnic group to include other racial minorities. In turn this increases Latino support for Black candidates who challenge White incumbents. I find that this effect is contingent on Black candidates signaling their support of Latino interest groups, and that racial resentment and panethnicity moderate treatment effects.
"The Gendered Bounds of Racial Solidarity"
Using observational mass attitude data and novel survey experimentation, this paper investigates a previously unexplored puzzle of African-American political identity and behavior. Michael Dawson asserted in his groundbreaking research that linked fate is not only a predictor of African-American political behavior, but that it overwhelms any class division within the racial group, leading Black Americans of all classes to support liberal economic policies. However, I find that Black women have markedly lower levels of linked fate than men, but are similarly liberal. This suggests that linked fate may not operate as hypothesized across gender lines. I analyze the oversample data from the 2012 ANES and find that linked fate predicts beliefs about race-targeted programs, discrimination, and race-based access to government for male respondents. However, Black women's racial attitudes and beliefs are unrelated to racial linked fate. Moreover, Black women who perceive greater gender-based discrimination have significantly lower racial group solidarity. I design new survey experiments on a large sample of African-Americans to test how priming respondents with information about political underrepresentation along racial and gendered lines affects expressions of racial group solidarity.
Published Work
The Double-Edged Sword of Disaster Volunteerism: A Study of New Orleans Rebirth Movement Participants. 2012. Journal of Political Science Education, 8(4): 311-332.
(with Caroline Heldman)
We examine the political and personal effects of disaster volunteerism with participants of the New Orleans Rebirth Movement (NORM) using four waves of pre- and post-surveys and qualitative analysis of participant journals. Significant increases are found in internal political efficacy, desire to be active in politics, and value placed on social justice issues, but disaster volunteerism also dramatically increases cynicism and emotional distress. Nearly every NORM participant in the study experienced emotional stress, and 1-in-5 self-medicated as a coping strategy upon one's return. Disaster volunteerism holds the potential to rapidly accelerate social justice consciousness and activism, even among those already inclined to be active, but the cost is high. Further research on this unique and increasingly popular type of community-based learning is needed.
Works in Progress
Papers available upon request.
Axes of Inequality: the Causes of Black-Latino Political Commonality (with Ariela Schachter)
This project investigates the relationship between social inequality and African-Americans’ perceptions of political commonality with Latinos. Previous work has assumed a causal relationship between these variables, but has offered little evidence to support this assumption. On a national sample of 1200 African-Americans, we present a vignette about educational disparities in which we vary the axis of inequality between class (low-income versus high-income students) and race (either Black versus White students, Latino versus White students, or Black and Latino versus White students) to prime respondents to consider group social position when evaluating their political commonality with Latinos. Most African-Americans report feeling quite high levels of political commonality with Latinos, but we find no evidence to support a causal effect of racial or economic inequality on commonality with Latinos. Our null effect is robust to a series of analyses, and our experiment provides convincing evidence that inter-minority political commonality is not constructed on perceptions of shared group social disadvantage. Our results have implications for the study of African-American politics and minority political behavior.
Abortion Attitudes and the American Voter
Abortion remains a salient and divisive issue in U.S. politics and is often considered a litmus test of candidate ideology, but its impact on voting behavior is less certain. This paper leverages a survey experiment in the 2008 ANES Time Series to interrogate the relationship between abortion attitudes and presidential vote choice. Citizen preferences over abortion are contextually dependent with two dimensions: whether the procedure is sought in a “traumatic'' or “elective'' situation. Abortion attitudes across both dimensions have little substantive effect on political behavior among partisan identifiers, but are predictive of vote choice among self-identified independents. Furthermore, while the elective dimension is linked to vote choice in the full sample and among Republicans, the traumatic scale predicts voting for non-partisans. These findings have implications for survey researchers, as well as for current political contests.
“Not Sure”?: The Gender Gap in American Trade Policy Preferences (with Lauren Prather)
This study examines the issue of measurement in the context of international trade policy preferences. We ask whether and how the measurement of trade preferences and individual skill level affect the presence and strength of the ”gender gap.” Scholars have consistently shown that women are more protectionist than men. This discovery is largely due to the inclusion of gender as a control variable, rather than it being explicitly the research subject. Focusing on gender in this analysis, we demonstrate that the gender gap is sensitive to changes in dependent variable measurement and construction. Moreover, our results suggest that proxy variables used to measure skill level may be gender specific. Education appears more predictive of men’s preferences than women’s, while income is more directly related to the preferences of women. These findings confirm that proper measurement is essential to the study of individual attitudes and that not doing so can augment differences that may not exist. Furthermore, this study encourages scholars to develop theories acknowledging the way in which gender conditions survey response and to consider how it should be incorporated into economic models.
The Reaches of Rape: Conflict-Related and Post-War Consequences (with Dara Kay Cohen)
While the consequences of wartime rape are often assumed to be ruinous, there has yet to be a systematic study of whether and how wartime rape affects the peace process and the post-war recovery of states. In this analysis, we explore how wartime rape disrupts the durability of peace, as well as whether wartime rape has discernable associations with a number of important post-conflict economic, social and political outcomes. We find that the strongest effect of wartime rape is on the difficulty of establishing a lasting peace; wars characterized by mass rape are more than 2.5 times more likely to be associated with more rapid peace failures, controlling for relevant features of the conflict. In addition, in the immediate post-conflict period, wartime rape is associated with significant declines in GDP per capita, indicating the economic costs of wartime rape. However, across a range of social and political outcomes, we find no systematic effect of wartime rape. Though wartime rape exacts severe costs in terms of the likelihood of continued fighting and economic losses, we conclude that the relationship between mass rape and many post-war outcomes is neither straightforward nor easily generalizable. The results have important implications for scholars and policymakers concerned with mitigating problems associated with wartime rape.
My research employs multiple methodologies including innovative survey experiments and survey analysis of the American National Election Study to test how multiple identities influence political attitudes and behaviors. Please contact me for drafts of working papers.
"Facing a Black Woman: the Irrational Response to Underperforming White Male Incumbents"
Does gender matter in an environment of racialized politics? Existing research has largely treated race and gender as independent identities, and generally ignored their interaction. Using an innovative survey experiment, this paper tests how race, gender, and qualifications together affect Whites’ support for candidates (measured as vote choice and campaign contributions), and how racial priming changes the effect of identity. The results support the double bind theory, whereby Black women are particularly disadvantaged. Informing respondents that crime has risen leads them to reward the incumbent with larger campaign donations when the challenger is a Black woman than when the challenger is a man—either Black or White—indicating gendered racial opposition to Black women’s candidacies. Surprisingly, highly qualified Black male challengers are actually advantaged relative to similarly qualified White male challengers when respondents are informed that crime is on the rise. A second survey experiment indicates that this counterintuitive finding may result from associations of Black men with violence, leading to stereotypic assumptions that Black male candidates are better equipped to handle crime policy. The findings suggest a need to better understand how so-called “racial” issues such as crime, elicit specific gendered effects as well, and vice versa.
"Conditions of Black-Brown Political Coalitions"
In this paper, I leverage random assignment in a survey of 1,000 Latino adults to test the effects of race/ethnicity, out-group ties, and racial threat on the boundaries of Black Latino electoral cooperation. Existing research on Black and Latino political relations is notable for its contrasting findings. Some scholars contend that there is a great deal of political cooperation between the two groups, while others suggest that conflict is often likely. Most of the previous literature on Black-Latino cooperation examines elite behavior, while relatively few scholars have explored whether there is an attitudinal basis for elite behavior among the African-American and Latino mass publics. This provides an important contribution to the literature by specifying under what conditions we are likely to observe political cooperation among Black and Latino Americans. I develop and test a theory that the threat of anti-Latino discrimination broadens Latinos’ conception of the in-group beyond the ethnic group to include other racial minorities. In turn this increases Latino support for Black candidates who challenge White incumbents. I find that this effect is contingent on Black candidates signaling their support of Latino interest groups, and that racial resentment and panethnicity moderate treatment effects.
"The Gendered Bounds of Racial Solidarity"
Using observational mass attitude data and novel survey experimentation, this paper investigates a previously unexplored puzzle of African-American political identity and behavior. Michael Dawson asserted in his groundbreaking research that linked fate is not only a predictor of African-American political behavior, but that it overwhelms any class division within the racial group, leading Black Americans of all classes to support liberal economic policies. However, I find that Black women have markedly lower levels of linked fate than men, but are similarly liberal. This suggests that linked fate may not operate as hypothesized across gender lines. I analyze the oversample data from the 2012 ANES and find that linked fate predicts beliefs about race-targeted programs, discrimination, and race-based access to government for male respondents. However, Black women's racial attitudes and beliefs are unrelated to racial linked fate. Moreover, Black women who perceive greater gender-based discrimination have significantly lower racial group solidarity. I design new survey experiments on a large sample of African-Americans to test how priming respondents with information about political underrepresentation along racial and gendered lines affects expressions of racial group solidarity.
Published Work
The Double-Edged Sword of Disaster Volunteerism: A Study of New Orleans Rebirth Movement Participants. 2012. Journal of Political Science Education, 8(4): 311-332.
(with Caroline Heldman)
We examine the political and personal effects of disaster volunteerism with participants of the New Orleans Rebirth Movement (NORM) using four waves of pre- and post-surveys and qualitative analysis of participant journals. Significant increases are found in internal political efficacy, desire to be active in politics, and value placed on social justice issues, but disaster volunteerism also dramatically increases cynicism and emotional distress. Nearly every NORM participant in the study experienced emotional stress, and 1-in-5 self-medicated as a coping strategy upon one's return. Disaster volunteerism holds the potential to rapidly accelerate social justice consciousness and activism, even among those already inclined to be active, but the cost is high. Further research on this unique and increasingly popular type of community-based learning is needed.
Works in Progress
Papers available upon request.
Axes of Inequality: the Causes of Black-Latino Political Commonality (with Ariela Schachter)
This project investigates the relationship between social inequality and African-Americans’ perceptions of political commonality with Latinos. Previous work has assumed a causal relationship between these variables, but has offered little evidence to support this assumption. On a national sample of 1200 African-Americans, we present a vignette about educational disparities in which we vary the axis of inequality between class (low-income versus high-income students) and race (either Black versus White students, Latino versus White students, or Black and Latino versus White students) to prime respondents to consider group social position when evaluating their political commonality with Latinos. Most African-Americans report feeling quite high levels of political commonality with Latinos, but we find no evidence to support a causal effect of racial or economic inequality on commonality with Latinos. Our null effect is robust to a series of analyses, and our experiment provides convincing evidence that inter-minority political commonality is not constructed on perceptions of shared group social disadvantage. Our results have implications for the study of African-American politics and minority political behavior.
Abortion Attitudes and the American Voter
Abortion remains a salient and divisive issue in U.S. politics and is often considered a litmus test of candidate ideology, but its impact on voting behavior is less certain. This paper leverages a survey experiment in the 2008 ANES Time Series to interrogate the relationship between abortion attitudes and presidential vote choice. Citizen preferences over abortion are contextually dependent with two dimensions: whether the procedure is sought in a “traumatic'' or “elective'' situation. Abortion attitudes across both dimensions have little substantive effect on political behavior among partisan identifiers, but are predictive of vote choice among self-identified independents. Furthermore, while the elective dimension is linked to vote choice in the full sample and among Republicans, the traumatic scale predicts voting for non-partisans. These findings have implications for survey researchers, as well as for current political contests.
“Not Sure”?: The Gender Gap in American Trade Policy Preferences (with Lauren Prather)
This study examines the issue of measurement in the context of international trade policy preferences. We ask whether and how the measurement of trade preferences and individual skill level affect the presence and strength of the ”gender gap.” Scholars have consistently shown that women are more protectionist than men. This discovery is largely due to the inclusion of gender as a control variable, rather than it being explicitly the research subject. Focusing on gender in this analysis, we demonstrate that the gender gap is sensitive to changes in dependent variable measurement and construction. Moreover, our results suggest that proxy variables used to measure skill level may be gender specific. Education appears more predictive of men’s preferences than women’s, while income is more directly related to the preferences of women. These findings confirm that proper measurement is essential to the study of individual attitudes and that not doing so can augment differences that may not exist. Furthermore, this study encourages scholars to develop theories acknowledging the way in which gender conditions survey response and to consider how it should be incorporated into economic models.
The Reaches of Rape: Conflict-Related and Post-War Consequences (with Dara Kay Cohen)
While the consequences of wartime rape are often assumed to be ruinous, there has yet to be a systematic study of whether and how wartime rape affects the peace process and the post-war recovery of states. In this analysis, we explore how wartime rape disrupts the durability of peace, as well as whether wartime rape has discernable associations with a number of important post-conflict economic, social and political outcomes. We find that the strongest effect of wartime rape is on the difficulty of establishing a lasting peace; wars characterized by mass rape are more than 2.5 times more likely to be associated with more rapid peace failures, controlling for relevant features of the conflict. In addition, in the immediate post-conflict period, wartime rape is associated with significant declines in GDP per capita, indicating the economic costs of wartime rape. However, across a range of social and political outcomes, we find no systematic effect of wartime rape. Though wartime rape exacts severe costs in terms of the likelihood of continued fighting and economic losses, we conclude that the relationship between mass rape and many post-war outcomes is neither straightforward nor easily generalizable. The results have important implications for scholars and policymakers concerned with mitigating problems associated with wartime rape.
