This is an open access article distributed under the terms of the Creative Commons
Attribution License (I just want to do God's will. And He's allowed me to go up to the mountain. And I've looked over. And I've seen the Promised Land. I may not get there with you. But I want you to know tonight, that we, as a people, will get to the Promised Land!
Dr. Martin Luther King, Jr., 1968
Oh, deep in my heart, I do believe
We shall overcome some day
“We Shall Overcome”
Keep hope alive. Keep hope alive! Keep hope alive! On tomorrow night and beyond, keep hope alive!
Jesse Jackson, 1988
Reflecting upon the words and ideas that have shaped Black political thought for centuries, one detects an unmistakable current of optimism—a heralding of an eventual dawn of Black liberation from a racially stratified society. Often conveyed through the language of Christian symbolism, this sentiment of Black hope wells up in the songs, speeches and sermons that animate Black individuals’ senses of racial group solidarity and political consciousness. Such racial consciousness is known to increase African Americans’ political behavior (Chong & Rogers, 2005; Miller, Gurin, Gurin, & Malanchuk, 1981).
I contend the hope narratives so prevalent within Black dialogic spaces also constitute a unique mobilizer of political action for African Americans. To uncover the unique function of hope in stimulating Black participation requires an integration of theories of appraisal, group-based emotions and group efficacy with literature on the political attitudes and beliefs of the Black body politic. I develop this integrative framework to offer two claims. One, African Americans possess a collective emotional sentiment that differs systematically from White Americans, and this sentiment influences Black individuals’ appraisals of cues of policy opportunity. Two, the emergence of hope in response to cues of opportunity produces shifts in African American appraisals of collective efficacy, resulting in increased motivation to take action to make good on the opportunity. Identifying the race-specific patterns of appraisal and participation among African Americans illustrates the unique capacity of hope to mobilize Black political behavior. It also defies conventional understanding, which indicates threats are generally more effective mobilizers of political action than opportunity.
Recent scholarship identifies how group-based emotions shape the political engagement and decision making of African Americans, revealing distinctions in the translation of emotions such as anger, enthusiasm and shame to participation in costly political actions (Banks, White, & McKenzie, 2018; Burge, 2014; Collins & Block, 2018; Phoenix & Arora, 2018; Thompson, 2017). Building on this work, I present findings from experimental and survey data that reveal racial differences in the mobilizing effects of hope on participation. Hope emerges as a strong and consistent mobilizer of participation among African Americans. This effect is not shared among their White counterparts.
I build a theoretical model of racial emotion regulation by drawing on scholarship connecting emotions to behavior, assessing the role of group identity in shaping emotional responses and perceptions of group efficacy in response to political stimuli, and interrogating the respective effects of threat and opportunity cues on political decision making.
Emotions and Political Behavior
Defined as a feeling of aspiration toward a particular goal, accompanied by positive feelings about the anticipated outcome, hope is a goal-oriented emotion that facilitates risk taking (Bar-Tal, 2011; Breznitz, 1986; Snyder, 1994) and stimulates creativity and problem-solving (Isen, Daubman, & Nowicki, 1987). Jarymowicz and Bar-Tal (2006) show that among people entrenched in violent conflict, those feeling hopeful can imagine a future that differs greatly from their present circumstances.
I contend the linkages between hope, creativity and problem solving carry particular impact for the political behavior of African Americans. This impact comes into clearer view when considering the roles of group identity and social context in shaping individuals’ emotional and behavioral responses to political stimuli. As Halperin, Sharvit, and Gross (2011, p. 85) note, “individual experiences of emotions occur within a certain societal context.” The socio-racial context shaping Black individuals’ perceptions of their collective political efficacy can make them especially responsive to the potential action-stimulating effects of hope.
Group Identity, Efficacy, and Group-Based Emotions
Group-based emotions are centered on an individual’s perceptions of closeness to and solidarity with the group, as well as her perceptions of the group’s collective efficacy (Halperin, Bar-Tal, Nets-Zehngut, & Drori, 2008). For example, people who closely identify with a political party are more likely to respond to political cues with action-stimulating emotions such as anger and enthusiasm, opposed to emotions that can trigger less direct and costly behavior, such as fear (Groenendyk & Banks, 2014; Smith, Seger, & Mackie, 2007). This is due to strong partisans’ perceptions that they are contributing to the good of a team with which they identify. On the one hand, this perception increases their sense of responsibility to the group, an idea consistent with the premise of social identity theory that the stronger one’s ties to the group, the more likely they are to act to advance the group’s interests. On the other hand, these close partisan ties increase individuals’ belief in their capacity to affect change.
The strong racial group ties among African Americans should make group-based emotions especially instrumental in shaping political engagement. Black individuals generally exhibit strong senses of in-group solidarity and racial consciousness, which are based in large part on shared perceptions that the group is marginalized by discriminatory structures in the U.S. (Dawson, 1994; Feagin, 1991; Simien, 2005). This solidarity in turn promotes political action among the group (Chong & Rogers, 2005; Miller, Gurin, Gurin, & Malanchuk, 1981).
Yet while African American generally possess the highest senses of racial solidarity among different racial groups, they also tend to express the lowest senses of collective political efficacy. Black people generally exhibit lower levels of trust in government, less confidence that politics works fairly, and greater dissatisfaction with the sociopolitical environment relative to Whites (Aberbach & Walker, 1970; Avery, 2006; Nunnally, 2012; Pierce & Carey, 1971). While African Americans express confidence in their political efficacy as individuals at rates commiserate with White Americans, they report significantly lower levels of political efficacy as a collective racial group (Phoenix, 2019). Thus, the group-based emotions generated among African Americans are a product of both strong in-group ties and weak perceptions of collective efficacy.
Past studies offer a framework for understanding how African Americans’ low perceptions of group efficacy should be expected to shape collective action in contexts of threat. van Zomeren, Spears, Fischer, and Leach (2004) summarize work by Folger (1987) and Tyler, Boeckmann, Smith, and Huo (1997) as indicating that when aggrieved groups perceive their marginalized status as unlikely to improve, they respond not with an action-mobilizing anger, but rather a debilitating resentment. But as Cohen-Chen and van Zomeren (2018) note, much less attention has been devoted to understanding how emotions interact with efficacy to shape collective action for groups encountering a prospective opportunity rather than a threat. Helping to fill this gap, they find that strong perceptions of group efficacy motivate increased action only in contexts in which hope for the desired outcome is high. They posit thusly that feeling hopeful is a “prerequisite for the motivational power of group efficacy beliefs in the context of collective action” (Cohen-Chen & van Zomeren, 2018, p. 56).
This perspective views the emotion state of hope and one’s sense of efficacy as distinct additive factors that can spur collective action when present together. This view is consistent with the appraisal theory of emotions, which asserts that individuals’ emotional responses to phenomena are rooted in their cognitive evaluations regarding how that phenomenon will advance or hinder their progress toward their goals (Lerner & Keltner, 2001; Scherer, Schorr, & Johnstone, 2001). These appraisals are influenced by individuals’ assessments of their group’s relative standing within the space. For instance, group members’ perceptions of their group’s collective disadvantage and the procedural fairness of governing institutions influence whether their emotional responses to threat cues will be anger, anxiety or resentment (van Zomeren et al., 2004; Weiss, Suckow, & Cropanzano, 1999). On the other hand, a group’s appraisal of whether a positive prospective change is both relevant to its goals and plausibly attainable determines whether they will generate an emotional response of hope (Breznitz, 1986; Thomas, McGarty, & Mavor, 2009).
I offer here an alternate account of hope and efficacy, viewing them not as additive but rather interactive. Whereas appraisal theory suggests the directional relationship flows from cognitive evaluations to emotional responses, Halperin et al. (2011) offer a multidirectional relationship wherein individuals’ broader emotional sentiments—their stable dispositions toward specific objects and symbols—influence their cognitive appraisals of phenomena.
A framework indicating that appraisals of efficacy precede the emergence of emotions suggests African Americans’ diminished group efficacy would inhibit the emergence of hope. Yet studies of Black personal and political life indicate that despite possession of a pervasive and engrained sense of sociopolitical inefficacy, African Americans nonetheless maintain optimistic sentiments (Utsey, Giesbrecht, Hook, & Stanard, 2008; Utsey, Hook, Fischer, & Belvet, 2008). This hopeful sentiment is rooted in part in the deep running Christian religious affiliations permeating the group (Mattis, Fontenot, & Hatcher-Kay, 2003; Mattis, Fontenot, Hatcher-Kay, Grayman, & Beale, 2004), as well as the affirming intragroup social ties that increase the salience of the group’s racial identity (Miller, 1999).
That this sentiment is deeply rooted in Black collective life in spite of the sense of group inefficacy suggests that such inefficacy need not preclude the emotion state of hope from playing a pivotal role in Black collective action. In fact, when attached to a concrete prospective opportunity, this sentiment can disrupt African Americans’ general perceptions of their sociopolitical environment. Signaling a departure from the modal state of affairs in which Black interests are typically subjugated, the positive prospective change represents an opportunity for Black collective action to be freed from the usual constraints emanating from a racially stratified system.
The emergence of hope then represents the removal of a critical barrier to Black collective action—the perception that such action will amount to little within the extant environment. Consistent with the claim by Frijda, Kuipers, and ter Schure (1989) that hope diminishes the role of efficacy in one’s appraisal, I contend the emergence of hope among African Americans represents a distancing of Black individuals from their usual senses of collective inefficacy. This distancing results from the shift in Black people’s appraisal of the environment. Because African Americans’ collective inefficacy is rooted in the perceived constraints of the environment rather than perceived limits on their ability, this shift can spur increased activity by fueling Black people’s belief that in this particular instance, they possess greater capacity to bring about the prospective change.
The potential carried by hope to disrupt the usual resignation regarding the sociopolitical environment should be uniquely felt by African Americans. Despite the increasing salience of White racial identity (Jardina, 2019) and a substantial proportion of the group reporting that anti-White bias is of greater significance than anti-Back bias (Norton & Sommers, 2011), White Americans still by and large perceive the sociopolitical environment as fairer and more responsive than their Black counterparts (Avery, 2006; Hurwitz & Peffley, 2005; Wilkes, 2015). Thus, hope emerging among White people in response to a positive prospective change should not fundamentally alter their appraisals of the environment. Accordingly, hope should not be expected to change White people’s perceptions of their group efficacy. Hope removes a potential impediment to Black collective action that does not exist to the same extent for White Americans. Hence, the potential effect of hope on White collective action is muted relative to its effect among Black people.
In response to cues of opportunity that can animate hope, African Americans’ distinct emotional sentiment shapes their subsequent appraisal of the cue, regulating their emotional and subsequent behavioral response to the cue in a racially distinct manner. I illustrate this model of racial emotion regulation in Figure 1 below.
Figure 1
Racial Emotion Regulation Model—Illustrating the effect of hope on group efficacy and behavior.
Cues of Policy Threat, Opportunity and Political Behavior
My claim regarding the action mobilizing force of hope adds a new dimension to scholarship on the respective motivating effects of threat and opportunity on political behavior. Studies by Brader (2005), Campbell (2003), Miller and Krosnick (2004), and Gadarian and Albertson (2014), reveal that people pay greater attention to cues of policy threat, and increase their participation in costly political actions to avert the threat.
I contend that as a consequence of their unique emotional sentiment forged as a racially conscious and socially marginalized group, African Americans exhibit unique responses to opportunity cues, uncovering a pathway for such cues to be distinctly mobilizing for the group. This is because such cues can interrupt the group’s collective disposition of inefficacy, generating a hope that motivates increased political behavior through causing Black individuals to reappraise their collective efficacy within the environment.
Accordingly, I expect hope engendered among African Americans to prime heightened senses of the group’s political efficacy. This expectation runs counter to extant literature, which links efficacy to anger rather than hope (Huddy, Feldman, & Cassese, 2007; Lerner & Keltner, 2001; Valentino, Gregorowicz, & Groenendyk, 2009). Informed by the model of racial emotion regulation, I conducted two studies to test a set of hypotheses regarding the racially distinct effects of hope on political participation and perceptions of efficacy among Black and White Americans.
Hypotheses
I expect to find that expressed hope carries a stronger mobilizing effect for Black than for White participants (H1). Accordingly, opportunity cues will elicit greater participation from African Americans relative to Whites (H2). Finally, I expect to find that hope engendered among Black Americans will prime greater perceptions of group efficacy (H3a) and a positive appraisal of the responsiveness of the policy environment (H3b). In contrast, reported hope will exhibit null association with White participants’ perceptions of group efficacy (H4a) and their appraisals of the policy environment (H4b).
Study 1
Method
Procedure
A survey experiment conducted in the Detroit metro area from May 2013 through May 2014 investigated racial differences in participants’ emotional and participatory responses to messages signaling policy opportunity. The experiment addressed a local area issue that carried political relevance for participants—privatization of Detroit’s water board. Proponents argued this move would lead to safer, more efficient and cost effective service, while opponents argued the move would cost local jobs and lessen accountability over residents’ water service. The pro arguments were mirrored in the opportunity condition to which participants were randomly assigned.
The treatment design provides high ecological validity by approximating how participants would be exposed to such messaging in real-world contexts. Treatments take the form of political flyers from a local political advocacy organization believed to be authentic by participants. Participants in the opportunity condition view images of contented people with the headline: Want safer water and lower rates? What you can do. This flyer contains arguments in favor of the privatization, emphasizing the effects of cutting wasteful spending, improving water quality standards, and creating the smallest increases to services in a decade. The flyer also invites the participant to contact the state legislature to support the proposed privatization.
Participants in the control condition view a flyer purportedly from an independent research group. This flyer contains no images of people, with the headline: Changes coming to your water service. What can you do? This flyer contains arguments from both proponents and opponents of the privatization. It then asks participants to take action on the issue by seeking more information about it and contacting their state representative (see Figure A1 in the Supplementary Materials for control and opportunity treatments).i
This flyer departs from the typical control condition, as rather than address an irrelevant issue, it presents the same issue absent emotion priming language. Thus, participants' prior awareness of and attitudes about the DWSD issue are as relevant to their responses in the control condition as in the treatment conditions.
Sample
A convenience sample of 287 adult (148 Black, 139 White) residents of the Detroit metro area were recruited for participation in the study via invitations to mailing lists of registered voters, advertising in local newspapers, and the researcher entering local area churches, universities and workplaces with invitations to take the survey. The average age of the sample was 32 years (SD = 15.65). Sixty-three percent of the participants were women, and 36% possessed college degrees. Consistent with racially segregated residential patterns, 74% of White participants resided in majority White zip codes, and 71% of Black respondents resided in majority Black zip codes.
Black and White participants varied as expected in perceptions of the sociopolitical environment. A pre-treatment question asks participants: how fairly would you say American society has dealt with people from your background? Consistent with my claim that African Americans generally perceive racially imposed bounds on their group, 64% of Black participants answered either “not at all fairly” or “a little fairly.” In contrast, only 22% of White participants selected either of these responses.
Measures
Manipulation check
The first post-treatment question is a five category measure gauging participants' opinion on the DWSD privatization. This question measures how effectively the treatments affected participants' views of the DWSD privatization, across variations in prior knowledge of the issue and attention paid to the flyer. Among Black participants, mean approval for privatization hovers right around the midpoint in the control condition (M = 0.48, SD = 0.23) and rises 13 points for those in the opportunity condition (M = 0.61, SD = 0.27). A t-test indicates this difference is significant (p = .03). White participants in the control condition sit just beyond the midpoint at (M = 0.54, SD = 0.21). White approval of DWSD privatization rises just seven points from the control to the opportunity condition (M = 0.61, SD = 0.18). This difference reaches one-tailed significance (p = .08).
Near the end of the instrument, participants answered multiple choice questions asking them to correctly identify elements of the treatment flyers they viewed. Just over 70% percent of participants correctly identified the region of Michigan represented by the political group named on the flyer. Just under 71% of participants correctly identified the make-up of the people in the photo on the flyer they viewed. Just over 50% of participants could correctly identify the name of the political group to which the flyer was attributed. These variations indicate participants varied in attention paid to and retention of the flyers. Although this likely imposed limits on the observable main effects, it better approximates how people process and respond to varying political messages in the real world.
Political participation
The first measure of respondents' participation is a direct invitation to sign their name to a letter being sent to the state legislature: Different groups are writing letters signed by state residents that will be sent to the Michigan legislature asking members to either support or oppose the DWSD restructuring. Please indicate whether you would be willing to add your name to either of these letters.
In addition to this direct measure of immediate participation, participants reported their likelihood of discussing the DWSD plan with others, attending a meeting or forum on DWSD, and contacting a public official about DWSD on a scale from 1 (not at all likely) to 5 (extremely likely).
Emotions
Participants' emotional responses to the flyers were measured via self-reports of the extent to which they felt each of the following emotions while viewing the flyer: angry, anxious, concerned, delighted, distressed, enthusiastic, frustrated, hopeful, motivated, optimistic, outraged, relieved and worried. Responses range from 1 (not at all) to 5 (extremely). Confirmatory factor analysis (CFA) categorized the multiple emotion variables into five proposed constructs. Component loadings for the proposed construct of hopeful yielded an adequate fit (see Table A1 in the Supplementary Materials for factor loadings with varimax rotation).
Efficacy
Participants' senses of collective efficacy were measured with a post-test question asking: How likely is it that people from your community are able to influence politicians on issues like this? Given participants' racially segregated residential patterns, I use this question as a proxy for participants' racial group efficacy. I use as a proxy for participants' perceptions of the policy environment’s responsiveness to their demands the question: In your opinion, how likely is it that the DWSD issue will turn out the way you want it to? Both questions use the same 1 to 5 scale.
All variables are rescaled to range in value from 0 to 1 (see Table A2 in the Supplementary Materials for mean values and standard deviations of all dependent variables, across treatment and participant race). Logistic regression and structural equation models were estimated to pinpoint the moderating effects of race on the associations between exposure to the opportunity treatment, reported hope, and participation and efficacy.
Results
Table 1 below displays the results from two ordered logistic regression models testing the effect of participant race on reported hope across the control and opportunity conditions. Black participants and White participants are equally likely to report feeling hopeful in the control condition. In the opportunity condition, African Americans report feeling more hopeful than Whites by a substantively large margin, though the effect falls just short of one-tailed significance.
Table 1
Logit Models Regressing Race on Participants' Reported Hope, Across Control and Opportunity Conditions
| Variable | Con.
|
Opp.
|
||
|---|---|---|---|---|
| B (SE) | p | B (SE) | p | |
| Black | -0.03 (0.53) | .96 | 0.67 (0.42) | .11 |
| N | 175 | 81 | ||
Hope, Race and Participation
To make direct comparisons of the action stimulating effects of hope on Black and White participants, I ran logistic regression models including an interaction between reported hope and participant race. Table 2 displays the main effects of participant race, hope and the interaction term across each form of action.
Table 2
Ordered Logit Models Interacting Subject Race and Reported Hope on Participation Likelihood
| Variable | Sign letter
|
Attend meeting
|
Discuss w/ others
|
Contact officials
|
||||
|---|---|---|---|---|---|---|---|---|
| B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | |
| Black | 0.36 (0.32) | .27 | 0.73 (0.33) | .02 | 0.21 (0.33) | .52 | 0.78 (0.33) | .02 |
| Hope | 0.67 (0.56) | .23 | 0.38 (0.58) | .52 | 0.76 (0.58) | .19 | 0.61 (0.58) | .30 |
| Black*Hope | 0.79 (0.91) | .38 | 1.12 (0.94) | .23 | 2.20 (1.18) | .06 | 1.48 (0.97) | .13 |
| N | 321 | 317 | 317 | 317 | ||||
The coefficient on the interaction term is positive across all four actions, but only reaches one-tailed significance for the action of discussing the restructuring with others. Calculating the predicted probabilities from these interactions, the marginal effects of hope on participation are substantially higher for Black than White participants—about 20 percentage points larger in the domain of signing the letter, 25 points larger for discussing the restructuring with others, 15 points larger for attending a meeting on DWSD, and about 40 points larger for contacting local officials.
The strong mobilizing effect of hope on Black participation portends that African American participants would be uniquely responsive to the opportunity treatment. Table 3 presents the effects of participant race on likelihoods of taking action on the DWSD restructuring across the control and opportunity conditions. In the control condition, Black participants are no more likely than White participants to take the immediate action of signing the letter to the state legislature. Yet Black participants report significantly greater likelihoods (at the .10 alpha level) of attending a meeting on the DWSD restructuring, discussing the restructuring with others, and contacting public officials regarding the issue compared to White participants.
Table 3
Logit Models Regressing Race on Participant's Likelihood of Taking Action on DWSD, Across Control and Opportunity Conditions
| Variable | Sign letter
|
Attend meeting
|
Discuss with others
|
Contact officials
|
||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Con.
|
Opp.
|
Con.
|
Opp.
|
Con.
|
Opp.
|
Con.
|
Opp.
|
|||||||||
| B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | |
| Black | 0.63 (0.58) | .30 | 1.23 (0.47) | .01 | 1.11 (0.52) | .03 | 0.90 (0.44) | .04 | 0.85 (0.50) | .09 | 0.76 (0.40) | .06 | 1.11 (0.52) | .03 | 0.52 (0.43) | .23 |
| N | 59 | 91 | 59 | 89 | 59 | 88 | 59 | 89 | ||||||||
Exposure to the policy opportunity treatment causes Black participants to be substantially more likely than their White counterparts to take the immediate action of signing the letter. Additionally, Black participants report a higher likelihood of attending a meeting addressing the DWSD restructuring and discussing the issue with others.
These results provide only partial support for Hypothesis 2, as Black participants were generally more participatory regardless of their assignment to the control or opportunity condition. To clarify the relationship between exposure to the policy opportunity frame, expressed hope and participation, I constructed structural models for Black and White participants separately. The models predict the effect of assignment to the opportunity treatment on participants’ participation likelihood, mediated by their reported hope.
Neither subsample reaches the recommended sample size of 200; however the ratios of participants to free parameters exceed the recommended ratio of 5:1 (Bentler & Chou, 1987). CFI and RMSEA indicate goodness of fit of the models for both racial groups (CFI = 1.00, RMSEA = 0.00). Yet chi-square indicates good fit for White, χ2(3, N = 129) = 6.50, p = .09, but not Black participants, χ2(3, N = 116) = 11.96, p = .01. Table 4 below displays the structural model parameters for the immediate action of signing the letter.
Table 4
Structural Parameter Models – Total and Indirect Effects of Opportunity Treatment and Hope on Letter Signing, Across Subject Race
| Variable | Black
|
White
|
||
|---|---|---|---|---|
| B (SE) | p | B (SE) | p | |
| Hope | ||||
| Total | 0.37 (0.19) | .05 | 0.20 (0.15) | .18 |
| Opportunity | ||||
| Total | -0.07 (0.10) | .53 | -0.12 (0.09) | .20 |
| Indirect | 0.05 (0.03) | .10 | 0.02 (0.02) | .29 |
| Direct | -0.12 (0.11) | .26 | -0.14 (0.09) | .14 |
Reported hope has a positive and significant effect on Black participation, but exerts a null effect on White likelihood of signing the letter. The opportunity treatment itself exerts null direct effects on participation for both Black and White participants. Exposure to this treatment yields small, significant increases in reported hopefulness among both Black (β = 0.15, p = .005) and White participants (β = 0.10, p = .08).
Thus, while the opportunity frame effectively engenders hope among both Black and White participants, only among Black participants is reported hope positively associated with taking action on the DWSD restructuring. The opportunity frame provides a clearer pathway to collective action for African Americans than their White counterparts (path diagrams displayed in Figure A2 in the Supplementary Materials).
Table 5 displays the structural parameters for the three prospective actions. The same patterns largely hold. Whereas reported hope exhibits a strong and substantially large positive effect on Black participants' intent to discuss the restructuring with others, it produces only null effects on White participants' intent to discuss the issue. Finally, reported hope elicits a strong and positive effect on Black participants’ likelihood of contacting elected officials, while once again exhibiting a null effect on White participants’ likelihood of contacting.
Table 5
Structural Parameter Models – Total and Indirect Effects of Opportunity Treatment and Hope on Prospective DWSD Action Likelihood, Across Subject Race
| Variable | Attend meeting
|
Discuss with others
|
Contact officials
|
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Black
|
White
|
Black
|
White
|
Black
|
White
|
|||||||
| B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | B (SE) | p | |
| Hope | ||||||||||||
| Total | 0.14 (0.11) | .21 | 0.02 (0.07) | .75 | 0.46 (0.12) | < .001 | -0.03 (0.08) | .71 | 0.24 (0.11) | < .001 | -0.01 (0.06) | .85 |
| Opp. | ||||||||||||
| Total | -0.05 (0.06) | .39 | -0.02 (0.04) | .58 | -0.09 (0.07) | .20 | -0.04 (0.05) | .44 | -0.13 (0.02) | .04 | -0.02 (0.04) | .85 |
| Indirect | 0.02 (0.02) | .24 | 0.00 (0.01) | .75 | 0.07 (0.03) | .02 | -0.00 (0.01) | .72 | 0.05 (0.02) | .04 | -0.00 (0.01) | .85 |
| Direct | -0.07 (0.06) | .24 | -0.03 (0.04) | .56 | -0.16 (0.07) | .02 | -0.04 (0.05) | .48 | -0.18 (0.06) | < .001 | -0.02 (0.04) | .58 |
Turning to the direct treatment effects on participation, exposure to the opportunity flyer exhibits significant negative direct effect on African Americans’ likelihoods of discussing DWSD with others and contacting officials. Were it not for its ability to activate an action-inducing hope among Black participants, the positive prospective opportunity frame would appear to inhibit collective action among Black participants. Accordingly, Hypothesis 2 is weakly supported. The structural equation models make clear the uniquely positive effect of hope on Black collective action, which translates to African Americans being mobilized indirectly by policy opportunity cues to a greater degree than their White counterparts. Had the opportunity treatment proven more effective at eliciting hope among Black participants, its effect on Black participation would likely be more apparent.
Hope, Race and Efficacy
I test the final set of hypotheses by examining the associations between reported hope and participants' perceptions of group efficacy on the DWSD issue. I ran ordered logistic regressions including an interaction between participant race and reported hope on the post-test indicators of collective efficacy and appraisal of the policy environment. Table 6 displays the main effects of participant race, reported hope, and an interaction term between the two, on participant perceptions of their community’s influence over the issue (group efficacy), and their belief that they will receive their preferred outcome (responsiveness of the environment).
Table 6
Ordered Logit Models Interacting Subject Race and Reported Hope on Group Efficacy and Environment Appraisal
| Variable | Community influence
|
Preferred outcome
|
||
|---|---|---|---|---|
| B (SE) | p | B (SE) | p | |
| Black | -0.35 (0.37) | .35 | -0.67 (0.42) | .11 |
| Hope | 0.29 (0.70) | .68 | -0.66 (0.72) | .36 |
| Black*Hope | 1.72 (1.23) | .16 | 2.93 (1.35) | .03 |
| N | 317 | 312 | ||
The interaction between participant race and reported hope achieves two-tailed significance in the domain of participants' belief that their preferences on the restructuring will be met. Calculating the marginal effects of hope for each subject group yields an effect of reported hope on perceptions of community influence that is about 27 points higher among Black participants. The effect of reported hope on participants’ sense of efficacy is about 39 points higher among Black participants than among White participants.
Discussion
Study 1 provided support for Hypotheses 1 and 3a through 4b, while offering less consistent support for Hypothesis 2. Exposure to the policy opportunity treatment engendered increased hope among both Black and White participants. But only for Black participants did this emotion state discernibly impact political behavior. Further, among Black but not White participants, expressed hope appeared to prime increased perceptions of group efficacy and a positive appraisal of the responsiveness of the policy environment.
The Study 1 treatment design enhanced the ecological validity of the study. The findings, however, remain limited to a relatively small sample in a particular geographic and temporal context. The null effects for White participants appear to reflect a fundamental distinction in racial groups’ responses to hope cues. Yet some of insignificant treatment and emotion effects on participation among Black participants may be attributable to the study’s reduced power. Study 2 addressed these concerns by examining a large sample in a nationally representative survey data set.
Study 2
Overview and Hypotheses
Study 2 analyzes the pooled American National Election Study (ANES), a large and nationally representative pre- and post-election survey of political attitudes conducted every Presidential election year since 1948. Beginning in 1980, the ANES has asked participants whether the Presidential incumbent and the major party Presidential candidates have made them feel each of the following emotions: angry, afraid, proud and hopeful. I examine racial differences in the propensity of a set of White and Black partisans to express hope toward favored partisans, and compare the effects of reported hopefulness on respondents’ vote participation and reported external efficacy.
Although past work indicates that African Americans expressed little optimism regarding the capacity of the Obama administration to achieve substantive policy gains for the racial ingroup (Hunt & Wilson, 2009), Black people may have nonetheless perceived that a descriptively representative President signified greater collective influence of the racial ingroup in politics (Bobo & Gilliam, 1990). Further, Collins and Block (2018) show that enthusiasm about the election propels Black turnout. Accordingly, the opportunity to elect (and re-elect) a descriptively representative candidate in Obama carried distinct potential to generate feelings of hope among African Americans, which I argue informed perceptions of their efficacy and incentivized voting.
I therefore operationalize the Obama era as a context of distinct political opportunity for African Americans. Accordingly, models are run separately for the years 2008 and 2012, in order to determine whether the presence of Barack Obama on the ballot produced a distinct effect of hopefulness on Black political behavior. I expect that during this era, African Americans will be more hopeful (H5) and more participatory (H6) relative to their White counterparts during this time period. Additionally, I expect that the hope expressed by Black Democrats toward Obama bolsters their reported external efficacy, while producing null effects for White respondents (H7).
I ran a series of logistic regression models to test the hypotheses. Due to the partisan affiliation of both Obama and a majority of Black survey respondents, analyses are limited to Democratic identifiers. This allows for examination of how racial emotion regulation creates distinct responses to common political figures from people sharing a salient political identity.
Method
Sample
Between the years 1980 and 2012, the ANES contains samples of 4,345 self-identified Black respondents and 22,676 self-identified White respondents. The 2008 and 2012 surveys contained oversamples of Black respondents. More than 84% of Black respondents in the sample identify as Democrats.
Measures
Emotion (Hope)
The first dependent variables is a dichotomous measure of respondent answers to the following question: Has [Democratic/Republican Presidential candidate or incumbent]—because of the kind of person he is, or because of something he has done—made you feel hopeful?
Political participation
Respondent participation is measured via a dichotomous self-report of whether or not respondents voted in the Presidential election. The act of voting offers respondents the most direct means to translate their affect for presidential figures to action.
The regression models include as controls the demographic variables of age, gender (coded 1 = woman), and whether respondents live in the South of the U.S. (1 = South). Also included are two variables assessing possession of material resources associated with voting: a five-category household income measure and a four-category measure of education attained. Indicators of respondents’ political and social engagement include a four-category measure of partisan identification strength, an index measure of respondents’ political and social distrust (coded so that higher values indicate greater distrust), and a six-category variable measuring frequency of religious attendance. The internal efficacy variable is a dichotomous measure of respondents’ agreement with the notion: sometimes politics and government seem so complicated that a person like me can’t really understand what’s going on (coded 1 = disagreement). These control variables are commonly used in political participation models (see Brady, Verba, & Schlozman, 1995).
The ANES contains no direct measures of respondents’ senses of racial group efficacy. In its stead, I use the external efficacy index, a summation of respondents’ agreement with two statements regarding the responsiveness of political actors to their input: (1) public officials don't care much what people like me think and (2) people like me don't have any say about what the government does. Responses to both are coded from 1 (strongly agree) to 5 (strongly disagree). This index measures respondents’ general perceptions of their influence within a given political environment.
Finally, a measure of incumbents’ evaluations of the Presidential incumbent is included. Incumbent approval is a four-category measure of respondents’ approval of the incumbent’s overall job performance. Compared to White Democrats, Black Democrats are significantly less approving of Republican incumbents (White M = 0.26, SD = 0.44; Black M = 0.12, SD = 0.32; p < .001) and significantly more approving of Democratic incumbents (White M = 0.80, SD = 0.40; Black M = 0.95, SD = 0.22; p < .001).
Results
Table 7 displays the results of multivariate logistic models regressing race of Democratic respondent on reports of hope toward Democratic Presidential figures.
Table 7
Logit Models Regressing Race on Democrats’ Reported Hope Toward Favored Partisans, Across Pre-Obama and Obama Years
| Variable | Pre-Obama 1980-2004
|
Obama 2008-12
|
||
|---|---|---|---|---|
| B (SE) | p | B (SE) | p | |
| Black | 0.17 (0.09) | .07 | 0.71 (0.13) | < .001 |
| Religious attendance | 0.12 (0.11) | .26 | 0.01 (0.17) | .96 |
| Woman | 0.15 (0.07) | .02 | -0.02 (0.10) | .85 |
| South | -0.23 (0.07) | < .001 | -0.08 (0.11) | .46 |
| Age | 0.31 (0.20) | .13 | -0.51 (0.30) | .09 |
| Household income | -0.23 (0.13) | .09 | 0.62 (0.20) | < .001 |
| Education | 1.13 (0.12) | < .001 | 1.15 (0.20) | < .001 |
| Party ID strength | 1.39 (0.13) | < .001 | 1.09 (0.19) | < .001 |
| Distrust in govt. | 0.03 (0.15) | .83 | 0.09 (0.20) | .64 |
| Incumbent approval | -0.37 (0.07) | < .001 | 1.22 (0.10) | < .001 |
| Constant | -0.44 (0.19) | .02 | -0.38 (0.26) | .14 |
| Pseudo R2 | .04 | .11 | ||
| N | 4920 | 3975 | ||
Results for the years 2008 and 2012 are shown separately in the second column in order to isolate the “Obama” effect on African Americans’ reports of hope. Black Democrats are indeed substantially more likely than their White counterparts to report hope in the Obama era. During the pre-Obama years, Black Democrats are also more likely than their White counterparts to express hope toward Democratic incumbents and candidates, albeit by a much smaller margin substantially. Consistent with Hypothesis 5, African Americans' proclivity to express in-partisan hope is magnified in an era marked by unique political opportunity for the group.
Race, Hope and Participation
Table 8 displays the results from logit models regressing Democrats’ reported hope on their likelihood to vote, including an interaction term between race and reported hope. Comparisons across the two columns reveal that hope plays a unique mobilizing role for African Americans, but only during the Obama era.
Table 8
Logit Models Interacting Race and Hope Toward Favored Partisans on Democrats’ Vote Likelihood, Across Pre-Obama and Obama Years
| Variable | Pre-Obama 1980-2004
|
Obama 2008-12
|
||
|---|---|---|---|---|
| B (SE) | p | B (SE) | p | |
| Black | 0.11 (018) | .54 | 1.58 (0.33) | < .001 |
| Hope toward Democrats | 0.28 (0.09) | < .001 | 0.80 (0.14) | < .001 |
| Black*Hope | 0.04 (0.21) | .83 | 0.87 (0.35) | .01 |
| Religious attendance | 1.03 (0.13) | < .001 | 0.65 (0.17) | < .001 |
| Woman | -0.03 (0.08) | .72 | 0.09 (0.09) | .32 |
| South | -0.42 (0.08) | < .001 | -0.45 (0.10) | < .001 |
| Age | 2.65 (0.23) | < .001 | 2.56 (0.28) | < .001 |
| Household income | 1.55 (0.15) | < .001 | 1.28 (0.18) | < .001 |
| Education | 2.37 (0.15) | < .001 | 2.08 (0.19) | < .001 |
| Party ID strength | 1.33 (0.15) | < .001 | 1.64 (0.17) | < .001 |
| Distrust in govt. | -0.21 (0.17) | .23 | 0.03 (0.18) | .89 |
| Incumbent approval | -0.30 (0.08) | < .001 | -0.06 (0.10) | .57 |
| constant | -2.52 (0.22) | < .001 | -3.02 (0.27) | < .001 |
| Pseudo R2 | .16 | .17 | ||
| N | 4833 | 3793 | ||
Whereas Democrats’ reported hope toward favored partisans is positively associated with turnout during the pre-Obama era, the effect of hope increases substantially during the Obama era. From the years 1980 through 2004, Black Democrats are no more likely than White Democrats to cast a ballot. In the years 2008 and 2012, Black Democrats are significantly more likely than their White co-partisans to vote. Additionally, in the Obama era the interaction term between Democrat race and hope is positive and significant at the .05 alpha level, indicating that hope exerts a stronger influence on voting among Black than White Democrats. These results strongly support Hypothesis 6.
Figure 2 displays predicted probability plots illustrating the differences in the marginal effects of hope on White and Black participants’ vote likelihood, during both the Obama and pre-Obama eras. The probabilities are calculated with all control variables set at their mean values. Prior to 2008, the marginal effect of reported hope on voting is virtually equivalent among the racial groups. But during the years African Americans have the opportunity to vote for a descriptively representative candidate, reported hopefulness carries an effect on Black vote likelihood that is about 13 percentage points greater than the virtually nil effect it exhibits on White turnout.
Figure 2
Predicted probability plots—Marginal effects of hope on vote likelihood, across Democrat race and election years. Includes 95% confidence intervals.
Race, Hope and Efficacy
Table 9 presents results from logit models regressing Democrats’ reported hope on their external efficacy, once again including an interaction term between race and reported hope.
Table 9
Logit Models Interacting Race and Hope Toward Favored Partisans on Democrats’ Reported External Efficacy, Across Pre-Obama and Obama Years
| Variable | Pre-Obama 1980-2004
|
Obama 2008-12
|
||
|---|---|---|---|---|
| B (SE) | p | B (SE) | p | |
| Black | 0.19 (0.18) | .29 | 0.29 (0.54) | .59 |
| Hope toward Democrats | -0.08 (0.08) | .35 | -0.34 (0.25) | .17 |
| Black*Hope | 0.10 (0.21) | .62 | 0.39 (0.55) | .49 |
| Religious attendance | -0.16 (0.11) | .15 | 0.22 (0.22) | .30 |
| Woman | 0.01 (0.07) | .89 | 0.17 (0.13) | .17 |
| South | -0.06 (0.07) | .41 | -0.08 (0.13) | .55 |
| Age | 0.64 (0.21) | < .001 | -0.77 (0.39) | .05 |
| Household income | -0.79 (0.13) | < .001 | -0.75 (0.24) | < .001 |
| Education | -0.99 (0.12) | < .001 | -1.09 (0.24) | < .001 |
| Party ID strength | -0.64 (0.15) | < .001 | -0.56 (0.25) | .02 |
| Distrust in govt. | 2.38 (0.15) | < .001 | 2.47 (0.25) | < .001 |
| Incumbent approval | 0.38 (0.07) | < .001 | 0.28 (0.14) | .04 |
| Constant | 0.32 (0.19) | .10 | 1.60 (0.38) | < .001 |
| Pseudo R2 | .08 | .08 | ||
| N | 4830 | 2035 | ||
Reported hope from both Black and White Democrats exhibits null effects on the external efficacy measure, across both the pre-Obama and Obama eras. As expected, Black Democrats express no more efficacy than White Democrats during the years 1980 through 2004. Yet despite a larger coefficient effect size, African Americans remain no more likely than Whites to report greater efficacy during the Obama era. Finally, the interaction term between Democrat race and hope is positive but exceeded by its standard error term in across both eras. These results do not support Hypothesis 7. Across all designated years, Democrats’ reported hope failed to exhibit empirically discernible positive effects on their external efficacy.
Discussion
Conducted on a national sample, Study 2 yields findings similar to Study 1, indicating that hope produces distinctly mobilizing effects for African Americans. Further, this action-motivating hope was only animated among African Americans in a context of political opportunity—operationalized as the Obama era. This study also demonstrates that the unique translation of hope to action among African Americans is robust to differences of socioeconomic status and social and political engagement.
Yet in contrast to the first study, these analyses did not yield evidence that the hope engendered among African Americans in the Obama era spurred greater efficacy. This set of null results may be a reflection of usage of efficacy measures that do not directly gauge respondents’ senses of their racial group’s collective capacity to exhibit political influence. Alternately, this may suggest that African American Democrats’ positive affect toward Obama did not fundamentally alter their modal skepticism regarding their collective political influence. In this context, hope may have stimulated Black political participation not through increasing perceptions of collective efficacy, but rather by minimizing the effect of those perceptions on African Americans' decision to turnout.
General Discussion
The studies presented here provide strong indication that contexts of opportunity can effectively mobilize collective action among a socially marginalized group. The emotion state of hope emerged as a critical pathway to action for African Americans, while exhibiting null or substantively small effects on the participation of White Americans. Taken together, these studies indicate that when African Americans face a concrete object of hope—represented either by a promising prospective policy change or a descriptively representative political actor—they respond in kind with increased motivation to take up political action.
Yet the studies varied in indicating whether hope mobilizes Black participation through enhancing group members’ sense of efficacy, preventing me from drawing conclusive results about efficacy’s role in mediating the effect of hope on Black participation. Whereas Study 1 revealed strong associations between Black participants’ expressions of hope and their expressions of community efficacy, Study 2 revealed no association between hope induced by Obama and general efficacy. Identifying alternate design decisions that would address this and other lingering issues offers a blueprint for future work to arrive at a more refined understanding of how hope engendered among African Americans can prime greater collective efficacy and mobilize political participation among the group.
Limitations and Future Research
Despite the design of the experimental treatments to include elements such as emotive faces (Banks & Valentino, 2012) and evocative wording about a relevant political issue (Brader, Valentino, & Suhay, 2008; Gadarian, 2010), the treatments exhibited limited effects on participants’ reported emotions. Following the lead of Albertson and Gadarian (2016), modifying the treatments so that the information about the DWSD restructuring was presented in a more stimulating medium, such as a video clip, may have elicited a stronger response from participants. This would more clearly delineate the racial differences in the activation of hope and its translation to behavior.
Both studies relied on self-reported responses to close-ended questions to measure participants’ emotional responses to the respective objects. This operationalization method leaves open the question of whether participants agree on what is meant by “hopeful” or “optimistic.” Further, whereas Study 1 allowed participants to exhibit variations in the degrees to which they felt the reported emotions, Study 2 allowed for only dichotomous yes or no responses. I was thus precluded from examining how variations in either intensity or frequency of hope felt toward the Presidential figures influenced turnout.
Alternate measurements of participants’ emotional responses to the objects include open-ended self-reports, or batteries of close-ended questions asking participants to report intensity and frequency with which they feel the given emotion. Additionally, imposition of a time series element can allow participants to report pre-test emotion levels as a reference point against which to compare their reported emotions in response to the prime, administered at a later time. This design element allows for exploration of the duration of the hope’s effect on African Americans’ assessments of their collective influence within the political environment, and their corresponding behavior.
Whereas Study 2 measured turnout, Study 1 did not measure participants’ participation in costly political behavior. Beyond the relatively cost-free act of signing their name to a letter address to the state legislature, only participants’ declared intent to engage in future action was measured. An alternate design can address this issue by creating concrete opportunities for participants to engage in real behavior, such as inviting them to make a monetary donation in real time.
Finally, neither study contained direct measures of participants’ perceptions of their racial group efficacy or the responsiveness of the political system to their racial group. Due to the entrenched racial residential segregation in Detroit, the community influence question is a suitable proxy for participants’ perceptions of their racial group efficacy. But the ANES relies on measures of external efficacy that utilize the nebulous concept of “people like me” as respondents’ referent group. Inclusion of questions that directly gauge participants’ perceptions of their racial group’s influence over politics can clarify the relationship between hope, group efficacy and behavior.
This is particularly important given the inconsistent trends emergent from the studies. Would Study 2 yield a stronger mediating role for group efficacy in the relationship between expressed hope and turnout if direct measures of collective racial efficacy were available? Or does a focus on the realm of national electoral politics (compared to local non-partisan politics) illustrate the recalcitrance of Black perceptions of collective efficacy to hope inspiring actors?
To summarize, future work can address the limitations of these studies by employing more visually or aurally stimulating emotion primes, offering a more comprehensive set of response options for participants to report the intensity or frequency of emotions felt in response to the treatments, including measures of direct participation in a variety of actions post-test, and providing direct measures of participants’ racial group efficacy, both pre- and post-test. Such design decisions will further illuminate the distinct effects of hope on Black perceptions of efficacy and political behavior. Additionally, such designs can provide insight into whether the effects of hope are immediate and short lived, or remain robust over time.
Despite the limitations of these studies, clear racial differences emerged in how Black and White participants responded to emotionally and behaviorally to policy opportunity contexts. Those differences carry strong generalizability to real-world settings, in which people are continuously exposed to competing messages about relevant political issues and actors, with their prior knowledge and perceptions of those issues and actors shaping their receptivity to those messages.
The studies also highlight the challenge of disentangling the causal effects of emotion and efficacy on behavior. While Study 1 revealed positive associations between hope and reported efficacy that were unique to African Americans, I was unable to demonstrate a positive association between efficacy and participants’ participation. Van Zomeren et al. (2004) propose a multi-pronged model in which group-based efficacy and group-based emotion produce distinct yet interrelated effects on collective action. The aforementioned refinement of instruments to directly measure individuals’ perceptions of their racial group’s political influence—before and after the hope-inducing prime—may prove effective at distinguishing whether hope’s distinct effect on Black participation is predicated on or independent of its effect on Black perceptions of group efficacy.
Usage of more precise measures of efficacy that prime specific group identities such as race, gender or religious affiliation can result in more refined analyses of how perceptions of collective efficacy shape political behavior—both in contexts of threat, which have been examined thoroughly, and in contexts of opportunity, which have been understudied. Additionally, these measures can clarify the mechanisms through which different emotions exert varying effects on political behavior across social group identities. Future work in this area can inform the efforts of political actors and groups seeking to mobilize African Americans, and perhaps other socio-politically marginalized groups more broadly, by highlighting opportunity cues as a viable pathway to increased political engagement.
Conclusion
Following the lead of Groenendyk and Banks (2014), I made use of experimental and survey data in order to identify causal relationships in one study and demonstrate broader applicability of the phenomenon under examination in another study. Both of the studies presented here yielded a strong mobilizing effect of hope on participation that was unique to African Americans.
The findings presented here indicate that the rhetorical appeals to hope long prevalent in Black political discourses—from Rev. Jesse Jackson’s spirited cries to keep hope alive to the Obama campaign’s impassioned declarations of yes we can—are far more than platitudes. These intonations may indeed offer windows into the unique pull of hope in keeping Black participation afloat in racially fraught political climates. Drawing on theories of appraisal, group-based emotion, and emotion regulation, I offer a model for understanding how race regulates emotional responses to political cues. This model can explain why hope elicits such a distinctly mobilizing effect on African American participation. The model also offers a rebuttal to the contention that cues of threat are uniformly more effective at stimulating political behavior than cues of opportunity. For African Americans and perhaps other socially marginalized groups, opportunity cues engendering hope may be sufficient to lift the tide of action.