Buried hatchets, marked locations: Forgiveness, everyday racial discrimination, and African-American men’s depressive symptomatology

Abstract

Everyday racial discrimination (ERD) is linked to pronounced depressive symptomatology among African-American men. Yet, many African-American men do not experience depressive symptoms following ERD exposure often because they employ positive coping strategies that offset its effects. Granting forgiveness is one coping strategy associated with less depression. However, extant findings about the mental health benefits of forgiveness are somewhat mixed and pay scarce attention to offenses which are fleeting, historically rooted, and committed outside of close personal relationships. Evidence further suggest age-related differences in forgiveness, ERD exposure, and depressive symptoms. We explore the extent to which three strategies of granting forgiveness of ERD—letting go of negative emotion (negative release), embracing positive emotion (positive embrace), or combining both (combined)— are associated with less depressive symptomatology in 674 African-American men (ages 18-79). Building on past findings, we also test whether these forgiveness strategies moderate the ERD-depressive symptoms relationship for men in different age groups (18-25, 26-39, 40). Higher combined and negative release forgiveness were directly related to lower depressive symptoms among 18-25 year-olds. We also detected a less pronounced positive relationship between ERD and depressive symptoms among men reporting high levels of combined (18-25 and 26-39 groups) and negative release (26-39 and 40+ groups) forgiveness. We observed a more pronounced positive ERD-depressive symptoms relationship among 18-25 and 26-39 year-olds reporting lower forgiveness. When faced with frequent ERD, younger African-American men may have the most difficult time burying hatchets without marking their location but experience more positive mental health benefits when they do.

Recent high-profile murders of African-American males (e.g., Trayvon Martin, Michael Brown, Oscar Grant, Eric Garner, Jordan Davis, Tamir Rice, and Freddie Gray) have reinvigorated scientific and public discourse about the persistence of racial discrimination and its grave health consequences. Central to this discourse lie critical questions about what strategies should be collectively and individually mounted to address the detrimental impacts of racial discrimination on the well-being of the African-American community and on the males within it who are already at a unique health disadvantage. A substantial body of research documents the significant decrements in individual physical and mental health status as a consequence of exposure to everyday and structuralized racial discrimination (Borrell, Kiefe, Williams, Diez-Roux, & Gordon-Larsen, 2006; Paradies, 2006; Utsey, Payne, Jackson, & Jones, 2002; Williams & Mohammed, 2009; Williams & Williams-Morris, 2000). African-American men not only report more exposure to racial discrimination than African-American women (Banks, Kohn-Wood, & Spencer, 2006; Borrell et al., 2006; Krieger & Sidney, 1996; Sellers & Shelton, 2003; Sims et al., 2012), but their disproportionate exposure is associated with greater risk for cardiovascular disease (Chae, Lincoln, Adler, & Syme, 2010), hypertension (Krieger & Sidney, 1996), substance abuse disorders (Terrell, Miller, Foster, & Watkins Jr, 2006) premature biological aging (Chae et al., 2016), and depression (Hammond, 2012; Matthews, Hammond, Nuru-Jeter, Cole-Lewis, & Melvin, 2012; Watkins, Hudson, Caldwell, Siefert, & Jackson, 2011). African-American men’s increased risk for depression in the face of racial discrimination is especially worrisome given depression’s association with elevated suicide risk (Joe & Kaplan, 2001; Minkoff, Bergman, Beck, & Beck, 1973) and high comorbidity with chronic disease mortality (Livingston, Leaf, & Rozal, 1994; Rabins, Harvis, & Koven, 1985; Rutledge, Reis, Linke, Greenberg, & Mills, 2006).

Perhaps in response to the more subtle presentation of racial discrimination in contemporary life, most of the current evidence-base focuses on the mental health consequences (e.g., depression) of everyday racial discrimination (ERD). ERD or systematic and institutional attitudes and behaviors (e.g., racial profiling by police), as well as daily slights (e.g., being mistaken for someone who serves others (e.g., maid, bellboy) that occur because of race (Essed, 1991; Harrell, 2000; Sue et al., 2007) are often referred to as “microaggressions.” The higher incidents of racial profiling and police-related killings of African-American males are devastating examples of how ERD disproportionately disrupts the lives of African-American males (Gelman, Fagan, & Kiss, 2007; Krieger, Chen, Waterman, Kiang, & Feldman, 2015; The Leadership Conference on Civil and Human Rights, 2011). Racial profiling incidents are sometimes linked to attire (e.g., hoodies) or instigated by seemingly age-appropriate behavior (e.g., playing loud music), which appear to put younger African-American males at increased risk. This increased risk is affirmed by data demonstrating disproportionately higher number of stop-and-frisks among younger African-American males (Milner, George, & Allison, 2016; Ridgeway, 2007). ERD experiences, like racial profiling, exact unique “wear and tear” on mental health, in part because its exposure is fleeting, persistent, and ambiguous. These stressor characteristics also shape the kinds of collective and individual coping strategies that are employed to protect mental health when ERD presents itself. Although depressive symptomatology is a commonly noted mental health consequence of everyday racial discrimination (ERD) (Banks et al., 2006; Hammond, 2012; Lincoln, Chatters, Taylor, & Jackson, 2007; Noh, Kaspar, & Wickrama, 2007; Schulz et al., 2006), it is evident that not all African-American men become depressed following ERD experiences. Theorists purport that African-American men preferentially employ active and potentially health-damaging (e.g., substance abuse) coping strategies when faced with ERD (Jackson & Knight, 2006). Few studies consider African-American men’s use of emotion-focused coping strategies that are associated with positive psychological resources or with their mental health. Our study addresses this scientific void.

Forgiveness is one emotion-focused coping strategy taking center stage in burgeoning sociopolitical discourses about, and social scientific explorations of, ERD. The positive mental health benefits of forgiveness have been documented across a number of studies (Davis et al., 2015). In the present study we focus on forgiveness because prior evidence indicates that the use of this emotion-focused coping strategy is associated with lower levels of reported depressive symptomatology (Brown, 2003; Maltby, Day, & Barber, 2004; Toussaint, Williams, Musick, & Everson-Rose, 2008b; Tse & Yip, 2009). We recognize that public debate about the virtues of granting forgiveness in the aftermath of recent race-related incidents have been especially charged in that they have tended to elevate, as a more enlightened coping response, the act of burying the hatchet without marking the site or location of transgressions. In the midst of these debates, we take note of evidence suggesting that there are factors that complicate the deployment of forgiveness following ERD. We must consider, for example, that African-American men are situated in a historical moment that some pundits have declared as “post-racial” (i.e., as a time and space in which race has lost its sociopolitical relevance). This post-racial discourse may make it particularly difficult for men to justify or appropriately legitimize their claims of ERD experiences and, consequently, may make it particularly difficult to bury hatchets. At stake is the question of how men cope with ERD in a time period more fraught with broader societal refusal to acknowledge that Black lives matter or that racism continues to operate as a political force.

In this exploration of the role of forgiveness of ERD on the lives of Black men, we attend to evidence that indicates that ERD’s unique depressionogenic effect on African-American men’s mental health may also be informed by masculine role norms that shape the kinds of coping strategies African-American men employ to offset ERD (Matthews et al., 2012). Such norms also discourage letting go of negative emotions and disclosing vulnerability (Hammond, 2012). In particular, extant research demonstrates that African-American men have a more difficult time forgiving ERD when they also endorse masculine role norms that encourage emotion restriction (Hammond, Banks, & Mattis, 2006). There is also evidence that younger males report stronger endorsement of masculinity norms (Levant & Richmond, 2008). Thus, we consider age-related differences in the ERD-depression symptoms relationship given findings that document age-related or life course differences in racial discrimination exposure (Hammond, 2012; Kessler, Mickelson, & Williams, 1999), in forgiveness (Toussaint, Williams, Musick, & Everson-Rose, 2001), and in depression (Kessler, Birnbaum, Bromet, et al., 2010; Kessler, Birnbaum, Shahly, et al., 2010).

The primary research questions motivating our investigation are: 1) After accounting for masculine role norms and ERD frequency, is forgiveness of ERD associated with less depressive symptomatology among African-American men?; 2) Does forgiveness of ERD moderate the association between ERD frequency and depressive symptomatology?; and 3) Are the direct and moderated associations between forgiveness of ERD and depressive symptomatology more pronounced among African-American men in different age groups? In addressing these questions, we first explore definitions, conceptualizations, and correlates of forgiveness. Next, we review literature demonstrating linkages between forgiveness, mental health and depression. Finally, we present our theoretical framework and study hypotheses.

Definitions and Conceptualizations of Forgiveness

Forgiveness has many lay and scientific meanings. Although empirical support for a singular definition of forgiveness does not exist, many researchers concur that forgiveness is situated within a specific interpersonal context and is a process of intra-individual, prosocial change toward a perceived transgressor (Pargament, McCullough, & Thoresen, 2000). Operational definitions of forgiveness vary in the extent to which they emphasize particular affective, cognitive and behavioral dimensions of the construct (McCullough & Worthington Jr, 1994). Within the affective domain, a number of scholars assert that forgiveness entails the retraction of negative emotions (e.g., resentment) and the evolution of a capacity to experience positive affect (e.g., compassion, love) towards an offender (Hargrave & Sells, 1997; McCullough & Worthington Jr, 1994; Sells & Hargrave, 1998). In the cognitive domain, forgiveness is believed to entail a willingness to appreciate the fallibility of the aggressor, the capacity to develop insight into the causes of the rupture in relationship, and the ability to relinquish ruminations about the offense, and ruminations over the desire to enact retribution against the offender (Hargrave & Sells, 1997; McCullough & Witvliet, 2002; McCullough & Worthington Jr, 1994). Behaviorally, forgiveness is conceptualized as the active effort to provide offenders with opportunities to engage in corrective action and or the willingness to extend compassion toward the offender (Hargrave & Sells, 1997; McCullough & Witvliet, 2002; McCullough & Worthington Jr, 1994). It is worth noting here that although some scholars argue for the centrality of reconciliation in the conceptualization of forgiveness (Frise & McMinn, 2010), Kim and Enright (2015) argue persuasively that while reconciliation is central in theological understandings of divine forgiveness (i.e., God’s forgiveness is accompanied by a desire for believers to return to right relationship with Him), reconciliation is not a requirement of forgiveness in the human realm.

In this exploration of forgiveness, it is also important to recognize that forgiveness is distinct from “forgivingness.” While the former is state-specific, the latter term, “forgivingness,” refers to trait characteristic wherein an individual is dispositionally oriented to grant forgiveness in the aftermath of transgressions (Berry, Worthington Jr, O'Connor, Parrott III, & Wade, 2005). We are concerned in this study with state-specific definitions of forgiveness. Consistent with the work of Worthington and Scherer (2004) and with the work of other scholars who endorse multidimensional conceptualizations of forgiveness (Maltby & Day, 2004; Maltby et al., 2004; Maltby, Macaskill, & Day, 2001; Worthington Jr, 1998), we conceive of forgiveness as an emotion-focused coping process that involves 1) the generation of positive thoughts, and 2) the elimination of negative thoughts, feelings, or behaviors towards transgressors (McCullough & Witvliet, 2002). Our interest here is in exploring empirically the extent to which the generation of positive thoughts and feelings or elimination of negative thoughts and feelings are central to mitigating depressive symptomatology among African-American men in the aftermath of ERD.

Religiosity and Forgiveness

Although forgiveness is often conceptualized as an individual level manifestation of personal choice and character, scholarship in philosophy, cultural studies, sociology, and theology point to links between forgiveness and religiosity. Indeed, divine forgiveness and self and other forgiveness are central in Christian theology (Kim & Enright, 2014; McCullough, Bono, & Root, 2005; Tsang, McCullough, & Hoyt, 2005). Importantly, in the U.S., where Christianity has been cultivated in a context of oppression, marginalization, and brutal, dehumanizing racial and gendered oppression, the theologies that are central to African-American faith traditions espouse complex and sometimes opposing positions on forgiveness. Some African-American theologians assert that forgiveness is the ultimate manifestation of authentic faith and is necessary for ensuring racial reconciliation and movements towards a just society (see Roberts, 2005 for example). Other African-American theologians posit that in the context of oppression, virtues such as love, hope and justice must take precedence over forgiveness (see Cone, 1977, for example). From this latter perspective, the rush to forgiveness can obscure or invalidate the righteous anger of victims, serve to mollify and prematurely absolve those who engage in oppressive action, and undermine any serious efforts to confront and dismantle unjust power arrangements and the structures that support them. The tension that is ever present in African-American faith life, then, is a tension between the divine mandate to forgive, and a “prophetic” religious tradition that seeks to balance the need for forgiveness (particularly the forgiveness of racism and other injustices) with a quest to transform power arrangements in a way that eradicates oppression, restores dignity and humanity, and ensures liberation and justice.

The complex relation between forgiveness and faith among believers generally, and among African-Americans in particular, has begun to lend credence to the complexity of the interplay between faith and forgiveness. Empirical research on the cultural and cultural ideological contributors to forgiveness supports the link between religious commitment (i.e., more frequent religious service attendance) and the likelihood of forgiving transgressions (Lutjen, Silton, & Flannelly, 2012; Wuthnow, 2000). In a nationally representative study of 1,379 U.S. adults participating in bible study groups, Wuthnow (2000) found most (61%) attributed their capacity to forgive with being immersed in these religious networks. A more recent investigation in a random sample of more than 1,600 U.S. adults (Lutjen et al., 2012) found higher forgiveness of oneself among individuals reporting more prayer, church attendance, and religiosity. Importantly, research suggests that relative to other groups, African-American adults report a significantly greater likelihood of engaging in forgiving behavior (Wuthnow, 2000). Further, despite empirical findings that African-American men score lower on indices of religiosity than their female counterparts, empirical work demonstrates that religiosity is a robust predictor of forgiveness among African-American men (Hammond et al., 2006). In particular, research demonstrates a significant positive relationship between subjective religiosity (i.e., the importance of religion in one’s life) and forgiveness of ERD among African-American men (Hammond et al., 2006; Sue et al., 2007). Taken together, the extant work on the religiosity–forgiveness link suggests that that the motivation to forgive and the likelihood of forgiving may be amplified among men who have embraced faith life as a meaningful part of their personal identity, and for those who are exposed to the lessons about forgiveness and compassion drawn from formal religious life (e.g., sacred texts, sermons). These indices of faith may provide individuals with compelling reasons to forgive and may provide men with effective methods for meaning-making and coping in the aftermath of transgressions (Pargament & Rye, 1998; Rye, 2005; Rye et al., 2000; Wuthnow, 2000).

Correlates of Forgiveness

In this study, we account for the extent to which forgiveness offsets ERD’s impact on depressive symptomatology after the effects of religiosity are considered. Given our focus on associations between forgiveness and mental health in the lives of men, we would be remiss if we did not attend to scholarship that elucidates linkages between masculine role norms, coping and forgiveness. We note again that we understand forgiveness as an emotion-focused coping strategy that informs both the meanings that individuals ascribe to race-related transgressions, and the way that individuals respond to such transgressions. Accordingly, we understand that any aspects of the self (e.g., personality style, beliefs, gender role identity) that influence emotions, coping preferences, or the meaning-making process will likely affect peoples’ capacity to forgive. Here it is important to acknowledge scholarship that suggests that forgiveness might also be understood as a response tendency rooted in dispositional characteristics (Brose, Rye, Lutz-Zois, & Ross, 2005; McCullough & Hoyt, 2002; McCullough & Worthington, 1999; Walker & Gorsuch, 2002), as well as empirical evidence that affirms that neuroticism (i.e., the tendency to easily experience unpleasant emotions), is most consistently and negatively associated with the likelihood to forgive (Ashton, Paunonen, Helmes, & Jackson, 1998; Maltby et al., 2001; Quartana, Schmaus, & Zakowski, 2005). We note, in particular, that in a study among college students attending religious and non-religious universities, Walker and Gorsuch (2002) observed lower forgiveness scores among individuals reporting higher neuroticism. Similarly, among 275 college students, Brose et al. (2005) found that higher levels of neuroticism were associated with willingness to grant forgiveness by releasing negative and generating positive affect and cognitions. This consistent association likely exists because neurotic individuals have a heightened vulnerability to negative emotion and perhaps a greater tendency to harbor anger (McCullough & Hoyt, 2002). In light of these findings, in this work we advance a model that accounts for neuroticism as a potentially important correlate of forgiveness of ERD.

While neuroticism is a personality feature that is potentially important in understanding forgiveness, we also take note of empirical findings that implicate masculinity norms as important forces in the ERD-forgiveness dynamic among men. Indeed, recent theoretical frameworks and empirical findings suggest that the way that masculinity norms play a role in the primary appraisal of racial discrimination (Hammond, Fleming, & Villa-Torres, 2016; Matthews et al., 2012) and that these norms are generally linked to diminished forgiveness (Hammond et al., 2006; Walker & Doverspike, 2001). In particular, masculinity norms encouraging emotion restriction were associated with lower forgiveness of racial discrimination scores particularly among younger African-American men and facilitators of forgiveness among older men (Hammond et al., 2006). Research on men’s mental health also suggests that masculine role norms may exacerbate the impact of ERD on depressive symptomatology among African-American men (Hammond, 2012). Specifically, in a sample of 674 African-American men recruited primarily from U.S. barbershops, Hammond (2012) found that the positive association between ERD and depressive symptoms was strongest among men who held tightly to masculinity norms encouraging emotion restriction or suppression. Hammond et al. (2006) argue that African-American men may who rigidly embrace masculinity norms have a difficult time forgiving ERD because doing so depends on individuals’ willingness to both make emotional disclosures about stressors or transgressions (Maltby et al., 2004). Men who strongly endorse traditional masculinity norms may perceive these kinds of disclosures and the displays of empathy associated with forgiveness as violations of the masculine code. These findings are useful in that they highlight the role of masculinity norms in shaping forgiveness and mental health outcomes among men. We honor extant findings by accounting for the role of masculinity norms in the forgiveness-depressive symptoms dynamic.

Forgiveness, Mental Health, & Depression

Epidemiological and psychology research identifies depression as a particularly pernicious mental health outcome among African American men (Watkins, Green, Rivers, & Rowell, 2006). Studies estimate the prevalence of depression at 5-10% among African-American men, and data gathered over a ten year period (1991-2001) demonstrate a more than two-fold increase in the prevalence of depression (from 2.48% to 6.48%) among African-American men (Ward & Mengesha, 2013). These prevalence data are especially concerning in light of recent studies that demonstrate that suicide (which is significantly associated with depression) has become a leading cause of death among African-American men over the past 2 decades (Walker, Salami, Carter, & Flowers, 2014). Importantly, a number of studies have pointed to a more pronounced risk for depression among African-American men exposed to racial discrimination (Walker et al., 2014; Watkins, 2012). Racial discrimination demeans and dehumanizes individuals, erodes hope, increases anxiety, contributes to cardiovascular disease and poor health outcomes, and disrupts individuals’ access to the opportunity structure and to critical resources including employment and health care (Clark, Anderson, Clark, & Williams, 1999; Watkins, 2012). For African-American men, racial discrimination may also serve as a challenge or threat to gender roles including the role of provider or breadwinner (Hammond et al., 2016). In community contexts, gendered forms of racism (e.g., racial profiling of men) can increase men’s levels of stress, carry substantial financial costs associated with lost wages and the need to secure legal protection against the assaults of transgressors, and cost men their lives. Watkins (2012) and Hammond et al. (2016) propose that the stresses associated with racism accumulate over the life-course, leaving men vulnerable to a range of poor health and mental health outcomes including depression as they age. What remains unclear is which factors moderate the relationship between racism and depressive symptomatology among men. In this study we investigate the moderating role of forgiveness in the ERD-depression relationship.

The literature connecting forgiveness and mental health including depression is steadily evolving, but generally establishes a linkage between a greater capacity to forgive and better health outcomes (Davis et al., 2015; Ellison & Levin, 1998; Sanchez-Gonzalez, May, Koutnik, & Fincham, 2015; Worthington Jr, Witvliet, Pietrini, & Miller, 2007). For example, forgiveness is related to better self-rated health (Berry & Worthington Jr, 2001) and existential well-being (Rye et al., 2001). Adults with a greater tendency to forgive also report fewer depressive symptoms (Brown, 2003; Hirsch, Webb, & Jeglic, 2011b; Krause & Ellison, 2003; Lawler et al., 2003) and are less likely to meet diagnostic criteria for a Major Depressive Episode (Toussaint, Williams, Musick, & Everson-Rose, 2008a). Researchers have also found a negative correlation between older adults’ willingness to forgive others of general interpersonal transgressions and reported depressive affect and somatic symptoms (Krause & Ellison, 2003). On the other hand, data drawn from undergraduate, community, and veteran samples indicate that unforgiveness, or holding on to negative emotions (i.e., resentment, bitterness, hostility, hatred, anger, and fear) toward transgressors (Worthington Jr & Wade, 1999), is positively correlated with depressive symptoms (Maltby et al., 2001). Results from a nationally representative probability sample of U.S. adults also indicate that forgiveness of others is related to greater psychological well-being (Tse & Yip, 2009) and lower levels of psychological distress (Toussaint et al., 2001).

With some exception (Hirsch, Webb, & Jeglic, 2011a), the small number of existing investigations that attend to the relationship between forgiveness and health focus largely on the mediating role played by forgiveness in stress-mental health associations (Lawler-Row, 2010; Snyder & Heinze, 2005). However, it is also plausible that forgiveness moderates such associations (Bierman, 2006; Bowen-Reid & Harrell, 2002; Ellison, Boardman, Williams, & Jackson, 2001; Strawbridge, Shema, Cohen, Roberts, & Kaplan, 1998; Wink, Dillon, & Larsen, 2005). The empirical evidence supporting this moderating association is nascent but instructive. One study of 372 diverse college students found that forgiveness of onself significantly moderated associations between inward-directed, and outward-directed anger, and suicidal thoughts and behaviours (Hirsch et al., 2011a). Also, a more recent cross-sectional investigation of 91 older adults discovered that self-forgiveness moderated the association between perceived burdensomeness and suicidal ideation (Cheavens, Cukrowicz, Hansen, & Mitchell, 2016). Theorists posit that forgiveness off-sets the negative effects of stress on mental health by nurturing positive affect, cognitions, and prosocial behavior (Worthington Jr et al., 2007). Among African Ameircan men, forgiveness may function similarly and might also moderate the association between ERD and depressive symptoms by attenuating anger. It is important to note that we do not posit that forgiveness (or any other personality trait) can fully protect African-Americans from the mental health consequences of ERD. However, to the extent that forgiveness serves as an individual-level index of resilience and higher order moral reasoning, the willingness of men to forgive ERD may attenuate the vulnerability to depression.

Age Considerations

Emerging findings also suggest that investigations of forgiveness, ERD, and depression should include a focus on potential age differences (Cheng & Yim, 2008; Girard & Mullet, 1997; Toussaint et al., 2001). This focus is supported by life-course developmental frameworks and research detecting age differences in religious commitment, ERD exposure, traditional masculine role norms endorsement, and in reported depression (Barrett & White, 2002; Diamond, 2006; Hammond, 2012; Stordal, Mykletun, & Dahl, 2003; Yip, Gee, & Takeuchi, 2008). The twelve month prevalence of major depressive disorder in the United States is approximately 5.8% with individuals ages 18-39, 7.2% for 40-59, and 3.4% for those ages 60 and older (Pratt & Brody, 2010). In a study of African-American men, Hammond (2012) found that ERD was more frequent among men in the 18-29 year old age group. This finding is consistent with national data indicating more frequent racial profiling by police officers among younger African-American men (Milner et al., 2016). The consideration of age-related variations in the present study speaks to the potential heightened vulnerability of younger African-American men to the impacts of ERD. Early conceptualizations of forgiveness described it as a stage and age-based process, whereby individuals become more forgiving over time (Enright, Santos, & Al-Mabuk, 1989). In a national probability sample of adults, Toussaint et al. (2001) found that middle and older adults reported higher levels of self-forgiveness and feeling forgiven by God. The propensity for adults to become more forgiving as they age is attributed to greater maturation in moral development and a shorter time perspective (Cheng & Yim, 2008; Robert D Enright et al., 1989). Cheng and Yim (2008) conducted a randomized experiment of 180 older and younger adults and learned that the former group was more willing to grant forgiveness especially under shorter time conditions. These findings suggest that a shorter perceived length of life among older individuals may enhance motivation to forgive. Findings from the National Survey of American Life, a national probability sample of African-Americans, demonstrates that among African-American men religious commitment (e.g., religious service attendance, subjective religiosity) declines during adolescence and early adulthood and rebounds as men approach adulthood and later life (Taylor, Chatters, & Levin, 2003). Hammond and colleagues (2006) found higher levels of forgiveness of ERD among older African-American men. This body of work suggests that older African-American men in our study will be more willing to grant forgiveness of ERD than their younger counterparts, and that forgiveness may produce different mental health benefits for African-American men in various age groups. This study explores these possibilities.

Theoretical Frameworks and Study Hypotheses

Building on transactional stress and coping theory (Lazarus & Folkman, 1987), we test two alternative models of the relationships between forgiveness, ERD, and depressive symptoms. Both models posit that forgiveness is associated with reduced depressive symptomatology. However, the first is an additive effects model. In this additive model, forgiveness of ERD is hypothesized to correlate with lower depressive symptomatology independent of the level of forgiveness. The second model is an interactive, or buffering, model (Barrera Jr, 1988; Wheaton, 1985). In contrast to the additive model, the buffering model suggests that forgiveness and ERD produce an interactive effect on depressive symptomatology. According to the buffering model, forgiveness might correlate with lower depressive symptomatology because it attenuates negative ERD response or appraisal. Based on this body of evidence, we hypothesize that after accounting for masculine role norms and ERD frequency:

H1: African-American men in all age groups who report lower forgiveness and more frequent ERD will report more depressive symptomatology;

H2: The positive relationship between frequent ERD and depressive symptomatology will be moderated by forgiveness strategy indicating a more attenuated relationship among men who score higher on the indices of forgiveness; and

H3: The attenuating effect of forgiveness strategies on the ERD-depressive symptomatology relationship will be more pronounced among older men.

Method

Data for the analyses were drawn from a cross-sectional study of African-American men’s health and social lives conducted in three independent waves from 2003-2010. We recruited participants from seven barbershops in Michigan, Georgia, California, and North Carolina (79.2%) and from two academic institutions and events (20.8%): a community college in Southeastern Michigan, and a historically Black university (HBU) in the south eastern region of the US. Fifty percent of the community college population was male and 22% were ethnic minorities. The HBU student population was 77% African-American and 33% male. The academic event was a 2003 conference for African-American male law enforcement professionals in Miami, FL.

Participant Recruitment Procedures

Participants were recruited using fliers, direct contact, and word-of-mouth. Barbershops were chosen as primary recruitment sites because they are trusted congregating spaces for African-American men from various socioeconomic backgrounds, and have been successfully targeted in research studies and interventions with this population (Hart Jr & Bowen, 2004) Eight barbershops characterized as “high volume” businesses (i.e., having a wait time of 30-60 minutes and serving a minimum of 30 customers daily) were approached about participation. “High volume” shops were preferred because men could use their wait time to complete the surveys. Initial contact with barbershop owners was made in person or by telephone and followed-up with a study brochure, copy of the survey, and consent forms, after which we obtained signed letters of support. We solicited and incorporated feedback from barbers into our final survey. One of eight barbershop owners declined to participate in the study. Of the men approached in barbershops, 90% verbally consented to participate; most completed the survey during the wait time. The most frequently cited reason for non-participation was time constraints. Receptionists and/or barbers invited patrons to participate in “a study about African-American men’s health”; men aged 18 or older and who self-identified as African-American were eligible to complete the survey. The onsite study recruitment staff were African-American men and women ages 18 and older who received institutional human subjects training. Such staff assisted with obtaining participant consent and clarifying questions about the study intent and human subjects protection. All respondents completed the self-administered survey independently and received a $25 gift certificate for a free haircut. Four-year, public historically Black colleges and universities and predominately White universities served as academic recruitment sites. Recruitment methods were similar at academic institutions. Specifically, study personnel approached African-American men in high-traffic areas at these locations (e.g., student union, cafeteria, conference exhibit halls) during lunch hours or other breaks; 86% of the men approached completed the survey and received a $25 gift card. The (Institution Name Removed to Permit Blind Review) Institutional Review Board approved all study procedures.

Measures

Forgiveness of racial discrimination

Forgiveness of racial discrimination was assessed with an adapted version of the 15-item Forgiveness Scale (Rye et al., 2001), which measures individual willingness to forgive others. The original Forgiveness Scale has two subscales, a) Absence of Negative, and b) Presence of Positive, which were previously identified by Rye et al. (2001). The Absence of Negative subscale (ten items) assesses the absence of negative thoughts, feelings, and behaviors toward the wrongdoer and will be referred to as “negative release.” The Presence of Positive (five items) assesses the presence of positive thoughts, feelings, and behaviors toward the wrongdoer and will be referred to as “positive embrace.” The adapted measure asked individuals how they “have responded when they have been wronged or mistreated because of their race or racism.” Participants responded to items on the full measure using a scale anchored with 1 (“strongly disagree”) to 5 (“strongly agree”). Reliability for this scale was acceptable (α = 0.73). Participant responses to the “negative release” subscale (e.g., “I feel hatred whenever I think about the person who wronged me”) and the ”positive embrace” subscales (e.g., “I pray for the person who wronged me”) were anchored in the same way as the full measure. Reliabilities for the negative release (α = 0.73) and positive embrace (α = 0.67) subscales were acceptable. These reliability coefficients are consistent with those reported by Rye et al. (2001). For each of the scales, a mean score was computed so that higher scores would indicate a greater willingness to grant forgiveness over racial discrimination.

Everyday racial discrimination (ERD)

The18-item Daily Life Experience (DLE) subscale of the Racism and Life Experiences Scales (RaLes) (Harrell, 2000) was used to assess the frequency of self-reported ERD or “microaggressions” (e.g., being ignored or not given service). A mean score was computed from responses ranging from 0 (“Never”) to 5 (“Once a week or more”). Higher DLE scores indicate more frequent ERD (Cronbach’s α=0.95).

Masculine role norms

Masculine role norms were assessed with the Restrictive Emotionality subscale of the Male Role Norms Inventory (MRNI) Restrictive a 7-item measure that assesses traditional masculinity norms around emotion disclosure (e.g., “a man should never reveal worries to others”) (Levant & Fischer, 1998). The MRNI has been validated among racial and ethnic minorities and prior studies report coefficient alphas ranging from .73 to .96. (Levant & Fischer, 1998; Levant, Rankin, Williams, Hasan, & Smalley, 2010). Research further indicates that African-American men score higher on dimensions of masculine role norms assessed by the MRNI (Levant, Majors, & Kelley, 1998). For the current study, a mean score was computed from responses ranging from 1 (“strongly disagree”) to 7 (“strongly agree”). Higher mean scores on this scale indicate greater restrictive emotionality. Cronbach’s alphas for this scale in our study was .79.

Depressive symptoms

A 12-item version of The Center for Epidemiological Studies Depression Scale (CES-D) (Radloff, 1977), a self-report scale developed for the general population, was used to assess depressive symptoms. The measure has been widely used and validated among a variety of racial/ethnic groups (Radloff, 1977; Stahl et al., 2008). Overall scores range from 0 to 60 with higher scores indicating more depressive symptomatology (Weissman, Sholomskas, Pottenger, Prusoff, & Locke, 1977). Responses ranging from 0 (“Rarely or none of the time”) to 3 (“Most or all of the time”) were summed to create an overall continuous score. Possible scores ranged from 0-36, with higher scores indicating more depressive symptomatology (Cronbach’s α = .79).

Sociodemographic and control variables

Socio-demographic variables assessed include age (18-25, 26-39, ≥40) education (≤High school, some college, and college/graduate or professional degree), marital status (currently married vs. unmarried), annual income (<$20,000, $20-39,999, ≥$40,000), employment (employed full or part-time vs. unemployed), and recruitment site type (barbershop vs. academic institutions or events). We include a measure assessing neuroticism or heightened vulnerability to negative emotion because it is an important personality substrate of forgiveness (Ashton et al., 1998; Maltby & Day, 2004; Maltby et al., 2004). Neuroticism was assessed with the 8-item neuroticism subscale of the NEO-PI (McCrae & Costa, 1987), which assesses the tendency to experience negative affectivity or emotional states. Studies investigating the role of ERD in mental health suggests controlling for this more stable personality characteristic to account for individuals who may be more sensitive to negative emotion and stress exposures (Harrell, Hall, & Taliaferro, 2003; Thompson, 1987). Participants responded to each item using a scale anchored with 1 (“strongly disagree”) and 5 (“strongly disagree”). Reliability was acceptable (α = 0.72), and a mean score was computed so that higher scale scores would indicate a greater tendency to experience negative emotions. Given the demonstrated relationships between religiosity and forgiveness (Rye, 2005), we also assessed subjective religiosity with a single-item question. Participants responded to this item (“How religious are you?”) using a scale anchored with 1 (“not at all”) and 4 (“very”).

Statistical Analysis

Simple bivariate analyses (χ2 and ANOVA) were first conducted to examine sample characteristics. The Bonferonni method was used to adjust all pairwise mean comparisons. Multivariate linear regression was used to examine the relationship between our primary independent variable (forgiveness of racial discrimination), everyday racial discrimination, masculine role norms, and depressive symptoms. The main effects of ERD, forgiveness of ERD were assessed to test Hypothesis 1 (i.e., more frequent ERD and lower forgiveness would be negatively related to depressive symptoms in all age groups). (See Model 1). Interactions were used to test Hypothesis 2 (i.e., forgiveness of ERD strategies would moderate the relationship between ERD and depressive symptoms) (See Model 2). Hypothesis 3 (i.e., that these moderated relationships would be more pronounced for older men) was assessed by computing interaction terms for ERD × forgiveness of ERD (all strategies) for each age group (See Model 2). Simple slopes and graphs were used to clarify significant interactions. All continuous predictor and control variables were mean centered (Aiken & West, 1991). For both models, we adjusted for recruitment site type, geographic region, level of education, income, marital status, and neuroticism. Multicollinearity was evaluated and found absent as evidenced by variance inflation factors (VIF) values of less than 5 (Hair, 2009). Data on forgiveness (Overall, Positive, Negative), subjective religiosity, and income were missing for 5.9%, 6.4%, and 8.0% respectively. Further analysis suggested that these values were missing at random. Hence, we used established multiple imputation procedures (Allison, 2000) to generate 5 complete data sets. Standardized beta coefficients and standard errors from these 5 data sets were examined independently and in aggregate. Since we did not observe any notable differences between values in our imputed and original data sets, we present results from the original data. All statistical analyses were performed with Predictive Analytic Software Statistics (SPSS Inc., 2010) and evaluated with two-tailed tests of significance using a 0.05 alpha level.

Results

Descriptive Statistics and Univariate Analyses

Overall, participants ranged in age from 18-79 years (M = 32.0, SD = 11.1). Most men were between 18-25 (48%), unmarried (73.4%), employed at least part-time (82.0%), and resided in the south (74.0%). Education and income were more evenly distributed (i.e. high school- 32.5%, some college- 37.5%, college/graduate/professional degree- 30.0% respectively; less than $20,000- 38.4%, $20,000-39,000- 31.0%, and ≥$40,000- 30.5% respectively). In general, the 18-25 year-old men were unmarried (90.2%), completed some college (46.5%), were employed (78.2%), had incomes of <$20,000 (55.1%), and were Southern residents (80.4%). Men recruited from academic institutions were significantly younger (72.1% were 18-25 years of age), unmarried (87.0%), reported completing some college (61.1%), had incomes of <$20,000 (70.5%), resided in the South (69.3%), and had higher levels of depressive symptoms compared to men recruited from barbershops (p <.01, M= 10.91). A higher percentage of ≥40 year-old men category completed a college, graduate, or professional degree (44.5%), were employed (85.5%), earned ≥$40,000 (54.2%), and resided in the South (62.3%).

Men recruited from barbershops had higher overall mean depressive symptom scores, F (1, 656) = 7.64, p = .006, and higher mean combined forgiveness scores, F (1, 639) = 5.81, p = .016, compared to men who were recruited from academic institutions. Mean CES-D scores were higher among 18-25 year-olds than men in all other age groups (p <.001-.002). Men ages 26-39 years had higher mean CES-D scores than those in the ≥40 year-old men (p = .002). Men in the 18-25 year-old age category also reported more frequent exposure to ERD (p = .002), lower subjective religiosity (p = .001) and lower mean combined (p <.001) and negative release (p = .001) forgiveness scores than ≥40 year-old men. Mean positive embrace forgiveness scores were higher among 26-39 year-old men than those reported by 18-25 year-old men (p = .04).

Tables 2, 3, and 4 display results from the age-stratified multivariate linear regression analyses investigating the association between the three forgiveness strategies, ERD, and depressive symptoms adjusted for recruitment site type, level of education, income, marital status, neuroticism, and subjective religiosity. Consistent with Hypothesis 1, men in the full sample who report less forgiveness of ERD (β = −.12, p <.01), and more ERD experiences (β = .20, p <.001) also reported more depressive symptoms regardless of the forgiveness strategy. The direct association between forgiveness strategies and depressive symptoms were age-dependent. Specifically, we found that combined (β = −.21, p <.001) and negative embrace forgiveness (β = −.22, p <.001) were associated with less depressive symptoms but only among men in the 18-25 year-old age group. No direct effect for positive embrace forgiveness was observed.

Table 2.

Multivariate Linear Regression Analysis of Associations Between Forgiveness of Racial Discrimination (Combined), Everyday Racial Discrimination, and Depressive Symptoms among Community-Dwelling African-American, 2003-2009

Characteristic	Model B (SE)
	Total Sample (N = 674)		18-25 (n = 325)		26-39 (n = 180)		40+ (n = 169)
	B	(SE)	B	(SE)	B	(SE)	B	(SE)
Model 1
Intercept	16.20***	(.79)	15.95***	(.90)	13.17***	(1.63)	10.46***	(1.48)
Forgiveness of Racial Discrimination (Combined)	−.12**	(.41)	−.21***	(.67)	−.02	(.70)	−.10	(.85)
Everyday Racial Discrimination (ERD)	.20***	(.17)	.18***	(.24)	.15*	(.33)	.29***	(.36)
Masculine Role Norms (Restrictive Emotionality)	.12***	(.17)	.10*	(.27)	.12	(.28)	.13	(.36)
Adjusted R2	.45***		.42***		.47***		.38***
Model 2
Intercept	16.14***	(.78)	15.53***	(.90)	13.10***	(1.61)	10.18***	(1.48)
Forgiveness of Racial Discrimination (Combined)	−.14***	(.41)	−.25***	(.69)	−.01	(.69)	−.12	(.85)
Everyday Racial Discrimination (ERD)	.18***	(.17)	.14**	(.24)	.14*	(.33)	.29***	(.36)
Masculine Role Norms (Restrictive Emotionality)	.12***	(.17)	.10*	(.26)	.12	(.27)	.13	(.35)
Forgiveness of Racial Discrimination (Combined)× ERD	−.11**	(.32)	−.13**	(.58)	−.12*	(.49)	−.13	(.61)
Adjusted R2	.46***		.44***		.48***		.39***

Note: B = Standardized Regression Coefficient; SE = Standard Error. Models adjusted for age (total sample only), recruitment site type, marital status, education, subjective religiosity, and neuroticism.

^*P < .05;

^**P <. 01;

^***P < .001

Table 3.

Multivariate Linear Regression Analysis of Associations Between Forgiveness of Racial Discrimination (Positive Embrace), Everyday Racial Discrimination, and Depressive Symptoms among Community-Dwelling African-American, 2003-2009

Characteristic	Model B (SE)
	Total Sample (N = 674)		18-25 (n = 325)		26-39 (n = 180)		40+ (n = 169)
	B	(SE)	B	(SE)	B	(SE)	B	(SE)
Model 1
Intercept	16.90***	(.79)	16.41***	(.80)	13.22***	(.1.63)	10.63***	(1.49)
Forgiveness of Racial Discrimination (Positive Embrace)	−.06	(.29)	−.09	(.47)	−.03	(.51)	−.06	(.59)
Everyday Racial Discrimination (ERD)	.23***	(.17)	.22***	(.25)	.16*	(.32)	.30***	(.35)
Masculine Role Norms (Restrictive Emotionality)	.14***	(.17)	−.13**	(.27)	.12	(.27)	.15	(.36)
Adjusted R2	.44***		.41***		.49***		.41***
Model 2
Intercept	16.55***	(.80)	16.45***	(.80)	13.29***	(1.63)	10.63***	(1.49)
Forgiveness of Racial Discrimination (Positive Embrace)	−.07	(.29)	−.10*	(.48)	−.02	(.51)	−.06	(.61)
Everyday Racial Discrimination (ERD)	.23***	(.17)	.22***	(.25)	.16*	(.32)	.31***	(.35)
Masculine Role Norms (Restrictive Emotionality)	.14***	(.17)	.14**	(.27)	.12	(.27)	.15	(.36)
Forgiveness of Racial Discrimination (Positive Embrace)× ERD	−.05	(.24)	−.04	(.40)	−.07	(.39)	.00	(.46)
Adjusted R2	.44***		.41***		.49***		.41***

Note: B = Standardized Regression Coefficient; SE = Standard Error. Models adjusted for age (total sample only), recruitment site type, marital status, education, subjective religiosity, and neuroticism.

^*P < .05;

^**P <. 01;

^***P < .001

Table 4.

Multivariate Linear Regression Analysis of Associations Between Forgiveness of Racial Discrimination (Negative Release), Everyday Racial Discrimination, and Depressive Symptoms among Community-Dwelling African-American, 2003-2009

Characteristic	Model B (SE)
	Total Sample (N = 674)		18-25 (n = 325)		26-39 (n = 180)		40+ (n = 169)
	B	(SE)	B	(SE)	B	(SE)	B	(SE)
Model 1
Intercept	16.21***	(.79)	16.06***	(.90)	13.17***	(1.64)	10.32***	(1.50)
Forgiveness of Racial Discrimination (Negative Release)	−.13**	(.36)	−.22***	(.53)	−.01	(.61)	−.10	(.79)
Everyday Racial Discrimination (ERD)	.18***	(.18)	.13**	(.25)	.15*	(.35)	.29***	(.37)
Masculine Role Norms (Restrictive Emotionality)	.12***	(.17)	.10*	(.27)	.12	(.28)	.12	(.36)
Adjusted R2	.45***		.44***		.49***		.41***
Model 2
Intercept	15.90***	(.73)	15.69***	(.92)	12.95***	(1.62)	9.812***	(1.47)
Forgiveness of Racial Discrimination (Negative Release)	−.13***	(.35)	−.23***	(.53)	−.00	(.60)	−.09	(.78)
Everyday Racial Discrimination (ERD)	.17***	(.18)	.11*	(.26)	.12	(.35)	.30***	(.36)
Masculine Role Norms (Restrictive Emotionality)	.12***	(.17)	.10*	(.27)	.12*	(.27)	.13	(.35)
Forgiveness of Racial Discrimination(Negative Release)× ERD	−.11**	(.25)	−.09	(.39)	−.14*	(.44)	−.19**	(.53)
Adjusted R2	.46***		.44***		.51***		.44***

Note: B = Standardized Regression Coefficient; SE = Standard Error. Models adjusted for age (total sample only), recruitment site type, marital status, education, subjective religiosity, and neuroticism.

^*P < .05;

^**P <. 01;

^***P < .001

Partial support was found for Hypothesis 2 (i.e., that forgiveness of ERD would moderate the relationship between ERD and depressive symptoms). The interaction between ERD and forgiveness of racial discrimination for the full sample was significant, indicating a more pronounced positive association between ERD and depressive symptoms among men with lower combined (β = −.11, p <.01; Figure1a) and negative release forgiveness of ERD scores (β = −.26, p <.01; Figure 2a). No significant interaction effects between ERD and positive embrace forgiveness were observed.

Figure 1.

Forgiveness of Racial Discrimination (Combined) as a Moderator of the Association between Everyday Racial Discrimination (ERD) and Depressive Symptoms among African-American Men (N = 674)

Note: CES-D = Center for Epidemiological Studies Depression Scale; ERD = Everyday Racial Discrimination; SD = Standard Deviation

Figure 2.

Forgiveness of Racial Discrimination (Negative Release) as a Moderator of the Association between Everyday Racial Discrimination (ERD) and Depressive Symptoms among African-American Men (N = 674)

Note: CES-D= Center for Epidemiological Studies Depression Scale; ERD = Everyday Racial Discrimination; SD = Standard Deviation

Hypothesis 3 (i.e., that the moderating role of forgiveness on the ERD-depressive symptoms relationship would be more pronounced among older men) was not supported. Our results suggests that there was no attenuation in mean CES-D scores among men who report less frequent ERD and granting forgiveness for ERD. There was a less pronounced, positive relationship between ERD and depressive symptoms among men in the 18-25 (β = −.13, p <.01) and 26-39 (β = −.12, p <.05) year-old age groups who reported using a combined forgiveness strategy. We also observed a less pronounced ERD-depressive symptoms relation among men in the 26-39 (β = −.14, p <.05) and ≥40 (β = −.19, p <.01) year-old age groups who reported using a negative release forgiveness strategy. Low combined and negative release forgiveness appeared to be associated with a more pronounced positive ERD-depressive symptoms relationship, particularly among men in the 18-25 year old age group (Figures 1b and 1c).

Discussion

Our study contributes to the growing scientific interest in identifying mechanisms linking racial discrimination to disparate health outcomes (Williams & Mohammed, 2009). In general, the study hypotheses were supported. Our main findings are as follows. First, ERD was associated with more depressive symptoms for men in every age group after controlling for key correlates (e.g., masculine role norms and subjective religiosity) and sociodemographic factors (e.g., marital status, education). Reifying previous study results (Lincoln, Taylor, Watkins, & Chatters, 2010; Watkins et al., 2006; Williams & Williams-Morris, 2000), this finding poignantly affirms that the negative mental health consequences of racial discrimination exposure persist for African-American men across the developmental life-course. Second, interaction effects reveal that the mental health consequences of exposure to ERD are particularly pronounced for men who score low on forgiveness. This finding affirms previous work indicating that religion-related resources buffer individuals from the health impacts of stress (Bierman, 2006; Ellison et al., 2001; Strawbridge et al., 1998). The exploration of three strategies of forgiveness—1) combined, 2) the letting go of negative emotion (negative release), and 3) embracing positive emotion (positive embrace)—provide a textured and more nuanced approach to the examination of the ways that forgiveness may function for men coping with everyday race-related transgressions. More specifically, analyses revealed that the ERD-depression link was stronger among men who were low on combined and negative release forgiveness.

Younger men (ages 18-25) in the sample reported experiencing ERD with the greatest frequency and scored higher on measures assessing depressive symptomatology than their older counterparts. We note, too, that the youngest subsample of men (18-25 year olds) scored lower than their older counterparts on combined forgiveness. Further, multivariate findings indicate that the ability and willingness to forgive was associated with depressive symptomatology for this age group such that young men who scored higher on forgiveness scored significantly lower on depressive symptomatology. It is plausible that this group may encounter ERD more frequently, and that they thus have more challenges with granting forgiveness, but that they receive the greatest mental health benefit (with regard to depression) when they do forgive. Our findings among this age group may be partly explained by developmental processes or may be a function of differences in the historical moments in which men are coming of age (i.e., differences in cohort).

From a developmental perspective it is reasonable to presume that, by virtue of their age and level of experience in negotiating the demands of the social world, younger African-American men may not have the same socio-emotional skills and protections as their older counterparts. To the extent that such skills and protections are less available, these young men may be less prepared to withstand, and may be more vulnerable to, the deleterious psychological effects of exposure to ERD. This explanation is in line with previous research demonstrating that for younger African-American men, endorsement of masculinity norms encouraging emotion restriction was associated with lower forgiveness of racial discrimination (Hammond et al., 2006). Research finds that suppressing negative emotions, being cognitively inflexible, or intolerant of distress can have broad mental health implications (Schmidt, et al., 2007, Wenslaff & Wegner, 2000). If younger African-American men are also attempting to restrict or shut down emotions associated with ERD, they may be able to temporarily bury hatchets but find their locations more marked in their cognitions and memories because of “post-suppression rebound effects” (Wenzlaff & Wegner, 2000). Post-suppression rebound effects associated with ERD may make it more difficult for younger men to dispense of negative affect further complicating the forgiveness process. Implications of such ideas would require society to validate the emotional experiences of younger African-American men as they relate to discrimination rather than rush them to forgive, avoid or curb intense emotional responses. It is also possible that these results reflect different demands associated with various ages and stages of development (e.g., the identity development needs are different for emerging adults than for men who are moving toward generativity). Taken together, these findings also suggest the need to support younger men in developing the variety of affect regulation skills needed to navigate ERD. Future work assessing exposure to ERD and its mental health correlated over the life-course among African-American men are needed to fully test this possibility.

A third crucial contribution of this work is its exploration of forgiveness of ERD and its roll in off-setting depressive symptoms among African-American men of various ages. High levels of overall and negative forgiveness served a buffering function for men, albeit differentially across age groups. However, low levels of overall and negative forgiveness appears to exacerbate the relation between ERD and depression for younger men. For example, even though our findings imply that combined forgiveness of ERD had a positive, off-setting effect on African-American men’s health, this effect was limited to men in the 18 to 25 year old and the 26-39 year old groups. Similarly, although forgiveness through the release of negative emotions moderated the relationship between ERD and depressive symptomatology, this effect was evident only in the 26 to 39 and the 40 and older age groups. Age differences in forgiveness scores raise questions about whether, for men, the capacity or willingness to engage in positive as well as negative forgiveness evolves over a lifetime or whether, consistent with a cohort analysis, the forgiveness process for young men who are coming of age in our present sociopolitical context is different than for their older counterparts who came of age in a somewhat different historical moment. For example. older men may have a higher threshold for the kinds of offenses that elicit the need to deploy forgiveness because of experience with more virulent kinds of racial discrimination characteristic of earlier historical times. This higher threshold among older men may also be attributable to greater maturation in moral reasoning (Enright, Gassin, & Wu, 1992) and affect regulation capabilities.

Building on the mean differences and moderation analyses, low levels of forgiveness in the face of discrimination appear to be more detrimental for young African-American men who reported lower mean levels of forgiveness and higher mean levels of ERD. These findings highlight the bind experienced by African-American men. Systems that perpetuate ERD remain despite proclamations of being a post-racial society. For example, data suggest African-American men experience a disproportionate amount of stop-and-frisks, which stem from law enforcement and criminal justice systems (Milner et al., 2016; Ridgeway, 2007). The bind of younger African-American men is particularly important to acknowledge given that such disproportionality results in higher social surveillance and would not exist if contemporary American society were truly post-racial. For men who continue to endure the consequences of ERD, the tendency to be less forgiving and mark the location of such transgressions could be considered a rational choice.

We were struck by the fact that a greater willingness to forgive by adopting positive views of and embracing positive affect in relation to their transgressors had no direct prophylactic effect on depressive symptoms nor did it moderate the effects of ERD on depression in any of the age groups. We assert that while forgiveness may have moral or humanistic value as men grapple with other psychosocial stressors, it may be less effective when fronting persistent, more ambiguous race-based stressors. The unilateral nature of granting forgiveness in the aftermath of ERD may also limit men’s capacity to generate positive affect, thoughts, and behaviors. Without opportunities for reconciliation, dialogue, or negotiated forgiveness, African-American men may have a harder time cultivating the desire to positively embrace transgressors. This finding implies that the kind of forgiveness promoted most centrally by religious doctrine or sacred texts (e.g., praying for and openly embracing one’s transgressors) may not always be health beneficial in the context of repeated race-related violence or abuses (Gordon, Burton, & Porter, 2004; James K McNulty, 2011). Researchers speculate that granting forgiveness in these cases may result in an erosion of self-respect and self-concept (Luchies, Finkel, McNulty, & Kumashiro, 2010). In other words, when ERD exposures are persistent and no immediate opportunity for remediation or social justice are present, granting forgiveness by generating positive thoughts and emotions may have little protective benefit or even unintended, negative mental health consequences for African-American men.

Two particularly important summative points emerge from our findings. First, our findings suggest that the relinquishment of negative emotions not the capacity to internalize positive feelings about transgressors drives the mitigating impact of forgiveness on depressive symptoms. In other words, acts of forgiveness that allow men to dispense with rumination, frustration, anger, and hurt may be more valuable in improving or preserving mental health than acts of forgiveness that may, by virtue of the focus on positive emotions, invite men into deeper relationship with others who have transgressed against them. These findings are consistent with previous research, which found that low levels of vengeance lessen the odds for depressive symptoms (Kendler et al., 2003). Second, our results imply that the capacity to forgive may be protective when men are younger (e.g., when men are in emerging adulthood), but may not be as predictive of depressive symptomatology as men get older.

Study Limitations

As is the case in all studies this study has limitations. First, the study relies exclusively on cross-sectional data, as such we can make no assertions regarding the direction of the relationships between study variables. In the absence of longitudinal data it remains unclear whether the cognitive biases inherent in depression may lead depressed individuals to experience ERD as especially salient or to be less forgiving. Longitudinal approaches to the study of ERD, forgiveness, and mental health among African-American men will help to clarify the directionality of these associations. Second, the study relied on the use of self-reported measures on aspects of experience (e.g., forgiveness and spirituality) that one might presume to increase socially desirable responses. Third, the sample, although demographically similar to the population of African-American men in the U.S., is not nationally representative (United States Bureau of Labor Statistics, 2008; United States Census Bureau, 2009) As such, these findings may not be generalizable. The psychometric properties of the CES-D have been widely assessed and the measure has been validated among racial/ethnic group members (Stahl et al., 2008; Wada et al., 2007; Weissman et al., 1977). However, this measure has been found to exhibit some bias when used among men and among African-Americans (Long Foley, Reed, Mutran, & DeVellis, 2002). Researchers will want to use caution in the conclusions they draw from the study findings. We also collapsed categories at the tail end of our sample’s age distribution into African-American men ages 40 and older. This decision was largely driven by the smaller number of men in the 50-59 (n = 45) and 60 and older category (n = 9). Prior studies examining depression onset and impacts, have collapsed age categories in a similar way (Weissman, Wickramaratne, Merikangas, & et al., 1984) or used 40 years and older as a cut point (Korten, Comijs, Lamers, & Penninx, 2012). However, we realize that doing so may have obscured important later life-course developmental differences in forgiveness of ERD and depression. Older African-American men who have more experiences with blatant racial discrimination may have different strategies and motivations for deploying forgiveness following ERD. Future studies will want to recruit a larger sample of African-American men so that these age-related variations can be fully investigated.

These limitations notwithstanding, this study has important strengths. First, by elucidating age differences in the ERD-depression matrix, this study highlights life-course variability in the landscape of African-American men’s depressive symptoms. Second, the study’s examination of the interplay between discrimination, positive psychological resources, and mental health. By focusing on both forgiveness of ERD across different age groups, the study examines the ways in which the availability of psychosocial assets across the life-course complicates our understanding of African-American men’s coping outcomes. Our investigation also adds to the growing scientific push towards studies that examine the conditions under which forgiveness may promote better health (McNulty & Fincham, 2012). In addition to its strengths, this study leaves us with important points for consideration. Our work is rooted in a recognition that the persistence of ERD in contemporary American life points to a critical need for sustained advocacy and structural change. Little can be gained from simply promoting forgiveness of ERD among African-American men without eradicating ERD in the places they live, work, play, and get healthcare. While we advocate for such change, we also recognize that social science can significantly enhance interventions with African-American men by unearthing the positive intra and interpersonal resources they leverage to combat exposure to ERD. The fact that African-American men make concerted, unilateral efforts to forgive even in threatening contexts speaks volumes about their strengths, moral reasoning, and capacity for resilience.

We must also consider the point that despite mounting evidence of an ERD link to depressive symptoms, all race related stressors are not equal. As such, future studies should explore similarities and differences in the forms of racism to which African-American men of different ages are exposed, and the extent to which these ERD experiences may prove to be especially taxing to a broad range of mental and physical health outcomes among emerging adult men, adult men, and men in middle to older adulthood. Indeed, we note here that while forgiveness may be associated with lower levels of depression, the use of that strategy may be differentially associated with other outcomes including anxiety, self-concept, and cardiovascular reactivity. Further, research should explore how, as men age, they use, adapt, and discard certain strategies (e.g., affective restriction, cognitive flexibility, forgiveness, faith) in an effort to adjust to the shifting stresses and coping demands of their environments. Our work advances scientific exploration of racism and mental health among African-American men. Specifically, this study demonstrates that the most potent forms of forgiveness for African-American men appear to be those that center explicitly on the capacity to forgive, not by forgetting race-based transgressions (burying hatchets), or marking their emotional location, but rather by loosening their intrapsychic grip on everyday life.

Table 1.

Characteristics of the Study Sample by Age

Characteristics a	18-25 (n = 325)	26-39 (n = 180)	40+ (n = 169)	P-Value b
	% or mean ± SD	% or mean ± SD	% or mean ± SD
Marital Status
Married	31 (9.8)	69 (40.1)	75 (45.5)	<.001
Unmarried	286 (90.2)	103 (59.9)	90 (54.5)
Education
High School or less	101 (31.8)	60 (34.5)	49 (29.9)
Some College	148 (46.5)	57 (32.8)	42 (25.6)	<.001
College, graduate, or professional degree	69 (21.7)	57 (32.8)	73 (44.5)
Employment Status
Employed	247 (78.2)	148 (85.5)	142 (87.1)	.023
Unemployed	69 (21.8)	25 (14.5)	21 (12.9)
Annual Income
<$20,000	167 (55.1)	43 (25.4)	24 (16.9)
$20,000-39,000	90 (29.7)	61 (36.1)	41 (28.9)	<.001
≥$40,000	46 (15.2)	65 (38.5)	77 (54.2)
Region
North	3 (0.9)	2 (1.1)	11 (6.6)
Midwest	26 (8.1)	7 (3.9)	8 (4.8)	<.001
South	258 (80.4)	135 (75.0)	104 (62.3)
West	34 (10.6)	35 (19.4)	44 (26.3)
Depressive Symptoms (CES-D Score)	12.71 ± 6.0	10.83 ± 5.3	8.68 ± 5.7	<.001
Forgiveness of Racial Discrimination (Overall)	3.21 ± 0.4	3.31 ± 0.5	3.40 ± 0.6	.001
Forgiveness of Racial Discrimination (Positive)	3.05 ± 0.6	3.21 ± 0.7	3.18 ± 0.7	.026
Forgiveness of Racial Discrimination (Negative)	3.32.± 0.6	3.38 ± 0.6	3.53 ± 0.6	.002
Everyday Racial Discrimination (ERD)	1.87 ± 1.2	1.75 ± 1.1	1.47 ± 1.1	.003
Masculine Role Norms (Restrictive Emotionality)	4.06 ± 1.1	4.11 ± 1.2	3.81 ± 1.1	.043
Neuroticism	2.59 ± 0.6	2.49 ± 0.7	2.35 ± 0.7	<.001
Subjective Religiosity	2.62 ± 1.0	2.72 ± 1.0	2.98 ± 1.0	.001

Note: CES-D = Center for Epidemiological Studies Depression Scale; SD = Standard Deviation.

^aThere were 674 respondents. However there were missing data for some participant characteristics. Data on forgiveness (Overall, Positive, Negative), subjective religiosity, and income were missing for 5.9%, 6.4%, and 8.0% respectively. Analysis suggests that the data were missing at random (MAR).

^bComparisons are based on the χ² statistic for categorical variables and the F statistic for continuous variables.