Self-Reported Experiences with Stigma

Reviews of self-reported experiences of stigmatization on markers of physical health (e.g., blood pressure, hypertension) reveal mixed findings (Harrell et al., Reference Harrell, Hall and Taliaferro2003) or conditional effects (Williams and Mohammed, Reference Williams and Mohammed2009). For instance, no consistent effects are seen comparing the blood pressure of African Americans who report experiencing less stigmatization in the past to that of those who have experienced more stigmatization in the past (Harrell et al., Reference Harrell, Hall and Taliaferro2003). A recent meta-analysis revealed that greater self-reported experiences of discrimination significantly but weakly predict greater stress responses (r = −0.11), and worse physical health (r = −0.13) (Pascoe and Smart Richman, Reference Pascoe and Smart Richman2009). Thus, an effect of perceptions of stigmatization on physical health is found but it is neither strong nor direct.

More consistent yet still weak effects are found when looking at the relations between people's self-reported stigmatization experiences and (a) mental health, and (b) well-being (Williams and Williams-Morris, Reference Williams and Williams-Morris2000). Of course, cross-sectional research suffers from interpretation problems given potential confounds. Thus, it is important to note that longitudinal studies reveal that, when controlling for initial mental health, experiences of stigmatization predict later mental health (Pavalko et al., Reference Pavalko, Mossakowski and Hamilton2003). Also, the more people report that they have experienced stigmatization, the worse their psychological well-being (e.g., lower self-esteem and greater depression) (Schmitt et al., Reference Schmitt, Branscombe, Kobrynowicz and Owen2002; Williams and Williams-Morris, Reference Williams and Williams-Morris2000). A recent meta-analysis revealed that regardless of their group memberships (e.g., ethnicity, mental disorder), the more people experience stigmatization, the worse their mental health (r = −0.28) (Mak et al., Reference Mak, Poon, Pun and Cheung2007). Thus, stigmatization is harmful for mental health and well-being. However, these effects are likely weak because they are conditional on other factors (to be reviewed later). Such conditional effects weaken the overall relation found between stigmatization experiences and mental health and well-being.

The anticipation of being stigmatized is also harmful for devalued group members. Stigma consciousness (Pinel Reference Pinel1999), stigma sensitivity (Inzlicht et al., Reference Inzlicht, McKay and Aronson2006), and group-based rejection sensitivity (e.g., race-based rejection sensitivity) all refer to the degree to which devalued group members are chronically aware of their stigmatized status and expect to be devalued. Among devalued group members, those who more strongly anticipate stigma report that they struggle more with self-regulation (e.g., motivation) (Inzlicht et al., Reference Inzlicht, McKay and Aronson2006), with academic performance (Brown and Lee, Reference Brown and Lee2005), and with depressive symptoms (Lewis et al., Reference Lewis, Derlega, Griffin and Krowinski2003). Thus, the anticipation of being stigmatized negatively affects devalued members' mental health, and their cognitive and scholastic performance.

In summary, the effects of self-reported expectations and experiences of stigmatization are negative yet weak. Stronger effects are found for psychological outcomes (e.g., depression, well-being) than for physiological stress responses and physical health. Yet, effects are often still conditional. It is possible that a reliance on targets' self-reported experiences will distort the relation found between stigmatization and outcomes because such measurement excludes subtle and undetected stigmatization.Footnote ² Therefore, I turn now to a consideration of the effects of experimental manipulations of stigmatization on devalued group members' health, well-being, and performance.

Experimentally Inducing Stigmatization

Experimental investigations consistently reveal that the experience of stigmatization (vs. no stigmatization) results in greater physiological reactivity, such as cardiovascular reactivity with moderate to large effect sizes (Blascovich et al., Reference Blascovich, Spencer, Quinn and Steele2001; Murphy et al., Reference Murphy, Steele and Gross2007; Williams and Mohammed, Reference Williams and Mohammed2009). Stigmatization experiences are manipulated in the lab by exposing devalued group members to prejudice, by reminding them of past experiences of discrimination, or by reminding them of their stigmatized group status, or of negative stereotypes about their group, etc.

Experimentally inducing stigmatization also causes devalued members to have worse well-being (Crocker et al., Reference Crocker, Cornwell and Major1993). For instance, when imagining that they are rejected from a course by a sexist professor, women feel more hostility (Major et al., Reference Major, Kaiser and McCoy2003a) and negative affect (Schmitt and Branscombe, Reference Schmitt and Branscombe2002a) than those who are rejected without any suggestion of sexism. When devalued group members are reminded of prejudice toward their ingroup, similar negative effects are found for well-being. For instance, Latinos who read about prejudice toward their ingroup feel more hostility and depressed affect than those who read about prejudice toward the Inuit (McCoy and Major, Reference McCoy and Major2003). Thus, experimental manipulations of stigmatization can lead to hostility, negative affect, and lowered self-esteem. However, as I will discuss later, these effects are often moderated by multiple factors.

Finally, experiences of stigmatization negatively affect people's performance in difficult test-taking situations. “Stereotype threat” occurs if, when struggling with a difficult test, a negative stereotype becomes salient concerning the ingroup's ability in a relevant domain (Steele Reference Steele1997). Stereotyped group members can then fear they will be evaluated in terms of the stereotype or that they will confirm the stereotype; these threats cause underperformance, compared with the performance levels of nonstereotyped group members (Steele Reference Steele1997). Importantly, when the threat of the stereotype is removed (e.g., by informing participants that tests are nondiagnostic or unbiased), so too is the performance gap between highly competent stereotyped and nonstereotyped group members (Steele et al., Reference Steele, Spencer, Aronson and Zanna2002).

Manipulations of stereotype threat affect performance for a variety of groups in stereotype-relevant domains, such as women's performance on math tests (Spencer et al., Reference Spencer, Steele and Quinn1999), Blacks' performance on intelligence tests (Steele and Aronson, Reference Steele and Aronson1995), older adults' performance on memory tests (Hess et al., Reference Hess, Auman, Colcombe and Rahhal2003), and gay and bisexual men's quality of interaction with children (Bosson et al., Reference Bosson, Haymovitz and Pinel2004). In addition to performance decrements on tests, among Blacks, stereotype threat has been shown to increase blood pressure (Blascovich et al., Reference Blascovich, Spencer, Quinn and Steele2001), and activate self-doubt (Steele and Aronson, Reference Steele and Aronson1995). Among women, stereotype threat can lead to disengagement with quantitative majors and occupations (Davies et al., Reference Davies, Spencer, Quinn and Gerhardstein2002). The effects of stereotype threat on stigmatized group members' test-taking performance are robust; however, they too are conditional on other factors (Schmader et al., Reference Schmader, Johns and Forbes2008; Steele Reference Steele1997).

To summarize, consistent negative effects of stigmatization are found with experimental studies for devalued group members' physiological reactivity and their performance on difficult tests. These effects are robust and they are of moderate magnitude. However, less consistency is found when investigating the experimental effects of stigmatization on people's well-being, as these processes are often conditional.

Taking Stock

Some might be tempted to conclude that due to findings for experiences of stigmatization that are often mixed and weak, we should not be too concerned with its consequences. Such a conclusion would be erroneous. The relation between self-reported experiences of stigmatization on health and well-being might be underestimated in part because stigmatization can be difficult to detect given its often subtle nature (Crosby Reference Crosby1984). Furthermore, when it is detected, devalued group members are often reticent to report that they have been victims of stigmatization (Crocker et al., Reference Crocker, Major, Steele, Gilbert, Fiske and Lindzey1998; Ruggiero and Taylor, Reference Ruggiero and Taylor1997).Footnote ³ Finally, weak effects can be aggregated across domains and over time to have a large impact and great practical significance (Martell et al., Reference Martell, Lane and Emrich1996). Therefore, weak effects do not translate into unimportant phenomena.

Others might infer that, because the effects of stigmatization on devalued group members are often conditional, the phenomenon is too complicated to make conclusions about responses to stigmatization. However, we should expect the effects of stigmatization to be conditional (i.e., moderated) given what we know about psychological resilience: in the face of risks, protective factors or assets create opportunities for the achievement of positive outcomes, such as well-being (Armstrong et al., Reference Armstrong, Birnie-Lefcovitch and Ungar2005; Masten Reference Masten2001). Thus, responses to risks, such as stigmatization, should depend on people's assets, making resilience neither exceptional nor guaranteed. Although some of these assets are based in the individual (e.g., disposition or psychological outlook), others stem from the environmental context, such as social networks, local institutions, and cultural frameworks (Hall and Lamont, Reference Hall and LamontForthcoming). Furthermore, I would suggest that the list of known moderators of responses to stigmatization is not a laundry list. Rather, it can be clearly organized into a framework to understand when and for whom the negative effects of stigmatization will be exacerbated or mitigated by drawing on the transactional model of stress and coping (Lazarus and Folkman, Reference Lazarus and Folkman1984).

Self-Relevance

A strong sense of group identification can, in some ways, make people more vulnerable to stigmatization (Major et al., Reference Major, McCoy, Kaiser, Quinton, Stroebe and Hewstone2003b). Identification can involve many elements; however, I will focus on the degree to which people see their group membership as central to their identity (McCoy and Major, Reference McCoy and Major2003). The negative effects of stigmatization on cognitive processing (Bair and Steele, Reference Bair and Steele2010) and on test-taking performance are stronger for those who are more identified with their ingroup (Schmader et al., Reference Schmader, Johns and Forbes2008). Similar effects are found for well-being. For instance, among Latino Americans who read that prejudice toward Latinos is pervasive, the stronger their ingroup identification, the more depressed affect they feel (McCoy and Major, Reference McCoy and Major2003). Thus, there is some evidence that stigmatization negatively affects well-being and cognitive performance more strongly for devalued group members who are more strongly identified with their ingroup.

There is some evidence that the interactive effects of ingroup identification and stigmatization operate through primary appraisals of harm and self-relevance. The more strongly people identify with their ingroup, the more they should perceive stigmatization as self-relevant and harmful; thus the greater stress they should experience (Branscombe et al., Reference Branscombe, Schmitt and Harvey1999). There is evidence that those with a stronger ingroup identification appraise stigma as more harmful (Rusch et al., Reference Rusch, Corrigan, Wassel, Michaels, Olschewski, Wilkniss and Batia2009) and as more threatening, which leads to more negative effects on well-being (McCoy and Major, Reference McCoy and Major2003). Thus, the more strongly people identify with their devalued ingroup, the more likely they are to perceive stigmatization as self-relevant and harmful; consequently, stigmatization leads to greater feelings of threat; as a result, well-being and cognitive performance are diminished.

When negative stereotypes exist about an ingroup's abilities for a specific domain, identifying with that domain can exacerbate the negative effects of stigmatization. Many studies reveal that the effects of stereotype threat are most damaging to individuals who place great value on, or who are highly identified with, the domain being tested (Steele Reference Steele1997). For instance, women who are told that a math test produces gender differences perform worse, compared with those who write the identical test but are told that the test does not produce gender differences; however, this effect is found only for those strongly identified with math (Keller Reference Keller2007).

The more strongly people identify with the domain (e.g., cognitive ability, math) for which stigmatization occurs, the more self-relevant and harmful stigmatization should seem, and consequently the more stressful it should be. This is because a domain can become part of the self-concept (Keller Reference Keller2007). The effects of stereotype threat on performance operate in part through an increased stress response (i.e., increased activation of the sympathetic nervous system and the hypothalamic-pituitary-adrenal axis), which impairs cognitive functioning (Schmader et al., Reference Schmader, Johns and Forbes2008).Footnote ⁴ Interestingly, stereotype threat effects are not mediated by self-reported anxiety. Rather, stereotype threat affects indirect (or implicit) measures of anxiety because people under stereotype threat work hard to monitor and suppress their anxiety (Johns et al., Reference Johns, Inzlicht and Schmader2008). Thus it seems reasonable to conclude that in the context of taking difficult tests, stereotyped group members who are more strongly identified with a relevant domain appraise poor performance as a more self-relevant and harmful threat; consequently, they experience greater stress and try to regulate it, which impairs their performance.

Harm

When experiencing negative treatment with ambiguous roots, that is, when it is unclear whether stigmatization is occurring, anticipation of stigma or beliefs about how pervasive stigmatization is can exacerbate its negative effects (Schmitt and Branscombe, Reference Schmitt and Branscombe2002a). The negative effects of ambiguous instances of stigmatization are more adverse for those who more strongly anticipate negative treatment based on group memberships or who perceive their ingroup to face more chronic discrimination (Mendoza-Denton et al., Reference Mendoza-Denton, Downey, Purdie, Davis and Pietrzak2002). For instance, the more Black Americans experience microaggressions (e.g., being mistaken for someone else of the same race), the worse their well-being, but only if they also believe that others have negative expectations of Blacks (Sellers et al., Reference Sellers, Copeland-Linder, Martin and Lewis2006).

Importantly, in ambiguous situations that might involve stigmatization, chronic beliefs about stigmatization drive stronger physiological stress responses (Townsend et al., Reference Townsend, Major, Gangi and Mendes2011). This is likely because the more devalued group members attribute negative treatment to racism or sexism, the more they construe the negative treatment to be personally meaningful, and consequently, the more psychological stress they feel (King Reference King2005). Thus, when devalued group members face ambiguous negative treatment, the more they perceive themselves and/or their group to be chronically stigmatized, the more likely they are to make attributions to stigmatization, which is construed as more harmful and self-relevant, leading to more stress; consequently, the worse their well-being.

In a seemingly contradictory way, when people clearly recognize that they are the targets of discrimination, the belief that society is meritocratic for all appears to make the experience of stigmatization worse for devalued groups (Major et al., Reference Major, Kaiser, O'Brien and McCoy2007). Among American ethnic minorities and women who believe that they have been discriminated against a lot in the past, the more they believe that society is meritocratic for all, the lower their self-esteem (Foster et al., Reference Foster, Sloto and Ruby2006; Major et al., Reference Major, Kaiser, O'Brien and McCoy2007). Similar experimental effects on their well-being (i.e., mood, life satisfaction, and self-esteem) are found for women and ethnic minorities who are either induced to perceive more discrimination (Major et al., Reference Major, Kaiser, O'Brien and McCoy2007) or who are the clear targets of discrimination (Foster and Tsarfati, Reference Foster and Tsarfati2005). Thus, devalued group members who experience discrimination have lower well-being if they have stronger beliefs that rewards, such as jobs and pay, go to the deserving. This might be because perceiving discrimination is in conflict with, and serves as a threat to, the worldview of those who believe that the world is a just place and that hard work is rewarded (Major et al., Reference Major, Kaiser, O'Brien and McCoy2007).

Primary Appraisals Summary

I argue that some known moderators of the effects of stigmatization operate through the primary appraisal process, that is, they make people perceive stigmatization as more self-relevant (i.e., identification with the ingroup, or with a negatively stereotyped domain) and/or as more harmful (i.e., chronic beliefs about stigmatization, or about meritocracy), which leads to greater feelings of threat, experiences of stress, and consequently negative effects for health, well-being, and performance. However, there are also important differences in how these processes might operate depending on the moderator. For instance, people who chronically believe that they or their group face a great deal of stigmatization likely find ambiguous negative behaviors to be more stressful than those less sensitive to stigmatization, perhaps because the former are more likely to interpret the event as discrimination. In contrast, people who believe that meritocracy exists likely find clear-cut, unambiguous experiences of stigmatization to be more threatening than those who believe that meritocracy does not exist, perhaps because the former experience a worldview threat that the latter do not. Finally, for some of the identified moderators there is solid evidence for the suggested mediated paths laid out in Figure 1 (i.e., ingroup identification), for some there is partial evidence (i.e., identification with the domain, anticipation of stigma), and for some, to date, no empirical evidence exists (i.e., the belief that meritocracy exists).

Psychological Resources

One's disposition is a type of psychological resource that can aid resilience to stigmatization. Among people with a weaker sense of control, lower self-esteem, and less optimism, more personal experience with discrimination is associated with lower well-being and more psychological symptoms. However, among those with a stronger sense of control, higher self-esteem, and greater optimism, the effect of experienced discrimination on mental health (Dion et al., Reference Dion, Dion and Pak1992) and well-being (Jang et al., Reference Jang, Chiriboga and Small2008) is greatly diminished. Thus, the negative effects of stigmatization are mitigated for those with a greater sense of control, self-esteem, and optimism.

Evidence suggests that people with greater psychological resources appraise stigmatization to be less threatening and themselves as better able to cope. For instance, very different reactions are found among women who read about sexism toward University students and alumni based on their general levels of optimism (Kaiser et al., Reference Kaiser, Major and McCoy2004). Compared with those lower in optimism, those higher in optimism appraised sexism as less threatening and themselves as better able to meet the challenges of sexism, which led to better emotional well-being (Kaiser et al., Reference Kaiser, Major and McCoy2004). Experimental manipulations can also affect these processes. For instance experimentally inducing participants to self-affirm, which raises self-esteem, can reduce the negative effects of stereotype threat on scholastic performance (Cohen et al., Reference Cohen, Garcia, Purdie-Vaughns, Apfel and Bruzustoski2009). Thus, it appears that having a stronger sense of control, higher self-esteem, and greater optimism leads people to see themselves as possessing the resources to deal with the challenge of stigmatizations making it less stressful. Consequently, stigmatization has less aversive effects on mental health and well-being.

Personal Resources

Personal resources are those attributes of an individual (e.g., SES) that can be drawn on to cope with stressors and should help foster resilience to stigmatization. Devalued group members with higher incomes, education, and occupational status should be buffered from stigmatization, compared with those with lower SES. There is evidence that the negative effects of discrimination on mental health are stronger for those with a lower level of educational attainment or lower income and weaker for those with a higher level of educational attainment or higher income (Kessler et al., Reference Kessler, Mickelson and Williams1999). This is true even though devalued group members with higher SES report experiencing more discrimination that those with lower SES (Kessler et al., Reference Kessler, Mickelson and Williams1999). For instance, a study of Hispanic employees found that the more they experience discrimination at work, the worse their job satisfaction and the higher their work tension; however, these effects were weaker for those with higher salary levels (Sanchez and Brock, Reference Sanchez and Brock1996). Thus, it appears that the effects of stigmatization on mental health and well-being are moderated by SES, but why?

Devalued group members with higher SES should be in a better position to respond to threats because they can harness economic and social resources to cope with stigmatization, compared with those with lower SES (Hall and Taylor, Reference Hall, Taylor, Hall and Lamont2009). People with higher SES might have more control and flexibility in how they respond to stigmatization. In addition, people with higher SES also have more psychological resources (e.g., a greater sense of control and higher self-esteem) that help to buffer the impact of stigmatization (Dion et al., Reference Dion, Dion and Pak1992). This could be because people with higher SES have jobs that involve more direction and control, so consequently they develop a greater sense of control (Link et al., Reference Link, Lennon and Dohrenwend1993). Alternatively, having a higher SES could provide alternate means of recognition for devalued group members (Lamont et al., Reference Lamont, Fleming, Welburn, Hall and LamontForthcoming). Thus, there are multiple routes through which incomes, education, and occupational status help to buffer devalued group members from the damaging effects of stigmatization.

Social Resources

Social resources are those attributes of a collective that can be drawn on to cope with stressors (e.g., group identity, networks, cultural repertoires, etc.) and should help foster resilience to stigmatization. To the extent that they are valued and a source of identification, people's ingroups can be social resources (Correl and Park, Reference Correl and Park2005), which can facilitate resilience to stigmatization. First, let us consider how having positive evaluations of one's ingroup can be an important resource for coping with stigmatization.Footnote ⁵ Among devalued group members, having more positive evaluations of the ingroup is associated with lower levels of depression and anxiety (Cassidy et al., Reference Cassidy, O'Connor, Howe and Warden2004). Furthermore, the negative effects of discrimination on mental health are mitigated for devalued group members who feel more positively toward their ethnic group (Rivas-Drake et al., Reference Rivas-Drake, Hughes and Way2008). Thus, the negative effects of stigmatization on mental health and well-being are lessened for those who positively evaluate their ingroup.

Importantly, there is evidence that these effects operate through stress processes (Neblett et al., Reference Neblett, White, Ford, Phillip, Nguyen and Sellers2008). For instance, the more mentally ill people value their ingroup (e.g., believe that the mentally ill are good), the more they perceive themselves as able to cope with stigma (Rusch et al., Reference Rusch, Corrigan, Wassel, Michaels, Olschewski, Wilkniss and Batia2009). This could be because when the self has been threatened (e.g., due to stigmatization), the self can be buffered by drawing on positive identities (Steele Reference Steele and Berkowitz1988), and valued group memberships provide people with a sense of self-worth (Tajfel and Turner, Reference Tajfel and Turner1986). Thus, devalued group members who value their ingroup should be more likely to see themselves as having the resources to cope with the challenge that stigmatization presents and consequently experience less stress and strain on mental health and well-being.

Second, let us consider how ingroup identification can be an important resource for coping with stigmatization. There is some evidence that the damaging effects of stigmatization are weaker for devalued group members who have a strong ingroup identification (Wong et al., Reference Wong, Eccles and Sameroff2003).Footnote ⁶ For instance, among Filipino Americans, the more they experience discrimination, the more depressive symptoms they have, but this effect is mitigated for those with a stronger group identification (Mossakowski Reference Mossakowski2003). In addition, much research has shown that perceiving more discrimination against one's group leads to greater ingroup identification, which in turn buffers negative responses to perceived discrimination, such as lowered self-esteem (Branscombe et al., Reference Branscombe, Schmitt and Harvey1999; Schmitt and Branscombe, Reference Schmitt and Branscombe2002b). Thus, either as a moderator or a mediator, ingroup identification can buffer the negative effects of stigmatization on mental health and well-being.Footnote ⁷

How might ingroup identification be a coping resource? Multiple potential processes might facilitate coping for devalued group members. First, group identification might promote access to cultural repertoires that provide alternate sources of group recognition (Lamont and Mizrachi, Reference Lamont and MizrachiForthcoming). Indeed, among African American youth, the typical effect of experiences of discrimination on felt stress is entirely mitigated for those who were frequently socialized by their parents to value Black culture, to be aware of discrimination, and to believe in racial equality (Neblett et al., Reference Neblett, White, Ford, Phillip, Nguyen and Sellers2008). Second, group identification might play such a beneficial role because it creates a sense of belonging or acceptance, which is a critically important human motive (Baumeister and Leary, Reference Baumeister and Leary1995). Third, ingroup identification might go hand in hand with having a better social network to draw on: the social support and material support offered by ingroup members can provide important coping resources (Schmitt and Branscombe, Reference Schmitt and Branscombe2002b). Social support from ingroup members can alleviate the negative effects of stigmatization on physiological stress responses (McNeilly et al., Reference McNeilly, Robinson, Anderson, Pieper, Shah, Toth, Martin, Jackson, Saulter, White, Kuchibatla, Collado and Gerin1995) and on psychological stress responses (Utsey et al., Reference Utsey, Ponterotto, Reynolds and Cancelli2000; but see Brondolo et al., Reference Brondolo, Halen, Pencille, Beatty and Contrada2009). In addition, being active in group-based organizations mitigates the negative effects of discrimination on well-being (Ramirez-Valles et al., Reference Ramirez-Valles, Fergus, Reisen, Poppen and Zea2005). Importantly, there is evidence that those who identify more with their ingroup are more likely to appraise themselves as having the resources they need to cope with stigma, compared with those who are less identified (Rusch et al., Reference Rusch, Corrigan, Wassel, Michaels, Olschewski, Wilkniss and Batia2009). Thus, for those with a stronger ingroup identification, stigmatization is less stressful and has less adverse consequences because devalued group members perceive themselves as having the resources needed to cope.

Secondary Appraisals Summary

In summary, I suggest that some moderators of the effects of stigmatization (namely, optimism, sense of control, self-esteem, high SES, evaluations of the ingroup, identification with the ingroup) operate through the secondary appraisal process. For those with higher levels of personal and social resources, sometimes the effect of stigmatization on outcomes is weakened and sometimes it is eliminated. This suggests that having more resources helps when facing challenges even if they are not sufficient to outweigh potential threats. For all these moderators, having greater resources makes people more likely to perceive stigmatization as a challenge rather than a threat; thus less stress is experienced; consequently, negative effects for health, well-being, and performance should be lessened. Some of these moderators might operate in a similar fashion. For example, optimism, a sense of control, and income might all affect people's beliefs about whether they have the resources they need to contend with stigmatization and whether they can do so effectively. However, other moderators might operate differently and through multiple mechanisms. For instance, ingroup identification might provide a cultural tool kit for understanding and responding to stigmatization (Swidler Reference Swidler1986), or access to social and material resources (e.g., social networks) that help devalued group members cope with the daily challenges created by stigmatization (Fleming et al., Reference Fleming, Lamont and WelburnForthcoming). Although there is consistent evidence that the proposed psychological, personal, and social resources outlined here affect secondary appraisals of ability to cope with stigmatization, the exact means through which people use such resources to cope are unclear (Fig. 2).