Stigmatization of War Veterans with Posttraumatic Stress Disorder (PTSD): Stereotyping and Social Distance Findings

Abstract

Posttraumatic stress disorder (PTSD) affects a significant portion of the US population, but there remains limited information on public responses to affected individuals. Diagnosed mental illnesses can lead to negative stereotyping by the public, who can then socially exclude or otherwise discriminate. This paper presents results of an experiment (N = 830) that assessed the extent to which workers with PTSD labels—either resulting from an auto accident or wartime military service—evoked negative stereotypes in a workplace scenario and social distance from study participants. Results demonstrated that PTSD from wartime service evoked perceptions of dangerousness and incompetence (similar to other severe mental illness conditions) but did not evoke worse perceptions of assertiveness (a requisite managerial trait). The condition also led to levels of social exclusion similar to some other mental health labels. These findings provide insight into the context of stereotyping and social distance for PTSD relative to other conditions.

Keywords

stigma posttraumatic stress disorder stereotyping

People with mental illness labels (PWMI) often struggle to manage symptoms associated with their illness and also have to manage negative treatment from the public in the form of stigmatization. This is a process whereby disclosed labels become associated with negative stereotypes, leading to discrimination (e.g., Link and Phelan 2001). Research has documented this process for people with various conditions, finding, for example, that people with schizophrenia are thought to be dangerous and face social exclusion (Link et al. 1999). Another subset of the population of PWMI are those individuals suffering symptoms consistent with posttraumatic stress disorder (PTSD). This condition affects approximately 10 percent of the US population over their lifetime (Kessler et al. 2012) and approximately 14 percent to 16 percent of deployed US military personnel (Gates et al. 2012). PTSD has garnered much media attention due to its link to the wars in Iraq and Afghanistan. It has even been cited as a “signature injury” of the wars (Kleykamp and Hipes 2015; Tanielian and Jaycox 2008). That being said, the condition can result from scenarios outside of traditional combat settings, such as military sexual assault or accidents (Eid 2003; Wilson forthcoming).

Noticeably absent from media and academic investigations into this condition is a consideration of how the public responds to PTSD labels. For example, there is limited information on how employers react to the knowledge that a job applicant or employee has PTSD. Veterans possess symbolic capital in the United States, which is the prestige given to a social group derived from their cultural value. For modern US veterans, this comes largely from the cultural desire to “support the troops” (Calhoun 2002; MacLean and Kleykamp 2014). Reflecting the high levels of support for military personnel and veterans, employment of veterans of the wars in Iraq and Afghanistan has been a prime concern at the highest levels of government. This was indicated by President Obama’s statement when signing the VOW to Hire Heroes Act, in which he claimed, “I proposed an even bigger tax break if a business hires a veteran who’s been unemployed for at least six months. And if a business hires an unemployed veteran with a disability related to their service, I proposed doubling the tax break we already have in place” (Obama 2011). But alongside portrayals of veterans and military personnel as deserving of public support are portrayals of veterans as “damaged” or otherwise victimized by their service (Kleykamp and Hipes 2015). Thus, it is unclear how PTSD compares to other mental illness labels. It could theoretically be linked with stigmatization consistent with mental illness or to public sympathy via the sacrifices made in service to the United States. Or both.

Many veterans fear that they will be perceived of as dangerous or otherwise troubled due to the stereotypes associated with their service (Kleykamp and Hipes 2015). This leads to the following research questions:

Research Question 1: How does the stereotyping and social distance elicited by a PTSD label compare to other mental illness labels?

Research Question 2: Does the public negatively stereotype people with PTSD similarly to other PWMI, or do they instead treat them more preferentially?

Research Question 3: Does the public discriminate against people with PTSD by desiring social distance from them?

The research presented here assessed stereotyping and discrimination toward people with PTSD to ascertain whether stigma toward this population exists and what form it takes. Measures focused on dangerousness and competence as these have been identified as key to mental illness stigma (Sadler, Meagor, and Kaye 2012). Because this study is focused on the labor market, there is also a dependent measure for a managerial trait (assertiveness), derived from a list used to assess workplace stereotypes (Schein 1973). This research consisted of an experiment that assessed the degree of stereotyping and discrimination directed toward a worker with a mental illness label who is otherwise presented as having a successful work history. An experiment examined the perceived characteristics of workers—both women and men—with five different mental illness labels: PTSD (resulting from an automobile accident or service during Operation Iraqi Freedom), an unspecified “mental illness,” schizophrenia, and depression.

Background

Experiencing trauma can have a lasting effect by inducing debilitating mental health symptoms. This was first codified for mental health professionals with the inclusion of PTSD in the Diagnostic and Statistical Manual of Mental Disorders III in 1980 (Scott 1990). Since then, the condition has been associated with military veterans, who experience high-stress conditions as a part of their work, including but not limited to wartime deployments and combat situations. Veterans who return to civilian life often face difficulties reintegrating, including maintaining healthy relationships with family and gaining meaningful employment (Kleykamp 2013; Lester and Flake 2013).

In the context of mental health, PTSD is a relatively unique diagnosis. First, it is seen as externally caused by trauma, which is explicitly linked in the name of the condition. This positions the condition as distinct from other diagnoses, which the public tends to view as genetic or biologically based (Conrad and Schneider 2010). Second, PTSD is a socially constructed condition whose origins were largely a result of a movement of Vietnam veterans, advocates, and anti-war psychiatrists such as Robert Lifton. These individuals were either directly experiencing the condition or familiar with Vietnam veterans’ mental health needs (Scott 1990). While mental illness labels are socially constructed (e.g., Horwitz 2002), PTSD stands out as having symptoms and causes and origins that are especially dependent on social context.

The Stigma Process and Predictions

People with disclosed mental illness labels can face negative treatment due simply to preconceived ideas of what their conditions entail (e.g., Hipes et al. 2016). Theories on stigmatization and stereotyping help explain how mental illness labels lead to unequal outcomes, including employment discrimination and social rejection. Stigma is a social phenomenon that results in discrimination via the following process: (1) People create distinctions based on characteristics and assign labels; (2) cultural norms dictate that certain characteristics are undesirable; (3) labeled individuals are set apart from nonlabeled individuals, creating an us/them division; (4) people create justifications for stigmatizing labeled individuals based on their perceived differentness; and (5) a power differential exists such that a more powerful group can stigmatize a labeled, devalued group (Link and Phelan 2001). This model shows how labels can become associated with stereotypes and lead to discrimination, such as social exclusion.

The first step in the stigma process is “distinguishing and labeling human differences” (Link and Phelan 2001:367). For PWMI, this process often occurs through the transmission of knowledge about a mental illness diagnosis as these seemingly discrete categorizations are often how mental health status is conveyed by medical professionals to patients. Once PWMI have been labeled as having a condition—and thus received a label that separates them from “normal” individuals—their mental illness labels can become associated with negative attributes or stereotypes (Link and Phelan 2001). Research has demonstrated that mental illness labels evoke damaging stereotypes such as incompetence, dangerousness, uncleanliness, worthlessness, and unpredictability (Link et al. 1999; Phelan and Link 2004). The dangerousness stereotype particularly remains a powerful tool leading to social exclusion and negative treatment of PWMI. From the long history of research demonstrating stigma toward PWMI, the overarching study prediction that situates PTSD alongside other labels is:

Hypothesis 1 (stigma of mental illness): (a) Study participants will negatively stereotype workers with mental illness labels. (b) Participants will desire social distance from workers with mental illness labels.

Public reactions to people with mental illness can be more severe when the person has a label of “mental illness” and is seen as more dangerous by the public (Martin, Pescosolido, and Tuch 2000). In an overview of studies, Angermeyer and Dietrich (2006) found that the public saw people with mental illness as especially dependent on others, unpredictable, and violent/dangerous. And alongside portrayals of Operation Iraqi Freedom veterans as deserving of sympathy have also been fears of veterans being depicted as insane or volatile (Kleykamp and Hipes 2015). While these portrayals are more infrequent than other types of stories about veterans, they could be more attention grabbing to casual viewers of media. We therefore predict that:

Hypothesis 2 (dangerousness stereotype): Participants will perceive of PWMI as more dangerous than nonlabeled individuals (a control condition). This will be especially true for conditions commonly linked with perceived violence and dangerousness.

Perceptions of competence differ based on the specific condition, but the public stereotypes PWMI as less competent than nonlabeled individuals (Sadler et al. 2012). Hipes, Lucas, and Kleykamp (2014) showed that war veterans with PTSD received lower status when working on a group task with a partner, which is one measure of perceived competence. Other research has shown that people with certain mental illness labels evoked perceptions of lowered competence compared to nonstigmatized individuals (e.g., Sadler et al. 2012). In addition to being a mental illness stereotype, competence is also an important component of what makes someone “managerial” (Schein 1973). This is thus an indicator of status in a workplace setting.

Hypothesis 3 (incompetence stereotype): Participants will perceive of PWMI as less competent than nonlabeled individuals (a control condition).

Social Responses to People with PTSD

Relative to other conditions, some research suggests that the public expresses sympathy toward people with PTSD in some scenarios. In one study, Vietnam War Syndrome (an earlier term similar to what is now called PTSD) elicited more pity than other mental-behavioral conditions. It was also seen as less controllable than other mental conditions (Weiner, Perry, and Magnusson 1988). In the meaningful setting of courtroom sentencing, a more recent study found that prosecutors were significantly more lenient toward veterans both with and without PTSD diagnoses compared to nonveterans. Furthermore, they viewed veterans as less at fault for their crimes and empathized more with them compared to nonveterans (Wilson et al. 2011). There is evidence, then, that PTSD labels elicit forms of sympathy from the public.

MacLean and Kleykamp (2014) investigated public responses to military veterans presented as having seen combat, noncombat veterans, and military contractors. They developed an extension of the original stigma model (Link and Phelan 2001) by including symbolic capital as a moderator/modifier that can alter the effects of negative stereotypes on discrimination. Results of their study demonstrated support for their hypothesis, such that the public saw combat veterans as more likely to have mental/behavioral issues than the other groups studied (a stigmatizing characteristic) but did not seek social distance from the group (perhaps due to their symbolic capital). Thus, labels (combat veteran) that evoked negative stereotypes (mental/behavioral issues) did not lead to discrimination (social exclusion). Their conceptual model is presented in Figure 1.

Figure 1.

Modified stigma model (MacLean and Kleykamp 2014:134).

As illustrated in this model, the stigma process need not necessarily lead to discrimination. This is true especially for those individuals whose stigmas are externally caused (e.g., by a traumatic event) and who possess high power and high symbolic capital (e.g., for military personnel and veterans; Maclean and Kleykamp 2014). One study has found greater sympathy for “Vietnam War syndrome” due to its war-related cause (Weiner et al. 1988). Following from the symbolic capital that military personnel and veterans have, it is expected that war veterans with PTSD will be subjected to less social exclusion than other types of mental illness. In the case of this paper, it is expected that war-related PTSD will lead to different levels of social distance than other mental illnesses due to the cultural support for military personnel and veterans. Following from this process, predictions are as follows.

Symbolic Capital Predictions

Hypothesis 4: Participants will desire more social distance from a worker with mental illness than a control condition.

Hypothesis 5: Participants will not desire more social distance from a war veteran with PTSD than a control condition.

Unlike for PTSD, research has yet to indicate that people with other mental illnesses evoke cultural sympathy relative to other stigmatized groups. For conditions such as schizophrenia, the public is more likely to believe that symptoms do notlie on a continuum but are more black and white (Schomerus, Matschinger, and Angermeyer 2013).¹ Therefore, people with schizophrenia can face detachment from society due to their perceived differentness, while war veterans with mental illness have been theorized to be less likely to face rejection due to their symbolic capital (MacLean and Kleykamp 2014).

Media coverage has documented the difficulties that individuals with PTSD can experience—or expect to experience—in the workplace, including the relationship between the condition and unemployment (Anderson 2014; Jason, Mileviciute, and Aaso 2011). For this reason, this study tests workplace-related stereotypes of people with PTSD, either from an automobile accident or wartime service, in terms of how dangerous, competent, and assertive participants perceive them to be. The workplace setting provides more meaningful measurements of traits due to the skills and characteristics required to succeed in a competitive environment. It can be a setting where individuals find meaning, earn money, and establish valuable social networks. Due to the stigma of mental illness and the documented difficulties for people with PTSD in the workplace, we expect that:

Hypothesis 6: Individuals with war-related PTSD will be seen as incompetent compared to nonlabeled individuals.

Hypothesis 7: Individuals with war-related PTSD will be seen as less managerial compared to nonlabeled individuals.

Methodology

Institutional Review Board approval at a major state university was applied for and obtained for this research. To test the predictions, experimental methods were employed in which participants read 1 of 12 randomly selected vignettes (conditions listed in the following) of a worker with a successful history who was said to have missed work time to attend to a mental health hospitalization (or missed no work time/had no label for the control condition). Conditions are:

Condition 1: Control (has not missed work time for a medical condition), woman

Condition 2: Mental illness (unspecified), woman

Condition 3: PTSD resulting from auto accident, woman

Condition 4: PTSD resulting from service during Operation Iraqi Freedom, woman

Condition 5: Schizophrenia, woman

Condition 6: Depression, woman

Condition 7: Control (has not missed work time for a medical condition), man

Condition 8: Mental illness (unspecified), man

Condition 9: PTSD resulting from auto accident, man

Condition 10: PTSD resulting from service during Operation Iraqi Freedom, man

Condition 11: Schizophrenia, man

Condition 12: Depression, man

Once participants had read their scenario, they judged whether they believed a series of traits were characteristic of the “type” of person they had read about in the scenario (see Appendix for a list of these traits). These attributes included stereotypes of dangerousness and incompetence as well as the trait of “assertiveness,” which an analysis indicated to be a key factor from a list of requisite managerial traits (Schein 1973). Finally, participants responded to questions assessing their personal sense of power (Anderson and Keltner 2012) and past contact with mental illness and indicated their desired level of social distance from the vignette subject. The Appendix provides some materials, and all remaining study materials are available on request.

Sample

One thousand and seventy participants from Amazon’s Mechanical Turk system responded to a Qualtrics survey intended to take between 10 to 15 minutes of time, in exchange for $0.75. Mechanical Turk is a tool to connect researchers with participants to carry out various types of studies that can be completed online. Participants agree to work on Human Intelligence Tasks (called HITs) in exchange for money, and researchers post their HITs to be completed. Many HITs can be completed within a few hours or even minutes, provided that pay is sufficient to attract participants. Thus, mTurk allows for extremely fast data collection relative to laboratory studies by streamlining participant recruitment and utilizing a massive subject pool. Research has demonstrated that mTurk’s subject pool is more diverse than traditional university samples and that data using mTurk samples is as reliable as traditional methods (Buhrmester, Kwang, and Gosling 2011). Nevertheless, just as with any sampling technique, recruitment techniques can ultimately impact the data (Paolacci and Chandler 2014). Only US participants were sampled to account for different concepts of the labor market outside of the United States.

The sample for this study was less racially diverse and more highly educated than the general population (as is typical of mTurk samples), but due to these factors, more conservative estimates are expected on the amount of stigma toward people with PTSD. For education level specifically, Corrigan and Watson (2007) asked participants to respond to a vignette of someone with either schizophrenia, drug dependence, or emphysema. They found that more highly educated participants were less likely to endorse stigma for both a person with mental illness and a person with drug dependence. This supports previous research (Pescosolido et al. 1999) that also found lower stigmatization of mental illness among more highly educated people.

In an attempt to account for racial differences in stigma, names of the vignette subjects were selected to be relatively race-neutral using data on name popularity by race, thus helping to ensure that the mental illness labels were the most salient aspect of the scenarios. After dropping any participants who did not complete the study or correctly answer questions designed as “attention checks,” the final sample included in analyses was 830 participants.

Table 1 presents the descriptive statistics for the sample demographics.

Table 1.

Sample Demographics.

Variable	N	Percentage
Sex
Male	462	55.7
Female	368	44.3
Other	1	0
Education level
High school or less	81	9.8
Some college	316	38
Four-year college+	433	52
Race
White/Caucasian	608	73.3
Black/African American	50	6
Asian	73	8.8
Other/multiracial/unspecified	99	11.9
Age (mean)	34	Range (18–71)
Total	830	100

Independent Variables

Demographic measures included participant sex, education level, age, and race. Education level was grouped into the categories of high school or less, some college/trade school, and four-year college degree or higher. Control variables for this study include a measure of the sense of power that participants feel in their daily lives. A sense of power has been associated with less stereotyping of others in at least one study (Lammers, Stoker, and Stapel 2009) and was included as an important control measure. Contact with mental illness in various arenas is another important control measure, with more contact typically associated with less stigma (Alexander and Link 2003).² We tested for interaction effects with these key control variables to see if they act as modifiers, and these tests did not produce significant findings. Because the effect of power or contact do not interact with vignette condition, they each affect vignette condition independently of the experimental condition. These results are not included in the paper but are available on request.

Dependent Variables

Study participants responded to a series of traits judging the vignette subject (see Appendix). To assess study predictions, we selected 49 traits from the list presented to participants. These traits were expected to cluster around the stereotypes of “dangerousness” and “competence,” thus traits that were synonyms were included. With this list of 49 traits, we conducted a factor analysis using orthogonal varimax rotation (Statistics Corner 2009; Vogt 1993).

To include another dependent measure related to general employability for the vignette subject, we also ran a separate factor analysis (using identical methods as previously described) of 34 of Schein’s (1973) original requisite managerial traits (list available on request). These characteristics were selected to not overlap with the constructs of dangerousness or competence. This analysis yielded two factors that together accounted for 70.9 percent of the variance among the items. However, one factor that captured “capability” is not included in analyses due to overlap with the competence factor. As an example, the capability trait loaded strongly on the competence factor and vice versa. The remaining factor was “assertiveness,” which is analyzed in the findings. Indices were created from these analyses and served as the main dependent variables assessing stereotyping. See Table 2 for these indices.

Table 2.

Dependent Variable Index Components.

Trait Variable	Loading
Competence
Consistent	.51
Capable	.74
Leadership	.73
Intelligent	.69
Skilled	.79
Analytical	.54
Competent	.78
Well-informed	.52
Qualified	.73
Dangerousness
Violent	.78
Threatening	.8
Devious	.53
Dangerous	.75
Vulgar	.54
Quarrelsome	.5
Harmful	.66
Assertiveness
Assertive	.71
Self-confident	.6
Authoritative	.68
Direct	.61
Decisive	.52
Firm	.64

Social distance is the main discrimination measure in this study. To test for differences in desired social distance from PWMI, four social distance items were combined into a scale ranging from 1 to 4, following from Link et al. (1987). The scale reliability coefficient for these items was .86, indicating high reliability. The mean of this scale was 2.73, and standard deviation was .73.

Analysis

Regression analyses were conducted to assess the hypotheses. Three models were ordinary least squares (OLS) regressions (assessing dependent variables of incompetence, assertiveness, and social distance), with the dangerousness model run as a logistic regression. Responses to the traits that made up the dangerousness factor were highly skewed. Due to the non-normality, a logistic regression was done instead of an OLS model. Both OLS and logit models produced similar results, and the OLS model is available on request. Interaction effects of variables thought to act as modifiers are included in some models (e.g., dangerousness interacting with condition to affect social distance), with some further models of interaction effects available on request.

Illness label and gender of the vignette subject were entered as separate variables, thus allowing for the estimation of interaction effects. The illness label variables include both women and men vignette subjects, and the vignette gender variable divides the 12 conditions in two. For example, the schizophrenia illness label variable includes all participants who read either a male or female vignette with schizophrenia as the illness label in the scenario. And for vignette gender, the male variable includes all six conditions that had a male person in the scenario. Models that compare all 12 conditions (instead of breaking down by illness label and vignette gender) are available on request and produce similar results.

We also confirmed these results with a structural equation model (SEM; results available on request). The SEM incorporated all of the index variables as their individual components and otherwise mirrored the regression analyses. The SEM model appears to be a reasonable fit, based on accepted goodness-of-fit measures (e.g., standardized root mean square residual, comparative fit index; Hooper, Coughlan, and Mullen 2008). For this paper, we prefer regression models as the results are easier to interpret.

Results

To address the research questions that explored the mechanisms associated with employment discrimination for people with PTSD and other PWMI, analyses first assessed whether people with different mental illness labels are negatively stereotyped relative to someone with no label. The stereotypes examined were dangerousness, incompetence, and unassertiveness.

Stereotyping of PWMI as Dangerous

Public perceptions that PWMI are more dangerous than nonlabeled individuals is perhaps one of the most damaging stereotypes, with repercussions for employment chances. PWMI must be expected to safely interact with, supervise, and coordinate with fellow co-workers. Analyses measured public perceptions of dangerousness for the PTSD conditions and other illness labels examined in this study, with all the people said to be workers with successful employment histories (see Appendix for study instruments).

A logistic regression of perceived dangerousness on the experimental conditions, contact with mental illness, sense of power, and control variables provided evidence of public responses to people with PTSD (Table 3). Perceived dangerousness was captured by participants who selected that the trait was either characteristic or somewhat characteristic of the vignette subject.

Table 3.

Logistic Regression Predicting the Log Odds of Participants Perceiving the Vignette Subject as Dangerous Based on Experimental Conditions, Sense of Power, Contact with Mental Illness, and Control Variables.

Variable	Odds Ratio	B	(SE of Odds Ratios)
Illness labels (reference = no illness label)
Mental illness	1.73	.55	(.61)
Posttraumatic stress disorder (PTSD; auto accident)	.83	−.19	(.31)
PTSD (Iraq War)	2.39***	.87	(.83)
Schizophrenia	2.52***	.92	(.92)
Depression	1.21	.19	(.46)
Vignette gender (reference = man)	.31*	−1.17	(.15)
Woman vignette ×
Mental illness	2.09	.74	(1.24)
PTSD (auto accident)	3.85	1.35	(2.3)
PTSD (Iraq War)	4.10	1.41	(2.36)
Schizophrenia	2.49	.91	(1.48)
Depression	1.23	.21	(.80)
Sense of power	.76***	−.27	(.05)
Contact with mental illness	.90	−.1	(.07)
Sex (reference = male)	.73**	−.31	(.12)
Age	.99	−.01	(.01)
Race (reference = white)
Black/African American	.89	−.12	(.31)
Asian	1.76	.57	(.49)
Other race	1.5	.41	(.36)
Education (reference = four-year degree+)
High-school or less	.51**	−.67	(.15)
Some college/trade school	.93	−.07	(.16)
Constant	.97	−.03	(1.95)
Likelihood ratio χ²	98.3***
Pseudo R²	.09
N	830

p < .05. **p < .01. ***p < .001 (two-tailed test).

Confirming Hypothesis 2 for two conditions relative to the control, results demonstrate that participants perceived of workers who were war veterans with PTSD as significantly more dangerous than workers with no stated condition (p = .012). Participants viewed the worker with PTSD from an automobile accident as no more or less dangerous than the control condition (p = .615). Similar to the war veteran condition, workers with schizophrenia also evoked perceptions of dangerousness (p = .012). These findings confirm Hypothesis 1, which predicted a general stigmatization of people with mental illness, as well as past research that has demonstrated negative stereotyping of people with schizophrenia as dangerous. Thus, war veterans with PTSD are stereotyped by the public as more dangerous than someone with no known mental illness.

When the vignette subject was a woman, participants perceived of her as less dangerous compared to the male conditions, which is expected because dangerousness is a stereotypically masculine trait. Also, participants who feel more powerful tended to see the vignette subject as less dangerous across conditions (p = .000). Finally, participants with no more than a high school education were less stigmatizing of the vignette subjects (p = .022) across conditions.

Stereotyping of War Veterans with PTSD as Incompetent

Another pervasive stereotype of people with PTSD and other PWMI that disadvantages them in the labor market is perceived incompetence (Sadler et al. 2012). The workers in the scenarios are said to have a successful employment history and thus should evoke perceptions of competence unless mental illness labels counteract this. This model was run as an OLS model.³ Results are presented in Table 4.

Table 4.

Ordinary Least Squares Regression of Perceived Incompetence on Experimental Conditions, Sense of Power, Contact with Mental Illness, and Control Variables.

Variable	Coefficients	(SE)
Illness labels (reference = no illness label)
Mental illness	.29**	(.090)
Posttraumatic stress disorder (PTSD; auto accident)	.09	(.090)
PTSD (Iraq War)	.20*	(.090)
Schizophrenia	.53***	(.100)
Depression	.23*	(.100)
Vignette gender (reference = man)	−.19*	(.090)
Woman vignette ×
Mental illness	.04	(.130)
PTSD (auto accident)	.17	(.130)
PTSD (Iraq War)	.24	(.130)
Schizophrenia	−.07	(.140)
Depression	.03	(.140)
Sense of power	−.11***	(.020)
Contact with mental illness	−.07***	(.020)
Sex (reference = male)	−.14**	(.020)
Age	.00	(.002)
Race (reference = white)
Black/African American	−.02	(.080)
Asian	.06	(.070)
Other race	.03	(.06)
Education (reference = four-year degree+)
High school or less	−.12	(.070)
Some college/trade school	.00	(.040)
Constant	2.42	(.130)
F statistic	8.01***
R²	.17
N	830

p < .05. **p < .01. ***p < .001 (two-tailed test).

Results indicate that mental illness labels evoke harmful stereotypes that the person is incompetent, relative to a control condition. This confirms Hypothesis 1. This also largely confirms Hypothesis 3, which predicted that mental illness labels would elicit perceptions of incompetence relative to a control. For all conditions except the person with PTSD from an automobile accident, participants perceived of the worker as significantly more incompetent than a worker with no label. The war veteran with PTSD evoked perceptions of incompetence (p = .028), confirming Hypothesis 7, while the person with PTSD from an auto accident did not differ significantly from the control condition (p = .306). None of the interaction effects between the illness labels and vignette gender were significant. Similar to the dangerousness model, women participants were less stigmatizing than male participants, and a higher sense of power was associated with less stereotypical responses. In this model, contact with mental illness was significantly associated with less stereotyping across conditions.

Stereotyping of People with PTSD as Less Assertive

This research attempts to connect previous knowledge on PTSD and mental illness stereotypes with the labor market by illustrating how people with PTSD are seen not only as dangerous and incompetent but also less managerial than nonlabeled individuals. As mentioned previously, a factor analysis identified assertiveness as a managerial trait, derived from Schein’s (1973) formulation. To assess participants’ perceptions of the vignette subjects as unassertive, an OLS regression was estimated on control variables, the experimental conditions, sense of power, and contact with mental illness. Results of this analysis are presented in Table 5.

Table 5.

Ordinary Least Squares Regression of Perceived Unassertiveness on Experimental Conditions, Sense of Power, Contact with Mental Illness, and Control Variables.

Variable	Coefficients	(SE)
Illness labels (reference = no illness label)
Mental illness	.41***	(.10)
Posttraumatic stress disorder (PTSD; auto accident)	.24*	(.10)
PTSD (Iraq War)	.03	(.10)
Schizophrenia	.39***	(.11)
Depression	.40***	(.11)
Vignette gender (reference = man)	−.19	(.11)
Woman vignette ×
Mental illness	.00	(.16)
PTSD (auto accident)	.10	(.15)
PTSD (Iraq War)	.27	(.15)
Schizophrenia	.20	(.16)
Depression	.08	(.16)
Sense of power	−.11***	(.02)
Contact with mental illness	−.04*	(.02)
Sex (reference = male)	−.11*	(.05)
Age	−.01*	−
Race (reference = white)
Black/African American	.04	(.10)
Asian	.03	(.08)
Other race	−.14	(.07)
Education (reference = four-year degree+)
High-school or less	−.07	(.08)
Some college/trade school	−.03	(.05)
Constant	2.80	(.16)
F statistic	6.26***
R²	.13
N	830

p < .05. ***p < .001 (two-tailed test).

Results indicate that participants viewed all mental illness conditions as less assertive than the control condition except the PTSD + Iraq War condition. Thus, results do not yield support for Hypothesis 7 in this case, which predicted that war veterans with PTSD would be seen as less assertive than a control condition (p = .086). There was a modest interaction effect, interpreted in the next paragraph. Participants’ sense of power was associated with less stereotyping of the vignette subject as unassertive across conditions (p = .000). Degree of contact with mental illness was associated with significantly less stereotyping of the mental illness conditions (p = .042). These are consistent with findings in the other models. Finally, women participants and older participants were less likely to rate the vignette subjects in stereotypical ways across conditions (p = .026 and p= .016, respectively).

Here, the analysis indicates that both PTSD conditions are the only mental illness conditions that are not negatively stereotyped as incapable relative to the control condition (p = .255 for the auto accident condition and p = .144 for the Iraq War veteran). One of the interaction effects, the PTSD from Operation Iraqi Freedom service × woman effect, is statistically significant and positive (p = .037). This demonstrates that for women relative to men, the effect of a PTSD + wartime service label evokes significantly stronger perceptions of incapability. Similar to the previous managerial model, sense of power, contact, and participant sex are all significant and negative. That is, participants who felt more powerful were less stereotyping of the vignette subjects across conditions, as were participants with more previous contact with mental illness and women participants relative to men.

Social Distance

Social distance is a key stigma outcome reflecting how mental illness labels can lead to social exclusion for people with PTSD and other conditions. In the workplace, an increased desire for social distance could mean less willingness to hire someone due to possible fears of dangerousness or awkwardness. Measures assessed the level of social distance participants desired from the different types of worker. To analyze this variable, a model regressed the social distance scale on the experimental conditions and the same controls used in the previous models (contact with mental illness, sense of power, and the demographic control variables).⁴

Hypothesis 4, consistent with stigma theory, predicted that participants would desire more social distance from a PWMI. However, due to potential public sympathy and symbolic capital, Hypothesis 5 predicted that participants would desire less social distance from a war veteran with PTSD than a control condition. Results are presented in Table 6.

Table 6.

Ordinary Least Squares Regression of Desired Social Distance on Experimental Conditions, Sense of Power, Contact with Mental Illness, and Control Variables.

Variable	Coefficients	(SE)
Illness labels (reference = no illness label)
Mental illness	−.11	(.11)
Posttraumatic stress disorder (PTSD; auto accident)	−.30**	(.11)
PTSD (Iraq War)	−.30**	(.11)
Schizophrenia	−.04	(.12)
Depression	−.11	(.12)
Vignette gender (reference = man)	.10	(.12)
Woman vignette ×
Mental illness	−.28	(.17)
PTSD (auto accident)	.14	(.17)
PTSD (Iraq War)	.06	(.17)
Schizophrenia	−.13	(.18)
Depression	−.30	(.18)
Sense of power	.01	(.02)
Contact with mental illness	−.12***	(.02)
Sex (reference = male)	−.03	(.05)
Age	.00	(.00)
Race (reference = white)
Black/African American	.21*	(.11)
Asian	.05	(.09)
Other race	.18	(.08)
Education (reference = four-year degree+)
High school or less	.11	(.09)
Some college/trade school	.05	(.05)
Constant	2.43	(.17)
F statistic	3.62
R²	.08
N	830

p < .05. **p < .01. ***p < .001 (two-tailed test).

Consistent with Hypothesis 5, findings demonstrate that participants desired significantly less social distance from both PTSD conditions relative to the control condition (p = .009 for the automobile accident condition and p = .008 for the Iraq War condition). None of the other mental illness conditions differed significantly from the control condition. Contact with mental illness was associated with less social distance across conditions (p= .000).⁵ We expect that the tendency of participants to provide socially desirable responses could have led to less stigma for some of the mental illness conditions and discuss this in the conclusion. For this reason, we interpret these findings tentatively here.

Discussion

PTSD, relative to other mental illness labels, can evoke perceptions of dangerousness and incompetence relative to a control condition. Thus, it is a condition associated with negative stereotypes just like some of the other conditions presented in these analyses. And yet, participants did not perceive of this person as less assertive than a control condition. All conditions were associated with significantly less social distance.

Results demonstrated that PTSD, especially when acquired due to wartime military service, can lead to negative stereotyping of an otherwise successful worker as dangerous, incompetent, and less assertive. These findings are the first of our knowledge to demonstrate that war-related PTSD evokes negative stereotyping. And these findings also place the condition in a broader context, including demonstrating how negative stereotyping can apply to both women and men. Nevertheless, participants did not perceive of the veteran with PTSD as less assertive or more worthy of social exclusion.

It is important to note that not all of the stereotypes evoked by the war-related PTSD and other conditions were identical. Rather, there were some differences, such as the war-related PTSD condition evoking lower perceptions of unassertiveness and incompetence relative to the person with schizophrenia. Also, depression appears to be more heavily stigmatized for some of the dependent variables, with the exception of dangerousness.

Conclusion

Stigma theory lays out a process whereby mental illness labels lead to inequality through stereotyping and forms of discrimination such as social exclusion. Social distance is one form of stigma thatcan negatively affect people with mental illness labels, leading to social marginalization and decreased life chances. The research presented in this paper is the first to examine the stigma of PTSD in an employment setting, placing the condition in the context of other more commonly studied conditions. These results represent some of the first findings to compare social responses to PTSD with responses to other mental illnesses, situating the label alongside labels known to be associated with negative stereotypes. Relative to a worker with no label, workers with war-related PTSD were stereotyped in more damaging ways. From research on the stereotype content model (Sadler et al. 2012), this research then places PTSD as a condition that evokes damaging perceptions of dangerousness and incompetence relative to other conditions.

Future research should measure social distance toward military veterans with PTSD in a more unobtrusive way, that is, in a way less likely to evoke social desirability bias. At least one other study has found some social distance toward veterans with PTSD relative to military personnel with no labels (Hipes et al. 2014). As the data stand here, it appears that participants felt a need to demonstrate a lack of stigma toward the vignette subjects with mental illness. Thus, we interpret the social distance findings tentatively.

These studies are experimental and therefore isolate and test specific mechanisms in controlled settings. We expect the various symptoms associated with different mental health conditions to affect how PWMI behave and interact with others. However, for the present studies, mental illness labels are key as mechanisms that can elicit negative stereotyping and negative treatment in interactions. These studies are not intended to shed light on the day-to-day interactions that take place between PWMI and their co-workers, which can have an impact on their treatment in key interactions. Further research might utilize ethnographic methods to assess the in-depth nature of interactions between labeled individuals and co-workers.

Efforts to combat stigma through public education and increasing the “normalcy” of mental illness are rooted in some of the same mechanisms found in this research. Should there be greater public support for PWMI and greater awareness of difficulties that this population faces, perhaps the link between stereotyping and discrimination can be reduced more broadly.

Footnotes

Appendix

Notes

References

Alexander

Laurel

Link

Bruce G.

2003. “The Impact of Contact on Stigmatizing Attitudes toward People with Mental Illness.” Journal of Mental Health 12(3):271–89.

Anderson

Cameron

John

Oliver P.

Keltner

Dacher

. 2012. “The Personal Sense of Power.” Journal of Personality 80(2):313–44.

Anderson

Pauline

. 2014. “Unemployment Biggest Predictor of PTSD Symptom Severity.” Medscape. Retrieved September 25, 2018 (http://www.medscape.com/viewarticle/824747).

Angermeyer

Matthias C.

Dietrich

Sandra

. 2006. “Public Beliefs about and Attitudes towards People with Mental Illness: A Review of Population Studies.” Acta Psychiatrica Scandinavica 113(3):163–79.

Buhrmester

Michael

Kwang

Tracy

Gosling

Samuel D.

2011. “Amazon’s Mechanical Turk: A New Source of Inexpensive, yet High-quality, Data?” Perspectives on Psychological Science 6(1):3–5.

Calhoun

Craig. J

. 2002. “Symbolic Capital.” P, 475 in Dictionary of the Social Sciences, edited by Calhoun

C. J.

New York: Oxford University Press.

Conrad

Peter

Schneider

Joseph W.

2010. Deviance and Medicalization: From Badness to Sickness. Philadelphia: Temple University Press.

Corrigan

Patrick W.

Watson

Amy C.

2007. “The Stigma of Psychiatric Disorders and the Gender, Ethnicity, and Education of the Perceiver.” Community Mental Health Journal 43(5):439–58.

Eid

Jarle

. 2003. “The Course of PTSD Symptoms Following Military Training Accidents and Brief Psychosocial Interventions.” Personality and Individual Differences 35(4):771–83.

10.

Gates

Margaret A.

Holowka

Darren W.

Vasterling

Jennifer J.

Keane

Terence M.

Marx

Brian P.

Rosen

Raymond C.

2012. “Posttraumatic Stress Disorder in Veterans and Military Personnel: Epidemiology, Screening, and Case Recognition.” Psychological Services 9(4):361–82.

11.

Hipes

Crosby

Lucas

Jeffrey

Kleykamp

Meredith

. 2015. “Status-and-stigma-related Consequences of Military Service and PTSD: Evidence from a Laboratory Experiment.” Armed Forces & Society 41:477–95.

12.

Hipes

Crosby

Lucas

Jeffrey

Phelan

Jo C.

White

Rick C.

2016. “The Stigma of Mental Illness in the Labor Market. Social Science Research 56:16–25.

13.

Hooper

Daire

Coughlan

Joseph

Mullen

Michael

. 2008. “Evaluating Model Fit: A Synthesis of the Structural Equation Modeling Literature.” Pp.195–200 in 7th European Conference on Research Methodology for Business and Management Studies. Retrieved September 25, 2018 (https://arrow.dit.ie/cgi/viewcontent.cgi?article=1046&context=buschmancon).

14.

Horwitz

Allan V

. 2002. Creating Mental Illness. Chicago: University of Chicago Press.

15.

Jason

Leonard A.

Mileviciute

Inga

Aase

Darrin M.

2011. “How Type of Treatment and Presence of PTSD Affect Employment, Self-regulation, and Abstinence.” North American Journal of Psychology 13(2):175–86.

16.

Kessler

Ronald C.

Petukhova

Maria

Sampson

Nancy A.

Zaslavsky

Alan M.

Wittchen

Hans-Ullrich

. 2012. “Twelve-month and Lifetime Prevalence and Lifetime Morbid Risk of Anxiety and Mood Disorders in the United States.” International Journal of Methods in Psychiatric Research 21(3):169–84.

17.

Keyes

Corey L. M

. 2002. “The Mental Health Continuum: From Languishing to Flourishing in Life.” Journal of Health and Social Behavior 43:207–22.

18.

Kleykamp

Meredith

. 2013. “Unemployment, Earnings, and Enrollment among Post-9/11 Veterans.” Social Science Research 42(3):836–51.

19.

Kleykamp

Meredith

Hipes

Crosby

. 2015. “Coverage of Veterans of the Wars in Iraq and Afghanistan in the U.S. Media.” Sociological Forum 30(2):348–68.

20.

Kleykamp

Meredith

Hipes

Crosby

MacLean

Alair

. 2018. “Who Supports U.S. Veterans and Who Exaggerates their Support?” Armed Forces & Society 44(1):92–115.

21.

Lammers

Joris

Stoker

Janka I.

Stapel

Diederik A.

2009. “Differentiating Social and Personal Power: Opposite Effects on Stereotyping, but Parallel Effects on Behavioral Approach Tendencies.” Psychological Science 20(12):1543–48.

22.

Lester

Patricia

Flake

Eric

. “How Wartime Military Service Affects Children and Families.” The Future of Children 23(2):121–41.

23.

Link

Bruce

Cullen

Francis

Frank

James

Wozniak

John F.

1987. “The Social Rejection of Former Mental Patients: Understanding Why Labels Matter.” American Journal of Sociology 92:1461–500.

24.

Link

Bruce G.

Phelan

Jo C.

2001. “Conceptualizing Stigma.” Annual Review of Sociology 27:363–85.

25.

Link

Bruce G.

Phelan

Jo C.

Bresnahan

Stueve

Pescosolido

Bernice A.

1999. “Public Conceptions of Mental Illness: Labels, Causes, Dangerousness and Social Distance.” American Journal of Public Health 89:1328–33.

26.

Lucas

Jeffrey W.

Phelan

Jo C.

2012. “Stigma and Status: The Interrelation of Two Theoretical Perspectives.” Social Psychology Quarterly 75(4):310–33.

27.

MacLean

Alair

Kleykamp

Meredith

. 2014. “Coming Home: Attitudes toward U.S. Veterans Returning from Iraq.” Social Problems 61(1):131–54.

28.

Martin

Jack K.

Pescosolido

Bernice A.

Tuch

Steven A.

2000. “Of Fear and Loathing: The Role of Disturbing Behavior, Labels, and Causal Attributions in Shaping Public Attitudes toward People with Mental Illness.” Journal of Health and Social Behavior 41:208–23.

29.

Obama

Barack H

. “President Obama: Hire a Veteran” Retrieved October 13, 2017 (https://obamawhitehouse.archives.gov/blog/2011/11/21/president-obama-hire-veteran).

30.

Paolacci

Gabriele

Chandler

Jesse

. 2014. “Inside the Turk: Understanding Mechanical Turk as a Participant Pool.” Current Directions in Psychological Science 23(3):184–88.

31.

Pescosolido

Bernice A.

Monahan

John

Link

Bruce G.

Stueve

Ann

Kikuzawa

Saeko

. 1999. “The Public’s View of the Competence, Dangerousness, and Need for Legal Coercion of Persons with Mental Health Problems.” American Journal of Public Health 89(9):1339–45.

32.

Phelan

Jo C.

Link

Bruce G.

2004. “Fear of People with Mental Illness: The Role of Personal and Impersonal Contact and Exposure to Threat or Harm.” Journal of Health and Social Behavior 45:68–80.

33.

Sadler

Melody S.

Meagor

Elizabeth L.

Kaye

Kimberly E.

2012. “Stereotypes of Mental Disorders Differ in Competency and Warmth.” Social Science and Medicine 74:915–22.

34.

Schein

Virginia

. 1973. “The Relationship between Sex Role Stereotypes and Requisite Management Characteristics.” Journal of Applied Psychology 57(2):95–100.

35.

Schomerus

Georg

Matschinger

Herbert

Angermeyer

Matthias C.

2013. “Continuum Beliefs and Stigmatizing Attitudes towards Persons with Schizophrenia, Depression, and Alcohol Dependence.” Psychiatry Research 209(3):665–69.

36.

Scott

Wilbur J

. 1990. “PTSD in DSM-III: A Case in the Politics of Diagnosis and Disease.” Social Problems 37(3):294–310.

37.

Smith

R. Tyson

True

Gala

. 2014. “Warring Identities: Identity Conflict and the Mental Distress of American Veterans of the Wars in Iraq and Afghanistan.” Society and Mental Health 4(2):147–61.

38.

Statistics Corner. 2009. “Choosing the Right Type of Rotation in PCA and EFA.” JALT Testing & Evaluation SIG Newsletter 13(3):20–25.

39.

Tanielian

Terri

Jaycox

Lisa H.

2008. Invisible Wounds of War. Santa Monica, CA: RAND.

40.

Vogt

W. P

. 1993. Dictionary of Statistics and Methodology. Vol. 1. Thousand Oaks, CA: Sage.

41.

Weiner

Bernard

Perry

Raymond P.

Magnusson

Jamie

. 1988. “An Attributional Analysis of Reactions to Stigmas.” Journal of Personality and Social Psychology 55(5):738–48.

42.

Wilson

Jennifer Kelly

Brodsky

Stanley L.

Neal

Tess

Cramer

Robert J.

2011. “Prosecutor Pretrial Attitudes and Plea-bargaining Behavior toward Veterans with Posttraumatic Stress Disorder.” Psychological Services 8(4):319–31.

43.

Wilson

Laura C

. Forthcoming. “The Prevalence of Military Sexual Trauma: A Meta-analysis.” Trauma, Violence, & Abuse. https://journals-sagepub-com.web.bisu.edu.cn/doi/10.1177/1524838016683459