The stigmatized hero? A review of UK opinion polls and surveys on perceptions of British veterans in UK society.

cultural representations of veterans focus on suffering and trauma (McGarry, 2012; McGarry & Ferguson, 2012). In this case, socially constructed images of veterans overlap to some extent with definitions of victims in a victimological context (McGarry, 2012; McGarry & Ferguson, 2012). However, although previous studies suggest that veterans may be related to heroizing and victimizing sentiments, little is known whether and to which extent members of the British public may hold heroizing and victimizing beliefs about veterans. While countries such as the US and other European countries have produced robust research that examines and compares public perceptions of the military and those who have served it, the UK failed to provide an exhaustive series of studies with nationally representative sample sizes (cf. Hines et al., 2015). Moreover, the few UK studies that represent rigorous and sound research on pertinent perceptions of the UK Armed Forces remain to date unrelated to each other. No scientific attempt has been made Findings from representative opinion polls and surveys provide information on pertinent perceptions of veterans in British society. The present project compares these findings and contrast publicly pertinent per ceptions of veterans with factually correct information. This allows researchers to determine erroneous and stereotyped perceptions of veterans that may be held by members of the British society. The present project reviewed 41 veteran-related question-items from 11 surveys/polls and grouped these question items into the categories 1) employment situation, 2) housing situation, 3) skills and experiences, 4) physical/psychological well-being, and, 5) addictions and antisocial behaviour. The comparison with factually correct information indicates that the majority of the respondents held a set of negative and erroneous stereotypes against veterans. As negative and erroneous stereotypes may hinder a successful transition into civilian society, future research should explore how these stereotypes may form.


Introduction
Previous research indicates that veterans from the UK Armed Forces are a focal point of public interest in UK society (cf. also Hines et al., 2015). As such, representations of veterans have penetrated almost exclusively all aspects of the British public life and are widely distributed in media (cf. Gibson, 2012;McCartney, 2011;Gibson & Condor, 2009). Social scientists who examined cultural representations of veterans that reside in public discourses, media, literature, and films suggest that these representations of veterans resonate particularly with heroizing sentiments.
Essentially, studies on rhetorical functions in relation to military service outline that associations between heroism and service may represent a deeply rooted cultural assumption (i.e. Gibson, 2012;Gibson & Condor, 2009;Coy, Woehrle & Maney, 2008). Culturally perpetuated descriptions of existential threats (i.e., "evil transgressors" vs. "righteous us") and symbolic values (i.e. "freedom" vs. "oppression") in relation to "serving the country," may therefore feed a cultural assumption that equates service with heroism, leading to an unconditional appreciation of veterans (Gibson, 2012). This, as Gibson (2012) notes, allows us to categorise one group of people to be treated with more respect than another. An example is subgroups in society that claim recognition and respect on behalf of veterans who were part of the same subgroup (cf. Gibson, 2012).
In addition to a public heroization of veterans, previous research identified a prevailing cultural notion that relates veterans to victimizing sentiments. These victimizing sentiments are characterised by introducing injury into the public's consciousness with public sympathy increasing over time (McCartney, 2011, Jones & Wessely, 2005. Particularly, if warfare or specific deployments are socially conceptualised as illegitimate, then cultural representations of veterans focus on suffering and trauma (McGarry, 2012;McGarry & Ferguson, 2012). In this case, socially constructed images of veterans overlap to some extent with definitions of victims in a victimological context (McGarry, 2012;McGarry & Ferguson, 2012).
However, although previous studies suggest that veterans may be related to heroizing and victimizing sentiments, little is known whether and to which extent members of the British public may hold heroizing and victimizing beliefs about veterans. While countries such as the US and other European countries have produced robust research that examines and compares public perceptions of the military and those who have served it, the UK failed to provide an exhaustive series of studies with nationally representative sample sizes (cf. Hines et al., 2015). Moreover, the few UK studies that represent rigorous and sound research on pertinent perceptions of the UK Armed Forces remain to date unrelated to each other. No scientific attempt has been made 65 Phillips: The Stigmatized Hero? A Review of UK Opinion Polls and Surveys on Perceptions of British Veterans in UK Society to evaluate publicly pertinent beliefs concerning veterans and thus to contrast and compare veteran-related question items from different opinion polls and surveys with each other. The current state of knowledge may therefore neither allow the observation of coherence, development, and shifts in opinions about veterans between different polls and surveys over time, nor to determine publicly pertinent erroneous beliefs about veterans. This is problematic as publicly prevalent beliefs of veterans may impact veterans' successful reintegration into society (cf. Hines et al., 2015;Burdett, 2014). This issue might be particularly relevant for veterans from the most recent missions in Iraq and Afghanistan, as this younger population of veterans may not only face negative stereotypes related to public perceptions of veterans, but might also be confronted with negative attitudes towards these missions (cf. Gribble et al., 2015;YouGov, 2015). In addition, negative public perceptions of veterans may deter recruitment and cause disaffection among those who are currently serving.
The present project addresses this problem by examining publicly prevalent beliefs and comparing these prevalent beliefs with factually correct information. Therefore, the present project draws on a range of existing representative surveys and polls and examines similarities and differences between these representative surveys and polls. The comparison between public perceptions of veterans and factually correct information allows us to determine erroneous beliefs and stereotypes, and so, to offer a comprehensive overview of findings from previous research. Finally, areas for future research are identified.

Literature Review
A systematic, wide-ranging literature review was conducted by accessing the university owned library resources 1 , the database UK Data Service 2 , and Google. In addition, reference lists of all included polls and surveys were handsearched, while also relevant known websites (i.e., Ministry of Defence 3 ) were screened to make sure all relevant surveys and polls were included.
The university owned library resources were utilised as these included an easily accessible and exhaustive list of opinion polls and surveys from a wide array of different publishers. In contrast, the database UK Data Service represented a more specific database, offering the UK's most extensive collection of social, economic, and population data resources. Lastly, Google served as "back-up" to make sure that no representative poll or survey, which captured social opinions of veterans, was overlooked.
The present study intended to provide a general overview of the ways in which veterans are being thought of in British society. Therefore, a variety of search terms were utilised. The databases were searched with a combination of the following search terms: "Opinion poll"/"Survey", "British"/"Britain"/"UK"/"United Kingdom", "public"/"society," and "Veteran"/"former service personnel"/"ex-service personnel." Opinion polls and surveys were selected that included at least one question item that referred to the veteran or veteran-related terms (i.e., ex-service personnel, former personnel).
To be included, polls and surveys needed to have representative sample sizes for the UK population. This means that the present study only included surveys with sample cohorts of 1,000-plus adults, ages 18 or over from different focus groups (i.e., region distribution, gender, education). To become representative of all adults in the UK, the auxiliary variables such as age, gender, or region were weighted to each survey respondent. Respondents in under-represented groups had a greater weighting (larger than one) than those in over-represented groups (weighting smaller than one). By using the weighted values in further analyses and not just the values of the variables, the polls' and surveys' findings were made statistically representative for the UK population as a whole (cf. BSA, 2012). In addition, only surveys that were published between the twenty-second of May 2011 (the day of the last British soldier's withdrawal from Iraq) and the first of November 2017 were included to ensure the recency of polls and surveys.
A total of 11 surveys and polls were selected that included 41 question items referring to veterans or veteran-related terms (i.e., former personnel, ex-service personnel). An overview of publisher and number of question items can be found in Table 1.

Instruments
A theme-based content analysis (CA) grouped question items into thematic clusters. CA is considered to be, "a technique used to extract desired information from a body of material […] by systematically and objectively identifying specific characteristics of the material" (Smith, 2000, p. 314). The application of CA as a methodology for the present study provided the rationale for extracting veteran-related question items and for coding these question items in accordance to similarity. Therefore, the 41 extracted question items were grouped and categorised into five emerging themes. Themes referred to perceptions of veterans' 1) employment situation, 2) housing situation, 3) skills and experiences, 4) physical/psychological well-being, and, 5) addictions and antisocial behaviour.
An overview of the five themes can be found in Table 2. Inter-rater reliability for question item categorisation was obtained from one psychology postgraduate student. A 100% inter-rater concordance for all 41 question items was obtained which suggested homogeneity in question-item categorisation. Thus, the question-item categorisation was considered sufficiently validated for further analyses. Although the researcher intended to conduct descriptive (comparison/interpretative analysis) and parametric (chi-square test, ANOVA/t-test, multiple linear regressions) data analyses, the latter exceeded the limits of the possible. This was the case, as the published versions of polls and surveys did not allow researchers to observe potential relationships between answers and sociodemographic characteristics like age, gender, or education.
The researcher was unsuccessful in contacting the relevant institutions and asking for permission to access the data sets. The researcher's request was either ignored, or financial contributions were requested that exceeded the project's resources. Thus, the impact of sociodemographic variables on veteran-related question items' answers could not be observed. Moreover, eyeballing thematic similarities in veteran-related question items indicated too few frequencies to conduct statistically relevant chi-square tests. Thus, tests that may have examined significant differences between factors such as time and changes in public understandings of veterans could not be conducted. Due to these limitations, the following will provide a descriptive, comparative, and interpretative in-depth analysis of opinion polls' and surveys' question items. However, though no parametric tests could be conducted, the descriptive analysis will be  sufficient to draw comparisons between socially prevailing opinions of veterans and factually correct information.

Results
The following presents an analysis of the themes into which veteran-related questions of representative opinion polls and surveys were grouped. Each section presents and discusses potential differences between opinions from polls and surveys and will compare societies' common perceptions of veterans with factually correct information.

Perceptions of veterans' employment situation, skills, and experiences
The following will discuss publicly pertinent perceptions that were categorised into "employment situation," and, "skills and experiences" in the light of factually correct information. Essentially, the results from representative surveys and polls indicate that members of the British public's acknowledge and appreciate the veterans' skills and experiences (cf. Table 4). For example, throughout the last decade, ICM surveys indicated consistencies in the agreement of the British public with statements like, "People who have left the Armed Forces have skills and experiences that help them to succeed in many different civilian careers" (68% to 78% between March 2012 to summer 2015, cf. ICM surveys March, 2012March, to summer, 2015. Similarly, also 78% of individuals responsible for recruitment in companies perceived veterans as, "Capable, experienced individuals with a lot to offer," and 60% evaluated the veterans' impact on the workplace as "positive" (Ashcroft, 2012). Here, particularly veterans' discipline (57%) and loyalty (30%) were appreciated (Ashcroft, 2012). Veterans were perceived to be more likely than civilians to possess many positive work-related attributes, i.e., to work well under stress (91% agreed for officers and 86% for other ranks), to have excellent time management skills (93% agreed for officers and 86% for other ranks), or to have better communication skills than civilians (91% agreed for officer and 81% for other ranks, cf. Ashcroft, 2012). Therefore, the public appreciation of the veterans' skills and experiences may reflect the Ministry of Defence's (MOD) considerable investments into the professional development and training of its members. Throughout their military career, all ranks progress through a structured network of courses (education and training) that provide service personnel with knowledge and skills. This is based upon the MOD's training and education policy (cf. MOD JSP 898, 2014) and consists of three overlapping areas: professional development, career management personal development, and elective personal development.
Professional development comprises generic courses with defined start and output standards necessary to undertake a military appointment, rank, or role. Organised by Service Schools, accredited service partner or online courses, the MOD provides service personnel with opportunities to improve numeracy or literacy skills. Career management is often not compulsory for all military appointments or rank promotions but intends to benefit the individual's career and therefore contribute to the personnel's motivation. Elective personal development includes training and education that support service personnel's interests or future employment. Although less directly contributing to military capability, elective personal development represents a central strand of professional development, as it improves the personnel's ability to respond to career development. Although a range of provisions exists within the three overlapping areas of professional development, career management personal development, and elective personal development, the main opportunities address literacy and numeracy, apprenticeships and vocational qualifications, command leadership and management training and-more generally-national degrees (i.e., GCSE, A -Levels, University Short Course Programmes, Postgraduate Studies).
In consideration of the MOD's investment into its members and consistently positive public reflections on the veterans' employability (cf.  Ashcroft, 2012), or at least as likely to be unemployed as a civilian individual (50.7%, BSA, 2012).
However, these perceptions are factually incorrect; the veteran-population that received Career Transition Partnerships (CTP) or CTP Future Horizons was factually less likely to be unemployed than civilians (MOD Career Transition Partnership Ex-Service Personnel Employment, 2018). CTP delivers different kinds of free resettlement services to all ranks of the UK military and makes the transition from military to civilian life as successful as possible by providing workshops for writing a CV, learning job-interview techniques, and applying for jobs. Although all veterans are entitled to receive CTP there are two exemptions. Firstly, veterans who were medically discharged (i.e. wounded, injured, sick) automatically qualify for the Core Resettlement Programme (CRP) regardless of time served. Where additional support is required, medically discharged veterans are entitled to access CTP Assists. CTP Assists offer specialist support that provides access to vocational assessments and a vocational opportunities portal or referrals to specialist employment consultants. Secondly, Service Leavers who completed less than four years of service or who faced compulsory discharge are not entitled to access the full CTP range of resettlement services but may use the CTP Future Horizons programme. CTP Future Horizons is a programme that offers the veteran the opportunity to speak to an employment advisor. Depending on the veteran's need for further support, access to advice from the Central Employment Team and the Career Advisor is provided.
Despite MOD (2018) statistics indicating unemployment being significantly lower for veterans than for civilians, a few caveats must be considered when drawing comparisons between the veteran and the civilian population. Firstly, there were notable differences in sociodemographic characteristics between the veteran and the civilian sample. For example, 90% of veterans were male compared with 49% of the UK population. Males are generally more likely to be employed than females (cf. ONS, 2018). Therefore, this gender disparity across compared populations may be driving employment rates of veterans being higher than those of the UK population. Secondly, employment and unemployment rates for veterans were estimated six months after they had left service, while the UK working population will have included individuals who had sought employment for longer than six months. Lastly, MOD transition reports only  include veterans who received CTP or CTP Future Horizon services. Therefore, MOD transition reports fail to include the entire veteran population. Interestingly, data from the most recent Royal British Legion's household survey (2014) indicated that individuals from the ex-service community aged 16-64 are more likely to be unemployed than the general population (8% of veterans vs. 5% of civilians), and more likely to be economically inactive (32% of veterans vs. 22% of civilians). This difference may be explained by the definition of "ex-service community." The "ex-service community" did not solely consist of the veteran population. Instead, the "ex-service community" comprised 700,000 individuals; from those around 120,000 veterans and 20,000 dependants were unemployed, around 130,000 veterans and 110,000 dependants were looking for work; and, around 160,000 veterans and 160.000 dependants were retired or in education. In this sense, it may be argued that the Royal British Legion's (2014) report may not be solely outdated by the MOD's (2018) publication, but may also provide less accurate information on the veterans' employment situation.
Moreover, previous research identified specific risk factors that are associated with a veteran's unemployment (cf. Burdett et al., 2018). Veterans were more likely to claim unemployment benefits when they were also Early Service Leavers, male, ex-army, less educated, leaving in an unplanned manner, and having claimed unemployment benefits before service. Moreover, antisocial behaviour and childhood adversities increased the likelihood of claiming unemployment benefits (Burdett et al., 2018). Interesting is veterans' risk-profile for unemployment related to the civilian's risk-profile for unemployment. Research indicated that low education, aggression, and childhood adversities lead to long-term unemployment in adulthood in the civilian population (i.e., Daly & Delaney, 2013;Bellis et al., 2013;Kokko & Pulkkinen, 2000;Caspi, 1998). In conclusion, unemployment may be less likely to be dependent upon discharge from the military, but upon factors that put both veterans and civilians equally at risk of suffering from unemployment.
In summary, although veteran and civilian risk-profiles for unemployment are comparable, opinion polls and surveys indicated a high proportion of the UK general public as holding incorrect beliefs about veterans' economic situation. While the British public was generally aware and appreciative of veterans' skills and experiences acquired through military service, individuals in British public also held the belief that veterans suffer from a greater risk of unemployment (cf. Table 3). This raises the question: Why was the veteran perceived as both highly skilled and employable, but similarly as encountering difficulties in the job market?

Perceptions of veterans' housing situation
Data from representative opinion polls and surveys suggest that a high proportion of the British public understands homelessness to be higher amongst ex-members of the Armed Forces (40%, Ipsos MORI, 2015;47%, Ashcroft & KCMG, 2017), or just about as likely as in the civilian population (44%, BSA, 2012; cf. Table 5). This perception is  York, 2008). Similarly, a homeless provision survey suggested that approximately three percent of accommodation project users and two to three percent of day centre users, direct access hostels and second stage clients were veterans (Homeless Link;SNAP, 2013;Jones et al., 2014). In conclusion, evidence suggests that veterans are slightly less likely than members of the civilian population to be homeless (6.1% of civilians are homeless according to the Adult Psychiatric Morbidity Survey, 2007). Essentially, research outlines that the decreasing rate of homelessness among the veteran population since the 1990s (Quilgars et al., 2017) responds to MOD initiatives and schemes. The MOD's housing policy is generally based on provisions of suitable accommodations for singles or families. These accommodations intend to enable the entitled person or family to carry out their duties. While personnel are serving, the MOD is providing financial aid for individuals to purchase homes. Service accommodations are approximately a third to a half cheaper than civilian equivalences, as these accommodations are being subsidized by the government. The entitlement to military-provided accommodation usually ends when personnel leave regular service. In general, personnel receive three months' notice to vacate the accommodation, which ideally coincides with the discharge date. Through the Joint Services Housing Advice Office (JSHAO), the MOD provides specialist housing information and advice on finding civilian housing options. This happens in "briefs"-presentations followed by discussions and advice sessions, which are available to service personnel at any point in their career. Moreover, the JSHAO gives guidance on relevant housing issues and service providers on their website and in a monthly magazine. The JSHAO also manages the MOD Referral Scheme, whereby those at risk of homelessness are assisted in finding vacant social housing.
Another supportive measure addresses home ownership. To enable service personnel to buy property, the MOD operates a scheme known as Long Service Advance of Pay (LASP), which entitles personnel with four or more years of trained service, to access a ten-year interest-free loan of £8500. Since 2010 the MOD also offers an interest-free loan of up to £75000 for personnel with four to six years' experience (Armed Forces Home Ownership Scheme, AFHOS). The most recently established scheme (Forces Help to Buy Scheme, est. 2014) offers interest free-loans of up to 50% of salary. Therefore, it is unsurprising that recent surveys indicated 59% of members from the Royal Navy to own a property, 54% of Personnel from the Royal Air Force do so, 50% of their colleagues from the Royal Marines, and 34% of Army-members, which may be an expression of their more nomadic existence in general (cf. AFCAS Survey, 2013).

Perceptions of veterans' behavior
This section splits veterans' behavior into three sections: First, general anti-social behavior, conceptualized as being imprisoned, will be discussed. Then, a short section will present publicly pertinent opinions and factually correct information on risk-drinking behavior and alcoholism among veterans. Lastly, drug and substance abuse among veterans will be discussed by contrasting public perceptions and factual information. The relevant question items that refer to this theme can be found in Table 6.

Anti-social behavior: Imprisonment
In summary, opinion polls and surveys indicate that the majority of the British public think of former Armed Forces personnel to be just as likely to be in prison (35%) or even more likely to be in prison (19%) than the population as a whole (cf. Ipsos  Yet, results from recent studies indicate that although the proportion of imprisoned veterans is less than a comparable sample of the general population, violent offending did not follow this trend and was overrepresented in the veteran sample (MacManus, 2013). Imprisoned veterans were more likely to have been convicted of more serious crimes and thus had longer sentences than members from the general public (63% of the veteran population reported to have sentences over four years vs. 53% of the general prisoner population; 39% of the veteran population reported to have sentences over 10 years vs. 26% of the general prisoner population; HM Inspectorate of Prisons, 2014; DASA, 2010). In conclusion, the most common offences committed by veterans differed from those committed by the general population. The veterans' most frequent offences included violence against a person (33% of veteran prisoner population vs. 29% of general prisoner population) and sexual offences (25% of veteran population vs. 11% of the civilian population; cf. DASA, 2010).
To understand and identify "risk profiles," recent studies conducted interviews with imprisoned veterans and consulted experts in research, mental health, social care, criminal justice and legal matters (i.e. Howard League, 2014;MacManus et al., 2013). The reoccurring themes were integral links between poor mental health, alcohol, substance misuse, and offending behavior (i.e. MacManus et al., 2013. In conclusion, the Howard League (2014) identified three "risk-profiles": The first group comprised veterans Table 6: Public perceptions of anti-social behavior, alcoholism and substance abuse. who had adverse childhood experiences and suffered from violence. These problems were associated prior to listing and were often accompanied by substance or alcohol misuse. The second group contained veterans whose difficulties arose during their service and which were often related to their "discharge"-meaning the release of their military obligations (i.e. mental health or physical injury, cf. also Fear et al., 2010). The third group contains veterans who experienced transition-related issues (i.e. finding a job). All three "risk-profiles" were found to be related to social isolation and exclusion, alcohol or substance abuse, and financial problems which are all more general, social "risk-factors" for criminal behavior (Howard League, 2014). In conclusion, the background risk profile of veteran-offender was similar to that of other, civilian offenders (cf. MacManus et al., 2015;DASA, 2010). Despite previous research successfully identified risk factors in offenders, to date there is only a small existing body of literature that quantitatively investigated causal factors in offending behavior (cf. MacManus et al., 2015). Previous research that addressed causal relationships between 'risk-profiles' and offending behavior did so exclusively in the context of deployment and combat experience. For example, MacManus et al., (2015) found that deployment was not independently associated with increased risk of offending behavior, though that serving in a combat role conferred an additional risk for offending behavior. Increased exposure to traumatic events during deployment also increased the risk of offending behavior, and, offending behavior was found to be strongly associated with post-deployment alcohol misuse, post-traumatic stress disorder (PTSD) and high levels of self-reported aggressive behavior . Rona et al. (2015) provided further evidence for these relationships. The results indicated strong intermediate associations between combat exposure, any mental health problem (i.e. PTSD, alcohol misuse), childhood adversities and anger.

Year of Publication Institution/Name of Publication Sample
In conclusion, evidence indicated that the small proportion of veterans who committed offences might be related to a complex interaction between their experiences in the military, mental health problems and their socioeconomic circumstances while transitioning from the military to civilian life (cf. MacManus, 2015). Thus, better collaborations between the criminal justice system and mental health agencies, greater support and interventions not just during the incarceration but also earlier in the criminal pathway may be required to help veterans that are experiencing difficulties in readjusting to civilian life (cf. MacManus, 2013).
However, the fact that veterans are approx. 30% less likely to be in prison than civilians was most often not acknowledged by members of the general public. Yet, although British society seems to be misinformed about the veterans' likeliness to be imprisoned, no scientific attempt has investigated this misperception in greater depth.

Risk-drinking behavior (alcoholism)
Opinion polls and surveys suggest that the majority of the UK public correctly identify high-risk drinking behavior and alcohol misuse as a common problem for Ex-Service Personnel. For example, evidence indicated that 51% of the general population would identify the veteran as much more likely or somewhat more likely to be addicted to alcohol (Ashcroft & KCMG, 2017). Correctly identifying Ex-Service Personnel's risk-drinking behavior is a recent trend. In earlier surveys, the majority of the UK population perceived both, civilians and Ex-Service Personnel to be equally likely to misuse alcohol (55.3%; BSA, 2012).
Research suggests alcohol misuse to be common among Service Personnel. For example, Service Personnel was found to have significantly higher levels of alcohol consumption than the general population (i.e. Thandi et al., 2015;Iversen et al., 2007), particularly when being in periods of ' decompression' -the period of leave given to Service Personnel after deployment (Hacker-Hughes et al., 2008). However, alcohol misuse was found to continue being a problem when Service Personnel leave the UK Armed Forces, being one of the Ex-Service Personnel's most common problems for which only a small proportion of veterans seek help (Murphy et al., 2016;Alcohol-Concern, 2012;Fear, Wood & Wessely, 2009).
In conclusion, the majority of today's British public seemed to be well aware of the veterans' alcohol consumption. The question why the general public may adopt negative facts but are seemingly more resistant to positive factually correct information (i.e. imprisonment, homelessness) remains although unaddressed.

Drug abuse
Evidence indicated the misuse of alcohol to be a far more significant problem for members and Ex-Members of the UK Armed Forces than the misuse of other substances such as drugs (Centre for Social Justice, 2014). Therefore, both, polysubstance misuse (misusing both, drugs and alcohol) and drug abuse were found to be minor issues amongst (Ex)Service Personnel (NAPO, 2009;Centre for Social Justice, 2014). For example, the 2011 study of the health needs of veterans living in Wales (Welsh Government Consultation Document, 2013) found that drug use among veterans was generally comparable with lower rates than in the general population. Similar tendencies were outlined in the Veteran's Aid Annual Report (2013) concerning the London veteran population, in the NAPO Report (2009) concerning the imprisoned veteran-population, and on a more national base, in the Centre for Social Justice Report (2014) concerning the veteran population as a whole. Therefore, previous research provided evidence for the veteran population being less likely to abuse drugs than the civilian population (cf. Home Office, 2013). In conclusion, also in the academic field prevalent research that discusses potential relationships between British veteran's substance abuse and problematic gambling attitudes (i.e. Roberts et al., 2018;Dighton et al., 2018) without further investigating these relationships may be misleading.
Interestingly, previous research indicated that people in British society may have erroneous beliefs regarding 77 Phillips: The Stigmatized Hero? A Review of UK Opinion Polls and Surveys on Perceptions of British Veterans in UK Society veterans' likeliness to abuse drugs. The majority of the British population understands veterans as either as likely as members of the general public to abuse drugs (BSA, 2012), or as more likely to be addicted to drugs than civilians (Ashcroft & KCMG, 2017). The question why this erroneous belief remains to be pertinent in UK society remained to date unaddressed.

Perceptions of the veterans' physical and psychological well-being
Healthcare provisions for the UK Forces are managed by the Defence Medical Service (DMS), which employs 7500 Armed Forces, 2200 civilian and 3500 reservist health professionals (DMS, 2018). The DMS provides healthcare for 147,030 UK Regular Forces Personnel, including servicemen within the UK, abroad, at sea and in some circumstances also for dependants of Personnel and entitled civilians. The DMS has no formal remit or funding set aside for veterans whose care lies within the competence of the NHS, as it does for reservists. Where appropriate the DMS also provides screenings and assessments for veterans and reservists only in conjunction with the individual's GP who remains primacy for the individual's care (DMS, 2018).
On enlistment, the responsibility of the recruit (except for reservists) transfers from the NHS to the DMS where the MOD initiates a health record. The health record is managed by the Defence Medical Information Capability Programme and includes a summary of pre-enlistment examinations and the healthcare commissioned by the DMS. It remains archived by the NHS and is retrievable upon discharge of the individual by the responsible NHS GP. Therefore, the Service Personnel's health remains well monitored and provides evidence for the overwhelming majority of Service Leavers departing in healthy and fit conditions (MOD, 2017). That most Service Leavers depart healthily may be related to the high levels of physical fitness required to serve in the UK Armed Forces. Before discharge, Service Leavers need to complete the "release medical process"-a consultation three months before discharge to identify and address remaining medical problems and a final consultation in the last days before discharge to review the medical condition of the Service Leaver. The Service Leaver is then provided with the FMed133, a paper document that includes a summary of the in-service healthcare and advice for future healthcare providers how to access the individuals' in-service medical history. The Service Leaver is advised to register with a dentist and an NHS GP to whom he shall pass the FMed155 onto. On leaving, the veteran is entitled to full NHS healthcare and under specific circumstance also eligible for priority treatment.

Veterans' physical health
The British Legion's UK household survey of the Ex-Service Community (British Legion, 2014b) outlined that working-age veterans (aged 16-64) were significantly more likely to report long-term illnesses that limit their activities. In response to the British Legion's (2014b) report, the MOD raised three main statistical concerns that the MOD took into consideration in its 2014 Annual Population Survey (MOD, 2016). Firstly, the MOD (2016) criticized differences in the population structures that were not considered in the Legion's report (British Legion, 2014b). The veteran proportion was predominantly male and older than the non-veteran population. While the MOD's annual population survey (2016) used standardizing procedures to take the population structures into account, the Legion's (2014b) failed to incorporate these procedures into their analyses. Secondly, the MOD (2016) outlined that the Legion's (2014b) large sample size may have produced statistically significant results, although the differences between observations were only trivial, as the effect size was not calculated. The MOD (2016) used the measure of effect to identify differences that were large enough to note and not solely statistically significant. Lastly, the MOD (2016) found the Legion's (2014b) standard test formulas to assume the data was from a simple random sample. The formula would have needed to be adjusted, using the design factor, to take account of both, the standardization and the sample design. In contrast to the Legion's (2014b) study, the MOD provided a full background quality report with information on statistical procedures provided. By taking these statistical adjustments into account, the MOD (2016) did not find any statistically significant differences between the veteran and the civilian population in terms of physical health problems. A comparison between the MOD's and the Legion's results can be found in Table 7. Similar tendencies were outlined in more recent research. For example, the MOD's (2017a) Annual Population Survey, indicated that there was no significant difference between veterans and civilians in terms of health problems limiting or affecting the amount and kind of activities (i.e. work that can be carried out). Veterans and civilians also commonly reported the same three longterm health conditions: heart, blood pressure, or circulatory related conditions (33% of both: working age veterans and civilians; 53% of both: retirement age veterans and civilians); legs or feet related conditions (35% and 27% of working age veterans and civilians; 37% and 34% of retirement age veterans and civilians); back or neck related conditions (31% and 25% of working age veterans and civilians; 23% and 22% of retirement age veterans and civilians); moreover, retirement age veterans also reported difficulties hearing (23%), with no statistically significant difference to retirement age civilians (17%; cf. MOD, 2017a;McManus et al., 2009).
While research outlined no significant differences between the veteran and the civilian-population, studies indicated significant differences within the veteran population. For example, Hotopf et al. (2006) found a weak association between the presence of multiple physical symptoms and deployment. Moreover, the impact of deployment on veterans' physical health also depended upon the deployment. In this sense, not all deployments affected veterans' health negatively-some deployments did more so than others (Unwin et al., 1999). Thus, instead of deployment being per-se physically damaging, the weak association between multiple physical symptoms and deployment may be explained by the physical challenges of deployment that are more apparent in some deployments. This is further evidenced by studies that examined the consequences of overstretch in the UK Armed Forces. For example, Rona et al. (2007) outlined that personnel who were deployed for 13 months or more in the past three years were more likely to have multiple physical symptoms. This was not the case if the personnel were not overstretched.
In conclusion, though generally spoken the majority of veterans from the UK Armed Forces do not statistically differ from the civilian population in terms of physical health, the physically straining nature of deployment may be related to physical symptoms. Goodwin et al., (2015) outlined that common mental health disorders may be more common amongst currently serving Personnel of the UK Armed Forces than in the civilian population. As Goodwin et al.'s (2015) study reflected proportions of currently serving Personnel, direct inferences to a veteran population could not be drawn.

Veterans' psychological health
The British Legion (2014b) indicated significant differences between veterans' psychological health and the civilian's. The MOD's (2016) report rejected the Legion's (2014b) findings due to the same statistical limitations concerning standardization procedures, sample size and measure effect as mentioned in the previous section (The veterans' physiological health). In contrast to the Legion's (2014b) survey, the MOD (2016) did not find any significant differences between veterans' and the civilians' psychological health. An overview can be found in Table 8.
Longitudinal studies also outlined insignificant differences between the veteran and the civilian population. Analyses of the MOD's Mental Health Reports between 2007 and 2012 indicated that the number of Personnel in need of mental health services ranged between 1.8% and 3.2% (MOD, 2017b). Moreover, the prevalence of mental health problems declined with age (cf. MOD, 2017b), a trend that is supported by evidence from KCMHR studies (cf. Wessely, 2012).
Also, estimates regarding the prevalence of PTSD in service-Personnel and veteran populations are considerably low. Longitudinal studies indicated that the PTSD prevalence ranged between 3% and 6%, which is comparable to estimates of PTSD prevalence in members of the general public (3% to 5%; APMS, 2009;Hunt, Wessely, Jones, Rona & Greenberg, 2014).
Similarly, suicide rates were not higher among the veteran population. Research outlined that veterans were less likely than civilians to commit suicide (Kapur, While, Blatchley, Bray, Harrison, 2009). Yet, although the overall suicide rates were lower for veterans, the rates for young veterans (16-24) and those classified as Early Service Leavers (less than four years of service) represented an exemption. Evidence suggests that these populations are at increased risk of suicide as a result of pre-service vulnerability, referring to childhood adversities and low educational attainment (cf. KCMHR, 2017).
Pre-service vulnerability was also associated with affecting the psychological health of veterans negatively. An extensive series of research indicated that psychological health problems may relate to a combination of intrinsic factors (individual factors) and extraneous factors (experiences during deployment; i.e. Jones et al., 2012;Iversen et al., 2008;Rona et al., 2007;Browne et al. 2007;Hotopf et al., 2006). For example, individuals from the regular (full-time) veteran population were more at risk to suffer from psychological health problems when pre-service vulnerability was combined with combat experience, being lower rank, ex-Army and unmarried (Iversen et al., 2008;Iversen et al, 2007). Reserve veterans, in comparison, were found to be generally at higher risk than regular veterans as reserve veterans did  Fear et al. (2010) found only a weak association between deployment and probable post-traumatic stress disorder in regular Personnel (full-time personnel) while a strong association between deployment and probable post-traumatic stress disorder for reservists (part-time personnel) was outlined. Alcohol misuse was significantly associated with deployment for both, regular and reservist Personnel.
For both veteran populations, regulars and reservists, evidence failed to find an effect of length of tour or number of tours on psychological health (KCMHR, 2017). Veterans' psychological health was found to be only negatively affected when the Armed Forces Harmony Guidelines were not followed or when the actual tour length exceeded expectations (KCMHR, 2017).
In conclusion, although research outlined that the majority of Service Leavers do well, a minority of veterans were identified to have experienced adverse mental health. This group was also found to be at risk of suffering from social exclusion and unemployment (Iversen et al., 2005). Steps to improve the knowledge and expertise of primary care services about veterans' mental health issues and to increase the availability of treatment options are therefore crucial to optimize the veterans' transition and economic situation.
Research indicated that the veterans' willingness to access these primary care services might be problematic (Mellotte et al., 2017;Jones et al., 2015;Iversen et al., 2011). This is the case as psychological health problems were found to remain subject to stigmatization, though not a to static stigmatization . Instead, the stigma associated with psychological health problems seemed to fluctuate in proportion to the frequency and severity of psychological health problems -the more severe the psychological health problems which were experienced, the less individuals were likely to access help services due to fear of stigmatization .
Studies indicated that the barrier of stigmatization in help-seeking behavior might be overcome by "initial enablers" that included social support, motivation, and the media (Mellotte et al., 2017;Iversen et al., 2011). "Treatment pathway enablers" were found to help overcoming the second, more practical barrier of help-seeking behavior by facilitating access to a veteran-specific service and establishing a good therapeutic relationship after having been diagnosed with a mental health problem (cf. Mellotte et al., 2017;Iversen et al., 2011). Moreover, recent research indicated an increase in the absolute number of veterans visiting Service Charities and the NHS for help with mental health problems (KCMHR, 2017). This increase in help-seeking behavior may reflect a modest reduction in the stigmatization of psychological health problems (cf. KCMHR, 2017).
In summary, there is no clear evidence that veterans' mental health is substantially worse than that of a comparable civilian population. A small minority of veterans may experience psychological health problems, particularly combat veterans, those with pre-service vulnerability, early service leavers and reservists.

Opinion polls and surveys
Despite the evidence suggesting veterans to be in good physical and mental health, public opinions towards veterans' health are negative. For example, as outlined in Table 9, statistics indicated that the UK general population perceived mental (82%) or physical (61%; Ashcroft & KCMG, 2017) health problems, or, mental and physical health problems (92%; Ashcroft, 2012) to be common among the veteran population. Similarly, the majority of the general public misperceives veterans to be more likely to commit suicide (54% in Ashcroft & KCMG, 2017;50.9% in BSA, 2012). In conclusion, the British general population seemed to have erroneous beliefs about veterans' state of health. The question why these erroneous beliefs are socially pertinent and resistant to change remains to date unresearched.

Discussion
Previous research indicated the majority of British veterans to be successful in beginning new careers, are of good health and unlikely to suffer from homelessness. Evidence from different polls and surveys indicated members of British society to hold a set of factually incorrect beliefs. In British society, veterans are commonly perceived as • being highly skilled individuals • struggling to find employment (factually incorrect) • more likely or equally likely to be homeless as members of the general public (factually incorrect) • just as likely as a member of the general public to be in prison (factually incorrect) • likely to engage in high-risk drinking behaviours (factually correct) • more likely to abuse drugs (factually incorrect) • being more likely to have suicidal tendencies and physical or psychological health problems (factually incorrect) In conclusion, the evaluation of opinion polls and surveys indicated that individuals in UK society might hold erroneous and negative beliefs about veterans. Thus, the depiction of veterans as victims that was found to be a common theme in media (c.f. Parrott, Albright, Dyche & Steele, 2018;Pinder et al., 2009) may resonate with publicly pertinent perceptions of veterans. In this sense, it may be possible that societally pertinent conceptualizations of veterans as being damaged victims may be related to influences from media. Essentially, the results did not overlap with any aspects that concerned the veteran-hero representation that was identified by previous research (i.e. Gibson, 2012;Gibson & Condor, 2009). This does not allow to draw the conclusion that individuals in British society do not understand veterans as heroes. Instead, the closed ended question format may not have allowed participants to relate veterans to heroism. This outlines a general problem with the present data that makes it difficult to explain these pertinent contradictory and erroneous beliefs. Surveys and opinion polls focused solely on quantitative closed-ended questions. This question format provided basic and descriptive information and failed to address false beliefs in an explanative nature. In this sense, original beliefs about veterans were not investigated as individuals answered given question items that may have had little significance for the individual's original understanding of a veteran. Moreover, it is questionable whether individuals thought about veterans when answering question items. Although, using terms that describe the veteran ("people leaving the Armed Forces", "former members of the Armed Forces"), or using related terms ("Ex-Service Personnel"), no survey or poll referred to the actual term "Veteran." The reasons, why surveys and polls used these terms instead of "veteran" were not outlined. Therefore, it is not entirely clear whether individuals would give the same answers when being introduced to the term "Veteran." Lastly, surveys and polls failed to conduct (or publish) substantial data analyses. For example, the polls and surveys did not provide any information on how sociodemographic characteristics may impact perceptions of veterans or whether any potential answering patterns of veteran-related question items were found. In this sense, future research may want to gain access to data from the publishers in order to conduct a thorough analysis of socially prevalent understandings of veterans.
While further analyses and comparisons between existing data sets would be beneficial for future studies, the present study fulfilled its intention. The findings provided a systematic overview of socially pertinent perceptions of veterans and assessed factual information to outline whether these perceptions were correct. Therefore, the present study provides an overview of stereotypical perceptions of UK Armed Forces veterans.