Principles of Social Psychiatry

STIGMA AND DISCRIMINATION 333 Is it important to undertake research about stigma job because he or she fully expects to fail in any such and discrimination? In short, yes, because it can lead application) [44]. This distinction between experi- to low rates of help-seeking, lack of access to care, enced and anticipated discrimination is closely related undertreatment and material poverty, and to social to what has been described as the difference between marginalization [43]. These effects can be the con- ‘enacted’ and ‘felt stigma’. ‘Enacted stigma’ refers to sequences of experienced (actual) discrimination (e.g. events of negative discrimination, while ‘felt stigma’ being unreasonably rejected in a job application) or includes the experience of shame of having a condi- they can be the consequences of anticipated discri- tionandthefearofencountering‘enactedstigma’[45], mination (e.g. when an individual does not apply for a and is associated with lower self-esteem. 26.4 IGNORANCE: THE PROBLEM OF KNOWLEDGE Whilethereisanunprecedentedvolumeofinformation in public attitudes towards people with mental illness, in the public domain, the level of accurate knowledge as shown recently in New Zealand and Scot- about mental illnesses (sometimes called ‘mental land [48,49]. In a campaign in Australia to increase health literacy’) is meagre [46]. In a population survey knowledge about depression and its treatment, some in England, for example, most people (55%) believed states and territories received this intensive, coordi- that the statement ‘someone who cannot be held nated programme, while others did not. In the former, responsible for his or her own actions’ describes a people more often recognized the features of depres- person who is mentally ill [47]. Most (63%) thought sion and were more likely to support help-seeking for that fewer than 10% of the population would experi- depression or to accept treatment with counselling and ence a mental illness at some time in their lives. medication [50]. Similarly, recent evidence compar- Interventions to improve public knowledge about ing trends between Scotland and England in public depression can be successful, and can reduce the attitudes towards people with mental illness are con- effects of stigmatization. At the national level, social sistent with a positive effect of the Scottish ‘See Me’ marketing campaigns have produced positive changes antistigma campaign [51]. 26.5 PREJUDICE: THE PROBLEM OF NEGATIVE ATTITUDES Although the term ‘prejudice’ is used to refer to than do stereotypes. Interestingly, there is almost many social groups that experience disadvantage, nothing published about emotional reactions to peo- e.g. minority ethnic groups, it is employed rarely in ple with mental illness apart from that which relation to people with mental illness. The reactions describes a fear of violence [52]. An example of a host majority to act with prejudice in rejecting a of such negative attitudes are the terms used by minority group usually involve not just negative school students towards people with mental health thoughts but also emotion such as anxiety, anger, problems, and in one English study, among 250 such resentment, hostility, distaste or disgust. In fact, terms used, none were positive and 70% were prejudice may more strongly predict discrimination negative [53]. 26.6 DISCRIMINATION: THE PROBLEM OF REJECTING AND AVOIDANT BEHAVIOUR Surveysofattitudeandsocialdistance(unwillingnessto imaginary situations or what they think ‘most people’ have socialcontact)usuallyask either studentsormem- who do, for example, when faced with a neighbour or bers of the general public what they would do in work colleague with mental illness. Important lessons

334 SOCIAL CONSEQUENCES AND RESPONSES have flowed from these findings. This work has empha- actual behaviour, without assessing such behaviour sized what ‘normal’ people say without exploring how directly. Such research has usually focused on hypothe- people with mental illness experience the behaviour of ticalratherthanrealsituations,neglectingemotionsand normal people towards them. Further, it has been the social context, thus producing very little guidance assumed that such statements (usually on knowledge, aboutinterventionsthatcouldreducesocialrejection.In attitudes or behavioural intentions) are congruent with short, most work on stigma has been beside the point. 26.7 GLOBAL PATTERNS OF STIGMA AND DISCRIMINATION Dowe know if discriminationvaries between countries mainly from urban sites, a number of common themes and cultures? The evidence here is stronger, but still emerged. The conditions most often rated as ‘mental frustratingly patchy [54]. Although studies on stigma illnesses’ were the psychotic disorders, especially and mental illness have been carried out in many schizophrenia. People with higher levels of education countries, few have compared two or more places, and tended to have more favourable attitudes to people few have included non-Western nations [55]. with mental illness. Alcoholism was considered to be In Africa, one study described attitudes to mentally the most common type of mental disorder. Most ill people in rural sites in Ethiopia. Among almost 200 people thought that a health professional needs to be relatives of people with diagnoses of schizophrenia or consulted by people with mental illnesses [36]. mood disorders, 75% said that they had experienced A great deal of work has studied the question of stigma due to the presence of mental illness in the stigma towards mentally ill people in Asian countries family, and a third (37%) wanted to conceal the fact and cultures [66–68]. Within China [69], a large-scale that a relative was ill. Most family members (65%) survey was undertaken of over 600 people with a said that praying was their preferred way of treating diagnosis of schizophrenia and over 900 family mem- the condition [56]. Among the general population in bers [70]. Over half of the family members said that Ethiopia schizophrenia was judged to be the most stigma had had an important effect on them and their severe problem, and talkativeness, aggression and family, and levels of stigma were higher in urban areas strange behaviour were rated as the most common and for people who were more highly educated. symptoms of mental illness [34]. In the field of stigma research we find that schizo- In South Africa [57,58], a survey was conducted of phrenia is the primary focus of interest. It is remark- over 600 members of the public on their knowledge able that there are almost no studies, for example, on and attitudes towards people with mental illness [59]. bipolar disorder and stigma. A comparison of attitudes Different vignettes,portrayingdepression,schizophre- to schizophrenia was undertaken in England and nia, panic disorder or substance misuse were presented Hong Kong. As predicted, the Chinese respondents to each person. Most thought that these conditions expressed more negative attitudes and beliefs about were either related to stress or to a lack of willpower, schizophrenia, and preferred a more social model to rather than seeing them as medical disorders [60]. explain its causation. In both countries most partici- Similar work in Turkey [61] and in Siberia and Mon- pants, whatever their educational level, showed great golia [62] suggests that people in such countries may ignorance about this condition [71]. This may be why be more ready to consider the individual to be respon- most of the population in Hong Kong are very con- sible for his or her mental illness and less willing to cerned about their mental health and hold rather grant the benefits of the sick role. negative views about mentally ill people [72]. Less Most of the published work on stigma is by authors favourable attitudes were common in those with in the US and Canada [11,33,63,64], but there are also less direct personal contact with people with mental a few reports from elsewhere in the Americas and in illness (as in most Western studies), and by women the Caribbean [65]. In a review of studies from (the opposite of what has been found in many Western Argentina, Brazil, Dominica, Mexico and Nicaragua, reports) [73].

STIGMA AND DISCRIMINATION 335 Little research on stigma has been conducted in is seen as less stigmatizing and is more often discussed India. Among relatives of people with schizophrenia openly. in Chennai (Madras) in Southern India, their main Little is written in the English language literature on concerns were the effects on marital prospects, fear stigma in Islamic communities, but despite earlier of rejection by neighbours and the need to hide the indications that the intensity of stigma may be rela- condition from others. Higher levels of stigma were tively low [68], detailed studies indicate that, on reported by women and by younger people with the balance, it is no less than we have seen described condition [74]. Women who have a mental illness elsewhere [86–89]. A study of family members in appear to be at a particular disadvantage in India. If Morocco found that 76% had no knowledge about the they are divorced, sometimes related to concerns condition, and many considered it chronic (80%), about heredity [75], they often receive no financial handicapping (48%), incurable (39%) or linked with support from their former husbands, and they sorcery (25%). Most said that they had ‘hard lives’ and their families experience intense distress because of the diagnosis [32]. Turning to religious from the additional stigma of being separated or authority figures is reported to be common in some divorced [35]. Muslim countries [37,90]. Some studies have found In Japan, mental illnesses are seen to reflect a loss of that direct personal contact was not associated with control, and so are not subject to the force of will- more favourable attitudes to people with mental ill- power, both of which lead to a sense of shame [76–78]. ness [91,92], especially where behaviour is seen to Although it is tempting to generalize about the degree threaten the social fabric of the community [61,93]. of stigma in different countries, reality may not allow A recent study used the Discrimination and Stigma such simplifications. A comparison of attitudes to Scale (DISC) in a cross-sectional survey in 27 coun- mentally ill people in Japan and Bali, for example, tries using language-equivalent versions of the instru- showed that views towards people with schizophrenia ment in face-to-face interviews between research staff were less favourable in Japan, but that people with and 732 participants with a clinical diagnosis of depression and obsessive-compulsive disorder were schizophrenia [94]. The most frequently occurring seen to be less acceptable in Bali [79]. areas of negative experienced discrimination were What different countries do often share is a high making or keeping friends (47%), discrimination by level of ignorance and misinformation about mental family members (43%), keeping a job (29%), finding a illnesses. A survey of teachers’ opinions in Japan and job (29%), and intimate or sexual relationships (29%). Taiwan showed that relatively few could describe the Positive experienced discrimination was rare. Antici- main features of schizophrenia with any accuracy. The pated discrimination was common for applying for general profile of knowledge, beliefs and attitudes was work or training or education (64%) and looking for a similar to that found in most Western countries, close relationship (55%), and 72% felt the need to although the degree of social rejection was somewhat conceal the diagnosis. Anticipated discrimination greater in Japan [80]. occurred more often than experienced discrimination. In a unique move aimed to reduce social rejection, This study suggests that rates of experienced discri- the name for schizophrenia has been changed in Japan. mination are relatively high and consistent across Following a decade of pressure from family member countries. For two of the most important domains groups, including Zenkaren, the name for this condi- (work and personal relationships), anticipated discri- tion was changed from seishi buntetsu byo (split-mind mination occurs in the absence of experienced disorder) to togo shiccho sho (loss of coordination discrimination in over a third of participants. This disorder) [81,82]. The previous term went against the has important implications: disability discrimination grain of traditional, culturally valued concepts of laws may not be effective without also developing personal autonomy, as a result of which only 20% of interventions to reduce anticipated discrimination, people with this condition were told the diagnosis by e.g. by enhancing the self-esteem of people with their doctors [83–85]. There are indications from mental illness, so that they will be more likely to service users and family members that the new term apply for jobs.

336 SOCIAL CONSEQUENCES AND RESPONSES 26.8 TAKING STOCK What sense can we make of all these fragments of initiatives have been launched in the UK aiming to information? Several points are clear. First, there is no improve public attitudes. The Royal College of Psy- known country, society or culture in which people chiatrists’ ‘Changing Minds’ campaign in England with mental illness are considered to have the same ran between 1998 and 2003. It advertised websites, value and to be as acceptable as people who do not showed campaign videos in cinemas, distributed leaf- have mental illness. Second, the quality ofinformation lets to the general public and health care professionals, that we have is relatively poor, with very few com- and created reading material for young people for use parative studies between countries or over time. Third, in the curriculum [46,97,98]. The Scottish Govern- there do seem to be clear links between popular ments ‘See Me’ campaign (2002–present) has a higher understandings of mental illness and whether people profile, is better funded and more extensive. It aims to in mental distress want to seek help or feel able to deliver specific messages to the Scottish population by disclose their problems [38]. The core experiences using all forms of media as well as cinema advertising, of shame (to oneself and for others) and blame outdoor posters, supporting leaflets in GP surgeries, (from others) are common everywhere stigma has libraries, prisons, schools and youth groups. It also has been studied, but to differing extents. Where compar- a detailed website containing interactive resources isons with other conditions have been made, then and its impact is regularly monitored and progress mental illnesses are more, or far more, stigma- reported in the public domain [49]. The investment of tized [95,96] – as noted, mental illness may be the public funds in government campaigns is an important ‘ultimate stigma’ [9]. Finally, rejection and avoidance step and evidence suggesting that ‘See Me’ may have of people with mental illness appear to be universal had a positive effect on attitudes in Scotland relative to phenomena. England is encouraging [52]. However, it remains the Prejudice and discrimination by the public against case that addressing public ‘knowledge’ and people with mental illness are therefore common, ‘attitudes’, as discussed above, does not necessarily deeply socially damaging, and are a part of more lead to a change in ‘behaviour’ and ‘discrimination’. widespread stigmatization. Stigma against people This remains an elusive goal and further work is with mental illness can contribute to negative out- needed to understand the complex relationships comes as well as perpetuating self-stigmatization and between these three elements of stigma and to identify contributing to low self-esteem. With a growing and develop evidence-based tools and interventions awareness about such stigma, a number of recent with which to tackle discrimination. 26.9 CONCLUSIONS If we deliberately shift focus from stigma to discri- level, social marketing techniques. Third, people who mination, there are a number of distinct advantages. have a diagnosis of mental illness can expect to benefit First, attention moves from attitudes to actual beha- from all the relevant antidiscrimination policies and viour, not if an employer would hire a person with laws in their country or jurisdiction, on a basis of mental illness, but if he or she does. Second, inter- parity with people with physical disabilities. Fourth, a ventions can be tried and tested to see if they change discrimination perspective requires us to focus not behaviour towards people with mental illness, without upon the ‘stigmatized’ but upon the ‘stigmatizer’. In necessarily changing knowledge or feelings. The key sum, this means sharpening our sights upon human candidates for active ingredients to reduce stigma are: rights, upon injustice and upon discrimination as (i) at the local level, direct social contact with people actually experienced by people with mental with mental illness [99–101] and (ii) at the national illness [7,24,41,102].

STIGMA AND DISCRIMINATION 337 REFERENCES mental disorders: an empirical assessment. American Sociological Review, 54, 100–123. 1. Cannan, E. (1895) The stigma of pauperism. Economic 17. Link, B. G., Struening, E. L., Rahav, M. et al. (1997) Review, I, 380–391. On stigma and its consequences: evidence from a 2. Thomas Hobbes of Malmesbury. Markes of the Absurd longitudinal study of men with dual diagnoses of Geometry, Rural Language, Scottish Church Politics, mental illness and substance abuse. Journal of Health and Barbarisms of John Wallis, Professor of Geometry and Social Behavior, 38 (2), 177–190. and Doctor of Divinity, London. Printed for Andrew 18. Link, B. G., Phelan, J. C., Bresnahan, M. et al. (1999) Cooke, 1657. Public conceptions of mental illness: labels, causes, 3. Gilman, S. L. (1982) Seeing the Insane, John Wiley & dangerousness, and social distance. American Journal Sons, Inc., New York. of Public Health, 89 (9), 1328–1333. 4. Gilman, S. L. (1985) Difference and Pathology: 19. Smith, M. (2002) Stigma. Advances in Psychiatric Stereotypes of Sexuality, Race and Madness, Cornell Treatment, 8, 317–325. University Press, Ithaca, New York. 20. Link, B. G. and Phelan, J. C. (2001) Conceptualizing 5. Goffman, I. (1963) Stigma: Notes on the Management stigma. Annual Review of Sociology, 27, 363–385. of Spoiled Identity, Penguin Books, Harmondsworth, 21. Link, B. G., Yang, L. H., Phelan, J. C. and Collins, P. Y. Middlesex. (2004) Measuring mental illness stigma. Schizophre- 6. Scambler, G. (1998) Stigma and disease: changing nia Bulletin, 30 (3), 511–541. paradigms. Lancet, 352 (9133), 1054–1055. 22. Jones, E., Farina, A., Hastorf, A. et al. (1984) Social 7. Hinshaw, S. P. and Cicchetti, D. (2000) Stigma and Stigma: The Psychology of Marked to Relationships, mental disorder: conceptions of illness, public atti- W. H. Freeman & Co., New York. tudes, personal disclosure, and social policy. Devel- 23. Sayce, L. (2000) From Psychiatric Patient to Citizen. opment and Psychopathology, 12 (4), 555–598. Overcoming Discrimination and Social Exclusion, 8. Mason, T. (2001) Stigma and Social Exclusion in Palgrave, Basingstoke. Healthcare, Routledge, London. 24. Chamberlin, J. (2005) User/consumer involvement in 9. Falk, G. (2001) Stigma: How We Treat Outsiders, mental health service delivery. Epidemiologia e Psi- Prometheus Books, New York. chiatria Sociale, 14 (1), 10–14. 10. Heatherton, T. F., Kleck, R. E., Hebl, M. R. and Hull, J. 25. Hinshaw, S. (2007) The Mark of Shame, Oxford Uni- G. (2003) The Social Psychology of Stigma, Guilford versity Press, Oxford. Press, New York. 26. Thornicroft, G. (2006) Shunned: Discrimination 11. Corrigan, P. (2005) On the Stigma of Mental Illness, against People with Mental Illness, Oxford University American Psychological Association, Washington, Press, Oxford. DC. 27. Littlewood, R., Jadhav, S. and Ryder, A. G. (2007) A 12. Wahl, O.F. (1999) Telling Is a Risky Business: Mental cross-national study of the stigmatization of severe Health Consumers Confront Stigma, Rutgers Univer- psychiatric illness: historical review, methodological sity Press, New Brunswick, New Jersey. considerations and development of the questionnaire. 13. Pickenhagen, A. and Sartorius, N. (2002) The WPA Transcultural Psychiatry, 44 (2), 171–202. Global Programme to Reduce Stigma and Discrimi- 28. Weiss, M. G., Jadhav, S., Raguram, R. et al. (2001) nation because of Schizophrenia, World Psychiatric Psychiatric stigma across cultures: local validation in Association, Geneva. Bangalore and London. Anthropology and Medicine, 14. Sartorius, N. and Schulze, H. (2005) Reducing the 8, 71–87. Stigma of Mental Illness: A Report from a Global 29. Kadri, N., Manoudi, F., Berrada, S. and Moussaoui, D. Association, Cambridge University Press, Cambridge. (2004) Stigma impact on Moroccan families of patients with schizophrenia. Canadian Journal of 15. Hayward, P. and Bright, J. A. (1997) Stigma and Psychiatry, 49 (9), 625–629. mental illness:a review andcritique. Journalof Mental Health, 6 (4), 345–354. 30. Ritsher, J. B. and Phelan, J. C. (2004) Internalized stigma predicts erosion of morale among psychiatric 16. Link, B. G., Cullen, F. T., Struening, E. L. et al. (1989) outpatients. Psychiatry Research, 129 (3), 257–265. A modified labeling theory approach in the area of

338 SOCIAL CONSEQUENCES AND RESPONSES 31. Corrigan, P. W., Larson, J. E., Watson, A. C. et al. stigma of mental illness. British Journal of Psychiatry, (2006) Solutions to discrimination in work and hous- 190 (3), 248–254. ing identified by people with mental illness. Journal of 43. Thornicroft, G. (2007) Most peoplewith mental illness Nervous and Mental Disease, 194 (9), 716–718. are not treated. Lancet, 370 (9590), 807–808. 32. Angermeyer, M. C. and Matschinger, H. (2005) Labe- 44. Rusch, N., Angermeyer, M. C. and Corrigan, P. W. ling–stereotype–discrimination. An investigation of (2005) The stigma of mental illness: concepts, forms, the stigma process. Social Psychiatry and Psychiatric and consequences. Psychiatrische Praxis, 32 (5), Epidemiology, 40 (5), 391–395. 221–232. 33. Estroff, S. E., Penn, D. L. and Toporek, J. R. (2004) 45. Jacoby, A. (1994) Felt versus enacted stigma: a con- From stigma to discrimination: an analysis of com- cept revisited. Evidence from a study of people munity efforts to reduce the negative consequences of with epilepsy in remission. Social Science and Med- having a psychiatric disorder and label. Schizophrenia icine, 38 (2), 269–274. Bulletin, 30 (3), 493–509. 46. Crisp, A., Gelder, M. G., Goddard, E. and Meltzer, H. 34. Alem, A., Jacobsson, L., Araya, M. et al. (1999) How (2005) Stigmatization of people with mental illnesses: are mental disorders seen and where is help sought in a a follow-up study within the Changing Minds cam- rural Ethiopian community? A key informant study in paign of the Royal College of Psychiatrists. World Butajira, Ethiopia. Acta Psychiatrica Scandanavica Psychiatry, 4, 106–113. Supplement, 397, 40–47. 47. Department of Health (2003) Attitudes to Mental Ill- 35. Thara, R., Kamath, S. and Kumar, S. (2003) Women ness 2003 Report. Department of Health, London. with schizophrenia and broken marriages – doubly 48. Vaughn, G. (2004) Like minds, like mine, in Mental disadvantaged? Part II: family perspective. Interna- Health Promotion: Case Studies from Countries tional Journal of Social Psychiatry, 49 (3), 233–240. (eds S. Saxena and P. Garrison), World Health Orga- 36. de Toledo Piza, P. E. and Blay, S. L. (2004) Community nization, Geneva, pp. 62–66. perception of mental disorders – a systematic review of 49. Dunion, L. and Gordon, L. (2005) Tackling the attitude Latin American and Caribbean studies. Social Psy- problem. The achievements to date of Scotland’s ‘see chiatry and Psychiatric Epidemiology, 39 (12), me’ anti-stigma campaign. Mental Health Today, 955–961. 22–25. 37. Al-Krenawi, A., Graham, J. R., Dean, Y. Z. and 50. Jorm, A. F., Christensen, H. and Griffiths, K. M. (2005) Eltaiba, N. (2004) Cross-national study of attitudes The impact of beyond blue: the national depression towards seeking professional help: Jordan, United initiative on the Australian public’s recognition of Arab Emirates (UAE) and Arabs in Israel. Interna- depression and beliefs about treatments. Australian tional Journal of Social Psychiatry, 50 (2), 102–114. and New Zealand Journal of Psychiatry, 39 (4), 38. Littlewood, R. (1998) Cultural variation in the stig- 248–254. matisation of mental illness. Lancet, 352 (9133), 51. Mehta, N., Kassam, A., Leese, M. et al. (2009) Public 1056–1057. attitudes towards people with mental illness in Eng- 39. Crocker, J., Major, B. and Steele, C. (1998) Social land and Scotland, 1994–2003. British Journal of stigma, in The Handbook of Social Psychology, 4th Psychiatry, 194 (3), 278–284. edn (eds D. Gilbert, S. T. Fiske and G. Lindzey), 52. Graves, R. E., Cassisi, J. E. and Penn, D. L. (2005) McGraw-Hill, Boston, Massachusetts, pp. 504–533. Psychophysiological evaluation of stigma towards 40. Corrigan, P. W., Markowitz, F. E. and Watson, A. C. schizophrenia. Schizophrenia Research, 76 (2–3), (2004) Structural levels of mental illness stigma and 317–327. discrimination. Schizophrenia Bulletin, 30 (3), 53. Rose, D., Thornicroft, G., Pinfold, V. and Kassam, 481–491. A. (2007) 250 labels used to stigmatise people 41. Rose, D. (2001) Users’ Voices. The Perspectives of with mental illness. BMC Health Services Research, Mental Health Service Users on Community and 7,97. Hospital Care, The Sainsbury Centre, London. 54. Littlewood, R. (2004) Cultural and national aspects of 42. King, M., Dinos, S., Shaw, J. et al. (2007) The Stigma stigmatisation, in Every Family in the Land (ed. A. H. Scale: development of a standardised measure of the Crisp), RoyalSociety of Medicine, London, pp.14–17.

STIGMA AND DISCRIMINATION 339 55. Fabrega Jr, H. (1991) The culture and history of 67. Leong, F.T. and Lau,A. S. (2001) Barriers to providing psychiatric stigma in early modern and modern Wes- effective mental health services to Asian Americans. tern societies: a review of recent literature. Compre- Mental Health Services Research, 3 (4), 201–214. hensive Psychiatry, 32 (2), 97–119. 68. FabregaJr, H. (1991) Psychiatric stigma in non-Wes- 56. Shibre, T., Negash, A., Kullgren, G. et al. (2001) tern societies. Comprehensive Psychiatry, 32 (6), Perception of stigma among family members of indi- 534–551. viduals with schizophrenia and major affective dis- 69. Kleinman, A. and Mechanic, D. (1979) Some observa- orders in rural Ethiopia. Social Psychiatry and Psy- tions of mental illness and its treatment in the People’s chiatric Epidemiology, 36 (6), 299–303. Republic of China. Journal of Nervous and Mental 57. Stein, D. J., Wessels, C., Van Kradenberg, J. and Disease, 167 (5), 267–274. Emsley, R. A. (1997) The Mental Health Information 70. Phillips, M. R., Pearson, V., Li, F. et al. (2002) Stigma Centre of South Africa: a report of the first 500 calls. and expressed emotion: a study of people with schizo- Central African Journal of Medicine, 43 (9), 244–246. phrenia and their family members in China. British 58. Minde, M. (1976) History of mental health services Journal of Psychiatry, 181, 488–493. in South Africa. Part XIII. The National Council 71. Furnham, A. and Chan, E. (2004) Lay theories of for Mental Health. South African Medical Journal, schizophrenia. A cross-cultural comparison of British 50 (3F), 1452–1456. and Hong Kong Chinese attitudes, attributions and 59. Hugo, C. J., Boshoff, D. E., Traut, A. et al. (2003) beliefs. Social Psychiatry and Psychiatric Epidemiol- Community attitudes toward and knowledge of mental ogy, 39 (7), 543–552. illness in South Africa. Social Psychiatry and Psy- 72. Chou, K. L., Mak, K. Y., Chung, P. K. and Ho, K. chiatric Epidemiology, 38 (12), 715–719. (1996) Attitudes towards mental patients in Hong 60. Cheetham, W. S. and Cheetham, R. J. (1976) Concepts Kong. International Journal of Social Psychiatry, of mental illness amongst the rural Xhosa people in 42 (3), 213–219. South Africa. Australian and New Zealand Journal of 73. Chung, K. F., Chen, E. Y. and Liu, C. S. (2001) Psychiatry, 10 (1), 39–45. University students’ attitudes towards mental patients 61. Ozmen, E., Ogel, K., Aker, T. et al. (2004) Public and psychiatric treatment. International Journal of attitudes to depression in urban Turkey – the influence Social Psychiatry, 47 (2), 63–72. of perceptions and causal attributions on social dis- 74. Thara, R. and Srinivasan, T. N. (2000) How stigmatis- tance towards individuals suffering from depression. ing is schizophrenia in India? International Journal of Social Psychiatry and Psychiatric Epidemiology, Social Psychiatry, 46 (2), 135–141. 39 (12), 1010–1016. 75. Raguram, R., Raghu, T. M., Vounatsou, P. and 62. Dietrich, S., Beck, M., Bujantugs, B. et al. (2004) The Weiss, M. G. (2004) Schizophrenia and the cultural relationship between public causal beliefs and social epidemiology of stigma in Bangalore, India. distance toward mentally ill people. Australian and Journal of Nervous and Mental Disease, 192 (11), New Zealand Journal of Psychiatry, 38 (5), 348–354. 734–744. 63. Link, B. G., Yang, L. H., Phelan, J. C. and Collins, P. Y. 76. Desapriya, E. B. and Nobutada, I. (2002) Stigma of (2004) Measuring mental illness stigma. Schizophre- mental illness in Japan. Lancet, 359 (9320), 1866. nia Bulletin, 30 (3), 511–541. 77. Hasui, C., Sakamoto, S., Suguira, B. and Kitamura, T. 64. Corrigan, P., Thompson, V., Lambert, D. et al. (2003) (2000) Stigmatization of mental illness in Japan: Perceptions of discrimination among persons with images and frequency of encounters with diagnostic serious mental illness. Psychiatric Services, 54 (8), categories of mental illness among medical and non- 1105–1110. medical university students. Journal of Psychiatry and 65. Villares, C. and Sartorius, N. (2003) Challenging the Law, 28 (Summer), 253–266. stigma of schizophrenia. Revista Brasiliera Psiquia- 78. Sugiura, T., Sakamoto, S., Kijima, N. et al. (2000) tria, 25, 1–2. Stigmatizing perception of mental illness by Japanese 66. Ng, C. H. (1997) The stigma of mental illness in Asian students: comparison of different psychiatric disor- cultures. Australian and New Zealand Journal of ders. Journal of Nervous and Mental Disease, 188 (4), Psychiatry, 31 (3), 382–390. 239–242.

340 SOCIAL CONSEQUENCES AND RESPONSES 79. Kurihara, T., Kato, M., Sakamoto, S. et al. (2000) 91. Arkar, H. and Eker, D. (1992) Influence of having a Public attitudes towards the mentally ill: a cross- hospitalized mentally ill member in the family on atti- cultural study between Bali and Tokyo. Psychiatry tudes toward mental patients in Turkey. Social Psychia- and Clinical Neurosciences, 54 (5), 547–552. try and Psychiatric Epidemiology, 27 (3), 151–155. 80. Kurumatani, T., Ukawa, K., Kawaguchi, Y. et al. 92. Arkar, H. and Eker, D. (1994) Effect of psychiatric (2004) Teachers’ knowledge, beliefs and attitudes labels on attitudes toward mental illness in a Turkish concerning schizophrenia – a cross-cultural approach sample. International Journal of Social Psychiatry, 40 in Japan and Taiwan. Social Psychiatry and Psychia- (3), 205–213. tric Epidemiology, 39 (5), 402–409. 93. Coker, E. M. (2005) Selfhood and social distance: 81. Desapriya, E. B. and Nobutada, I. (2002) Stigma of toward a cultural understanding of psychiatric stigma mental illness in Japan. Lancet, 359 (9320), 1866. inEgypt. Social Science and Medicine, 61 (5),920–930. 82. Takizawa, T. (1993) Patients and their families in 94. Thornicroft, G., Brohan, E., Rose, D. et al. (2008) Japanese mental health. New Directions in Mental Global pattern of anticipated and experienced discri- Health Services, (60), 25–34. mination against people with schizophrenia. Lancet, 83. Goto, M. (2003) Family psychoeducation in Japan. 373, 408–415. Seishin Shinkeigaku Zasshi, 105 (2), 243–247. 95. Lai, Y. M., Hong, C. P. and Chee, C. Y. (2001) Stigma 84. Kim, Y. and Berrios, G. E. (2001) Impact of the term of mental illness. Singapore Medical Journal, 42 (3), schizophrenia on the culture of ideograph: the Japanese 111–114. experience. Schizophrenia Bulletin, 27 (2), 181–185. 96. Lee, S., Lee, M. T., Chiu, M. Y. and Kleinman, A. 85. Mino, Y., Yasuda, N., Tsuda, T. and Shimodera, S. (2005) Experience of social stigma by people with (2001) Effects of a one-hour educational program schizophrenia in Hong Kong. British Journal of Psy- on medical students’ attitudes to mental illness. Psy- chiatry, 186, 153–157. chiatry and Clinical Neurosciences, 55 (5), 501–507. 97. Crisp, A. H., Cowan, L. and Hart, D. (2004) The 86. Karim, S., Saeed, K., Rana, M. H. et al. (2004) College’s anti-stigma campaign 1998–2003. Psychia- Pakistan mental health country profile. International tric Bulletin, 28, 133–136. Review of Psychiatry, 16 (1–2), 83–92. 98. Crisp, A. (2004) Every Family in the Land: Under- 87. Al-Krenawi, A., Graham, J. R. and Kandah, J. (2000) standing Prejudice and Discrimination Against Peo- Gendered utilization differences of mental health ser- ple with Mental Illness, Royal Society of Medicine vices in Jordan. Community Mental Health Journal, Press, London. 36 (5), 501–511. 99. Link, B. G. and Cullen, F. T. (1986) Contact with the 88. Al-Krenawi, A., Graham, J. R., Ophir, M. and Kandah, mentally ill and perceptions of how dangerous they are. J. (2001) Ethnic and gender differences in mental health JournalofHealthandSocialBehavior,27(4),289–302. utilization: the case of Muslim Jordanian and Moroccan 100. Pinfold, V., Toulmin, H., Thornicroft, G. et al. (2003) Jewish Israeli out-patient psychiatric patients. Interna- Reducing psychiatric stigma and discrimination: eva- tional Journal of Social Psychiatry, 47 (3), 42–54. luation of educational interventions in UK secondary 89. Cinnirella, M. and Loewenthal, K.M. (1999) Religious schools. British Journal of Psychiatry, 182, 342–346. and ethnic group influences on beliefs about mental 101. Pinfold, V., Huxley, P., Thornicroft, G. et al. (2003) illness: a qualitative interview study. British Journal of Reducing psychiatric stigma and discrimination – Medical Psychology, 72Pt (4), 505–524. evaluating an educational intervention with the police 90. Loewenthal, K. M., Cinnirella, M., Evdoka, G. and force in England. Social Psychiatry and Psychiatric Murphy, P. (2001) Faith conquers all? Beliefs about the Epidemiology, 38 (6), 337–344. role of religious factors in coping with depression 102. Kingdon, D., Jones,R.and Lonnqvist,J.(2004) Protect- among different cultural-religious groups in the UK. ingthehumanrightsofpeoplewithmentaldisorder:new British Journal of Medical Psychology, 74Pt (3), recommendationsemergingfromtheCouncilofEurope. 293–303. British Journal of Psychiatry, 185, 277–279.

27 Taking inequality's measure: poverty, displacement, unemployment and mental health Kim Hopper Nathan Kline Institute for Psychiatric Research and Mailman School of Public Health, Columbia University, New York, USA When, in late eighteenth century Germany, Johann abnormal element in social life’ [1]. Modernity Peter Frank christened the ‘people’s misery’ as the domesticated poor relief, even as its industrial order ‘mother of disease’ he laid down the premise for much created the swelling ranks of the wandering of the research into the social determinants of health poor [2–6]. Novel configurations of disease and dis- that would follow in the next two centuries. What is order were also among its mixed bequest. not so evident in that formulation is the extent towhich By the mid-nineteenth century, studies in ‘social the people’s misery had become a social product, medicine’ had begun to sketch what would eventually instead of simply the caprice of heaven operating become coordinated analyses of the relations between through drought, crop failure, floods, war and plague. the institutions and practices of social life and pre- As the settled verities of a feudal world slowly unra- vailing patterns of health and illness. By and large, this velled, the social order of a nascent capitalist economy sprawling body of work has dealt with the varieties of took shape. Labour became increasingly mobile, live- hardship that are the predicament of the destitute: poor lihood less certain, families more dispersed. Custom hygiene and housing; lack of food, clothing and fuel; gave way to competition and the security of a manor- uncertain, ill-paid and dangerous work – or none at all; bound peasantry yielded to the risky lot of the landless exposure to industrial and domestic hazards; perva- proletariat. Penury spread, the condition of the depen- sive stress, and the ever-present threats of debt and dent poor grew more precarious and the informal violence; along with a devil’s gamut of attendant institutions of local aid repeatedly proved unequal to ‘vicious habits’ [7–12]. However, hardiness and resi- the need. Once the work of virtue, charity became liency have also been part of that story. Just a century increasingly rationalized as the state took over the after Frank’s proposal for a ‘medical police’ (or, as we burden of social assistance. No longer unusual, desti- would have it, a public health authority) was made, tution emerged ‘as a normal and not, as heretofore, an Emile Durkheim put sociology on the map with his Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

342 SOCIAL CONSEQUENCES AND RESPONSES study of Suicide (1897). The book is best known of distinction – class, status and power – identified by for Durkheim’s painstaking demonstration that the modernity’s theorists (Marx and Weber especially), differential rates of this most private of acts are also have proven remarkably durable. They continue to ‘social facts’; that suicide’s uneven prevalence was provide useful analytic distinctions (e.g. see Refe- a reflection of variation in the degree of integration rence [28]). That said, most of the research to be and cohesiveness that exists in communities. Less reported on here was conducted during (or just after) well-known is his observation on the protective the last quarter-century of the twentieth century, value of poverty. The poor, wrote Durkheim, are precisely the period during which modernity’s late less susceptible to suicide because, ‘schooled’ to do category-bending, boundary-breaking, time/space- without and disciplined by scarcity to want less, compressing incarnation – globalization –cameof they are less liable to disappointment: ‘Poverty age [29–33]. (Even before globalization’s ascent, protects against suicide because it is a restraint in some questioned the claim that modernity imposed itself’ ([13], p. 254). clear boundaries between realms of living and com- Most analysts, as this chapter will rehearse, rightly partmentalized domains of experience, Latour [34] persist in documenting and analysing the destructive prominent among them.) That temporal coincidence effects of inequality and deprivation on well-being is awkward for our purposes, because the impact of and mental health (e.g. see Reference [14]). However, globalization is likely to register in ways that will there are those even today who are struck by substantially rework the documentary record ‘resourcefulness, resiliency and adaptation’ among reviewed here. Globalization seems well on its way even those who, by material standards, are very poor to redrawing the contours of inequality and the or experience substantial adversity [15–20], though contexts of disadvantage, even as it reworks social the identified sources of that hardiness are often far ledgers of comparison [35,36]. Capital flight and cries from Durkheim’s throttled desire. In addition, outsourcing have greatly expanded the compass of while there are clear hazards – moral, analytic and competition. Within developed economies, too, job political – to overemphasizing the durable ‘agency’ holdings may ‘churn’ even as gross employment of the poor, recent attempts at reframing poverty (e.g. rates remain high, adding uncertainty and insecurity see References [21] to [26]) make it clear that a to work [37]. Ever more fluid and transient positions purely structural account will be a partial one. By the of belonging and citizenship, of measuring up and same token, analyses of social determinants of making out, may make the project of identity – of health that neglect the ‘pathologies of power’ written locating oneself as a person of consequence, within into the embedded interests and established practices a place and among a people reliably regarded as that sort risk and resources in mortality’s roulette [27] home – increasingly problematic [38–41]. Through- will badly misrepresent the social machinery of out this unsettled terrain, lasting states of dislocation health. and uncertainty are remaking the landscape of Two cautionary notes are in order before reviewing psychiatric risk and replotting the course of subse- the evidence of poverty’s toll. The first deals with quent disorder. Nowhere is this clearer than in the epoch. Modernity is the name given to that vertigi- epidemiologically emblematic figure of the second- nous period when tradition’s hold gave way and generation migrant [42]. familiar forms of livelihood and community slowly The secondcautionary note has to dowith language. eroded, when it seemed ‘everything solid melts into Easy recourse to such keystone nouns as poverty and air’ even as an ‘iron cage’ of rationality displaced disorder has also become problematic: the first, long-standing canons of justification and meaning. because research on both international and domestic Along with technical progress, economic expansion, fronts has repeatedly demonstrated the insufficiencies productive power and population growth, modernity ofincome-based measures alone (see, for example, the laid down fresh contours of inequality. With the UN’s Human Development Reports); the second, not notable additions of gender, race/ethnicity, some- only because enquiry into the toll of disadvantage times culture and lately disability, the original axes extends beyond the bounds of formal diagnoses [43],

TAKING INEQUALITY’S MEASURE 343 but also because the terrain of unmet need for psy- correctives, unwillingness to face the prospect of chiatric care remains poorly mapped. In low-income living with mental illness, not knowing where to turn countries especially, this is due to scarce resources, for help or what to expect, ease of access, reassuring multiple therapeutic options and persistent stigma lay counsel or clergy support, anticipatory stigma – all (see, for example, the Lancet series on global mental of these probably play a role in what is often, in health [44]). Even in more affluent economies, the practice, a process of muddling through [47–51]. delay between apparent need (or reported distress) and Therefore, if the initial advisory above alerts the initiating treatment contact remains substantial [45]. reader to the provisional character of the empirical Help-seeking is a social process, not a private deci- record assembled here, this second caution is meant to sion, managed by the networks (intimate and less so) suggest that even the categories used to organize, inwhichpeopleareembedded atthetimeofcrisis[46]. parse and take stock of that record may well be headed Initial resolve to seek alternative explanation and for substantial revision. 27.1 RECLAIMING THE SOCIAL Once prevalent, then long stranded on the outskirts of were opened to public scrutiny in ways that life inside a psychiatry dominated by neuroscience and genetics, the state hospitals had never been [62–65]. an interest in how social arrangements may shape the Much of that scrutiny proved unsparing and critical. rates and prospects of mental disorder has shown Users of services provided first-person accounts of resurgence of late. Two distinct phases can be identi- living with severe mental illness, ‘reframing’ taken- fied. An earlier search for psychosis-free zones of for-granted assumptions of need and care as rankling civilization having been discredited and largely aban- issues of power and control [66–69]. Progressive doned, postwar psychiatry found fresh cause for opti- elements in psychiatric rehabilitation raised the bar mism in community-based mental health instead, on aspiration, even as they came to question their own seeking ‘psychosocial’ factors that would influence paternalism, and a once-distinctive public health con- course and outcome and charting the commodious cern with social justice [70] elbowed its way into landscape of ‘stress’ [52] (So commodious, indeed, discussions of mental health policy that went well that on the service side a few discordant voices beyond formal medicolegal issues of confinement. worried about mission creep, a vaguely defined What had been largely symbolic protests against ‘community psychiatry’ charged with the mundane misrecognition – the stain and sequelae of stigma – anxieties and difficulties of the everyday (e.g. see found its neomaterial voice [71]. Like its counterpart Reference [53]).) Cross-cultural epidemiology was in the social determinants of health literature, the surely part of what prompted this call to look beyond record subsequently compiled makes for sobering diagnosis for determinants of prognosis [54–60]. But reading. Poverty traps people in dead-end prospects, on the homefront in Europe and the US, deinstitutio- lowers quality of life, shadows efforts to reclaim nalization was the historical provocation. Once moral standing, complicates social life, contributes returned to society, no longer confined to special- to poor course of illness and bulks large as a major purpose ‘abeyance mechanisms’ conveniently located impediment to recovery from severe mental illness: elsewhere [61], the lives of former inmates (and, more these are the recurrent themes in contemporary pressing still, those who would henceforth be spared attempts to take the measure of lives marked by long-term hospitalization) became the stuff of policy psychosis [72–77]. Where close documentation of decisions, both deliberate and inadvertent. Politics, everyday lives allows, ingenuity and resiliency may markets and social science crowded a stage once also appear in that picture (e.g. see References [78] owned by asylum superintendents. The consequences and [79]). of those decisions, intended and unforeseen (and the However, epidemiology’s charge is not simply to interests, assumptions and intentions behind them), track and assess the consequences of declared public

344 SOCIAL CONSEQUENCES AND RESPONSES policy. As the early chroniclers of modernity taught need to be unpacked when attempting to track the us, inequality’s reach is deeper and subtler. That impact over time of spatial concentrations of poverty inequality may be implicated – insidiously and cumu- and other forms of disadvantage associated with latively, as well as abruptly and spectacularly – in the it [91]. Doing so may raise tricky conceptual and aetiology (not merely the course) of psychiatric dis- methodological issues [92,93], but the promise is a order is the more ambitious claim of a resurgent social heady one: instead of banishing environments to the psychiatry. Indeed, the notion that certain species of distal and given – ‘the realm of the invariant and disadvantage (especially when severe and/or durable) potentially confounding’ ([94], p. 44) – it enlists should occasion vulnerability, distress, ill-adaptive context as an active ingredient in the social structuring compensatory behaviour and overt breakdown, and of risk. Like Durkheim before it, the new social do so in ways that comparative analysis should be able psychiatry seeks to understand not only how external to detect and interrogate, has long been a founding constraints limit choice and opportunity but also how article of faith in that project. After too long a mor- they become embodied as distinctive physiological atorium, renewed activity (even a ‘renaissance’) is variant, interpretive schema, and native bent [95]. It clearly under way [80,81] as the limits of a narrowly asks, to borrow from another sociologist, how local biomedical approach are reassessed [82]. Social psy- worlds shape a distinctive habitus [96] and how that chiatry echoes (even as it seems largely uninformed changes over the life course [97]. by, but see References [83] and [84]) the larger, better Simultaneously reflective of this larger epidemio- established enquiry into the ‘fundamental’ social logical project to reclaim context and to rethink how causes of health and illness, as well as that enterprise’s it works as a creative process and established condi- contentious debate over mechanisms of influence tion (e.g. [98]), and emblematic of public health’s [28,85–89]. In the new social psychiatry, no matter turn to the social sciences for framing theory, analytic how finely calibrated at the individual level, ‘risk’ tools and investigative methods (e.g. see Refer- must be embedded in appropriate social/cultural ences [28] and [99]), the reclamation of the social in matrices as well [90]. ‘Macro-level’ variables not only psychiatry began with its own version of the long- help shape the sort of people individuals become, but standing social gradient in mortality [11] – the durable also constrain the range (and bias the inclinations) of association of psychiatric disorder with socioeco- their choices. ‘Neighbourhood effects’, for example, nomic status. 27.2 SOCIAL CLASS AND LIFE STRESS: THE CHARACTER OF DANGER REVISITED It may have lagged behind infectious and chronic Meyer’s practice of charting a person’s ‘story’ diseases [100], but an epidemiology of mental through use of a ‘life chart’ (see Reference [108], disorders was recognizably social from the start. 418ff.) – also kept social psychiatry from too com- Beginning in the early twentieth century, field reports fortable a home in medicine. The analytic move to documented variations in onset and course of psycho- causal inference in the second half of the twentieth sis across classes and cultures, tribes and times, century, however, began by shamelessly borrowing the sacred and the secular [60,101–105]. Arduous from biology. Hans Selye’s protean notion of stress ‘community prevalence’ studies charted clinically (or ‘general adaptation syndrome’of 1936 [109] and na€ ıve psychiatric ailments in far-flung European com- 1956 [110]), along with the dynamic understanding munities [106]. Ecological studies mapped zones of of organism–environment fit captured by the term urban instability against rates of mental illness, on the adaptation (e.g. see Reference [111]), were imported rationale that pathologies of place might find reper- to put the social psychiatric enterprise on firm phy- cussion in disorders of mind [107]. A long-standing, if siological footing. In this way, researchers contrived minority, clinical interest in the everyday challenges to give an otherwise alien explanatory model a mea- that patients wrestled with – typified in Adolf sure of face-validity and legitimacy. Stress provided

TAKING INEQUALITY’S MEASURE 345 an empirically plausible mechanism, a way of ground- sustaining processes were subjected to extraordinary ing otherwise free-floating social science theorizing. demands – this is how they acquire ‘the character of (The same hybridizing move can be seen, nearly a danger’ – also meant that clinically informed correc- half-century later, in invocations of ‘social defeat’ and tive measures would move from the clinic to the ‘costly signalling’ – both constructs stemming from community. Medicine, Virchow argued, would have ethological studies – to explain elevated rates of to be a ‘social science’. psychosis among second-generation migrants [112] Stress proved the versatile mechanism needed to and refusal of services on the part of psychiatrically accommodate multiple parsings of the social field. disabled homeless women [113], respectively.) Occupational structures were deconstructed and For social science the shift from sociology to reconfigured, shifting relationships examined, role epidemiology entailed a double transformation. conflicts and ‘status incongruities’ unpacked, hierar- Social position had first to be converted into differ- chies decoded and eventfulness docketed. Various ential risk. Markers of distinction or disadvantage buffering agents (social analogues to host resistance) (class, status, gender, race/ethnicity, religion) had to had to be factored in as well, because the impact of the be shown to be reliable guides to a greater or lesser same event could be variable depending upon coping likelihood of adverse health outcomes. Theoretically and support resources. Compounding matters further, driven, descriptive epidemiology often sufficed to the varieties of distress occasioned by initial stressors demonstrate such discrepancies. The second, more could serve as secondary provocations in their own difficult, task involves both documentation and expla- right. However, such complexities lay well down the nation: to elucidate how unevenly distributed risks postwar research road. In the beginning, there was become actualized in the lives of social agents in ways simply a list: an early, hugely influential effort to that plausibly dispose them to psychiatric distress and itemize and rank ‘life events’. Holmes and Rahe’s disorder. Here, hypothesis-driven research design and ‘Schedule of Recent Experiences’ (1967) standar- increasingly refined statistical techniques of causal dized the inventory of such events and made it eco- inference played important roles. (A host of measure- nomical and efficient [119]. (It could be self-adminis- ment and definition problems, as we will see, would tered or made part of a brief interview.) Their work have to be solved as well.) would spawn a small industry of derivative studies, From the start, then, the ‘social’ in psychiatric attempting to sort out the valence, magnitude, impact epidemiology took in Durkheim’s concern with and source of life events, to winnow (e.g. subtracting threats to order and integration, Weber’s insistence ‘wife begins work’ or ‘family member added’) or that ‘life chances’ constrained life choices and Marx’s enlarge the catalogue itself (e.g. adding losing custody obsession with durable structures of inequality. The of children, experiencing discrimination or homeless- upshot, latent in the uneasy kinship between deviance ness) and to assess various resources and coping and pathology [114] and underlying persistent con- strategies that buffered their effects (e.g. see Refer- cerns with diagnosis and confinement as a means of ences [120] and [121]). Research soon revealed that it social control (e.g. see References [115] to [117]), was was negative life events that were harmful [122–126], a tense, Janus-faced legacy: on the one hand, seeking particularly events that were ‘fateful’ [127] or psychiatric counterparts to grim bills of mortality that ‘unscheduled’ [128]. Culturally scripted transitions consistently disfavoured the poor [11]; on the other, built into the life course, even if momentous and ever alert to the misuse of ‘medicalized’ rationales for daunting, would seem not at issue, but in late what were actually police or judicial actions. If the modern societies ‘transitions’ themselves were first found common cause with the larger struggle for becoming less well marked and scripted [129] and health equity [118], the second would resonate with difficulties in negotiating age-graded changes are not user accounts of mistreatment. Our concern here is uncommon [130]. the link to social determinants of health. As Leighton Event-driven assays were eventually complemen- and colleagues noted [104], Rudolph Virchow’s insis- ted by painstaking attempts to take stock of dis- tence that pathology developed when ordinary life- advantage’s unrelenting toll: the ‘ongoing difficulties’

346 SOCIAL CONSEQUENCES AND RESPONSES [131], daily hassles [132] and routine disruptions, the various agegroups [153–155]. (On other axes of social children’s world of financially strapped and quarrel- distinction – race/ethnicity and gender, especially – ling parents [133], and the demoralization of crushed the record remains stubbornly mixed.) In a represen- hopes and dead-end job prospects common in lower- tative US household study using a structured clinical class lives [134,135]. Such ‘structural strain’ [136] interview, for example, researchers in the National was a matter of social arrangements, not fateful Comorbidity Survey (NCS) found that prevalence happenings, and was built into the quotidian pulse of rates for most psychiatric diagnoses (panic-related family life, work and play [137]. Still other difficulties conditions and substance abuse excepted) were high- stemmed from competing commitments to valued est in the lowest income and education groups [156]. roles, such as those that working mothers or caregivers Findings also suggested that comorbidities, disorders routinely face [138,139]. stacked or clustered together, are linked to SES. However, even ‘events’ should not be seen – the A decade later, using both diagnoses and a novel still-common practice of resorting to ready-made latent class analysis of multivariate disorder group- checklists notwithstanding [140] – as self-contained ings, the NCS-Replication found the same ‘broadly stressors. Not only did they interact with individual consistent’ patterns of disorder and disadvantage, with coping styles but the details of their occurrence had to an especially prominent group of highly comorbid be embedded in meaning-making contexts if their major depression [157]. impact was to be properly gauged [141–143]. (The That the association between socioeconomic same divorce could be liberating for one partner, status and psychiatric disorder in the NCS was con- catastrophic for the other [137] – though it would sistently stronger for 12-month than for lifetime take more than mere listing to determine this.) (ever-diagnosed) prevalence suggests that disadvan- Delayed effect was also an issue; adverse conditions tage may affect the course as well as the onset of in early life could have an impact much later in disorders [158] (cf. Reference [11]). Likewise, buffer- life [144] (cf. Reference [145]) and a cumulative ing resources, from effective coping strategies to impact over time. Nonevents, expected but failing to assertive attitude to social allies, ‘tend to come in materialize, could also occasion stress [146], espe- packages’ [143]. However, such clustering also illus- cially when normative at a certain life stage [147]. trates a standing measurement problem. Many alleg- Finally, exploration of the contingencies that con- edly stressful events (gaining or losing a spouse, tained ‘prelude’ to and actual ‘occurrence’ of an event moving a home, changing jobs) have direct or indirect could reveal marked differences in the extent to which implications for one’s support network. Causal order- the person was implicated as a generative ing becomes problematic – in retrospect, it is difficult agent [148,149]. Even so, and illustrative of the sort to tell whether compromised support exacerbated the of confounding commonplace in the field, some peo- event’s impact or was one of the effects itself [159]. ple thought to ‘produce’ their own stressful events There are subtleties and easily missed contours to were subsequently shown to be suffering the late the relationship between life stress and psychological effects of early childhood adversity [150], such as disorder as well. A major negative event is commonly homelessness [151,152], though the mechanism via implicated in an initial episode of depression, but which delayed vulnerability operates remains open to recurrences can be set off by minor or moderate inquiry. events [160,161]. The line between clearly demar- These formidable difficulties and limitations not- cated events and the ‘hardships and problems’ that are withstanding, the yield of stress-driven research has routine features of a neighbourhood or way of life can been impressive. In particular, it has put beyond be difficult to draw in practice [143]. The psychic toll reasonable doubt the consistently found inverse of disadvantage may have to be reckoned not only in relationship between socioeconomic status (SES) ‘hierarchy stress’ or ‘status anxiety’ [88] but also (measured in various ways) and psychiatric distress in the ‘moral’ opprobrium of subordinate status – (whether assessed as global rates, scores on symptom/ withheld recognition, devalued parenting style, and impairment scales or specific disorders), and across locally prized linguistic and presentation proficiencies

TAKING INEQUALITY’S MEASURE 347 that score poorly with those calledfor in broader social each class bracket, owing to the added adversity of participation [24,162,163]. In the event, what matters minority status. Social selection recognizes the drag is not ‘allostatic load’ on an organism but the nagging that minority status imposes, on healthy and unhealthy sense of having failed to live up to (sub)cultural members alike, but sees the healthier member of ideals [164,165]. Sometimes, too, it can be a simple advantaged groups rising up the class ladder. Conse- matter of being ‘too poor to participate in the promise quently, it predicts lower rates of disorder among of a culture’ [166] (cf. Reference [167]). Like the ethnically advantaged groups within a class bracket, tendency for clinicians to mistake circumstantial owing to the diluting effect of healthy members demoralization for poor motivation [168], such com- stranded there. plexities call for close scrutiny and raise vexing In a longitudinal study in Israel (with the disadvan- measurement problems. (Again, from the start, there taged group Jews of North African heritage), the have been difficulties in unravelling potential con- results differed by type of disorder. Social selection founding of undesirable events and symptoms/effects was the better fit for schizophrenia, but social causa- of disorder [159,169].) tion better explained the patterns observed for anti- Most durable, however, has been the explanatory social personality and substance use disorders among conundrum typified by duelling theories of causation men and (though less well) for major depression and selection in accounting for the relationship among women [174]. Even with this elegant design, between socioeconomic status and psychiatric disor- a number of cautions are warranted. Minority status der. Each has its claim on the research literature may be constant over the life span but its social [11,170–172]. The first holds that structural disadvan- reception, meaning and consequence need not be. tage (perhaps in interaction with developmental, Moreover, shared colour can be a poor proxy for genetic or other ‘diathesis’ (or vulnerability) factors) diverse heritage within broadly defined minority ‘breeds’ disorder: it overloads the organism, produ- groups [175], and those within group distinctions have cing distress, hapless efforts to cope and eventual to be shown to have epidemiological import [176]. breakdown. The second (echoed in the health capital The expression and experience of devaluation in literature) sees ‘downward drift’ as the mechanism unequal registers of social distinction may be quite behind the association: poor psychological health, and different [94,177]. Discrimination involves institu- especially frank disorder, ill-equip someone for life in tional and interpersonal transactions in ways that class a competitive economy (or marriage market), and the does not [178]; class position lends itself to self-blame net effect of repeated failure to measure up is slippage in ways that discrimination may not [172]. down the class ladder. Using an ingenious research With extended practice, it became clear that the design, Dohrenwend and Dohrenwend [173] put the event-listing approach was fraught with additional question to a quasi-experimental test. The approach methodological difficulties as well, difficulties that builds upon the ‘natural experiment’ offered by ethnic could help explain puzzling inconsistencies in the assimilation. It rests on four assumptions: that while research record. Some of these were inherent in the class may be (somewhat) mutable over the life course, elicitation practice itself, the limitations built into ethnic identity is not; that in a relatively open class survey interviews as retrospective accounting exer- society, merit and effort will be rewarded with upward cises [179,180]. To begin with, ‘what counts’ as an social mobility, even if starting positions confer unfair event proved to be open to (sometimes idiosyncratic) advantage; that serious psychiatric conditions handi- interpretation, much more so than had been antici- cap that upward progress (and dispose downward); pated [181]. Some of this could be simple misunder- and that ethnic minority status operates independently standing (listing as the death of a close friend someone as a status-related disability (owing to prejudice and long absent from the respondent’s life; some due to discrimination). With an appropriately configured face-saving manoeuvres (minimizing one’s role in a social setting, then, competing predictions are divorce); some, to genuine confusion (does ‘laid off’ derived. Social causation predicts higher rates of refer only to job loss without cause?); some, due to disorder for ethnically disadvantaged groups within the constraints of survey interviews, which are rife

348 SOCIAL CONSEQUENCES AND RESPONSES with opportunity for ‘working misunderstandings’ their life circumstances and mental health. Details of between interviewer and respondent [182,183]. (With everyday life (e.g. financial and marital difficulties) respect to stressful life events, there is simply no good and recent events (usually within the past year) are reason to assume a common working knowledge of carefully elicited and ratings of severity made. The ‘the spirit of the event between subject and latter, using close readings of ‘biographically deter- investigator’ [140].) Recall is not only imperfect, but mined circumstances’, are meant to capture embedded biased as well by the subject’s own understanding of meaning and impact – the level of ‘threat’ that most what to search for [180], as well as by one’s current (culturally similar) others would have experienced. symptomatic state [184] (so-called ‘state-dependent’ (An extensive manual for the Life Events and Diffi- or ‘mood congruent’ recall [185]). Classification of culties Schedule (LEDS) set forth detailed rules for events may also be subject to the informant’s own defining, distinguishing constellations among and changing understanding of such things over time. rating elicited experiences.) Social class (measured Gender dispositions/practice may affect reporting by occupation, education and assets) was strongly proclivities [186]. Evendisciplined efforts to carefully related to depression: fully a quarter of the lower- circumscribe the domain of events, factor in collateral class residents (versus 5% of the middle-class counter- accounts (like pairs of sisters corroborating each parts) reported recent or chronic depression. Stressors other’s account of early abuse and neglect [187]) and were also distributed by class, with over half (54%) of have respondents themselves rate severity of impact the lower-class women (versus 30% of the middle- have their limitations [188]. Asking informants to class women) reporting a recent major event or recall and evaluate putatively stressful life events in ongoing life difficulty. However, what proved parti- the checklist protocol, it became clear, activated a cularly hazardous was the combination of lower class lively, behind-the-scenes negotiation, not only with status and the presence of one or more young children the rater and instrument but with one’s own past at home. Lower-class women in that life stage were as well. Not surprisingly, concordance between eight times more likely (40 versus 5%) to report recent self-report checklists and more structured inventories or chronic depression than their middle-class counter- (see below) is poor, less than 50% by some parts [197], and, in a gender-specific finding of the reckonings [140]. importance of support that others have gone on to The upshot, as Meyer had anticipated long ago, is explore (e.g. see Reference [198]), they were also half that in order faithfully to transcribe both occurrence as likely to report having someone to confide in. and effect, researchers would need an equipment The Camberwell study was thus able to show how, upgrade. Semi-structured, skilfully probing inter- for women, structure (social class), developmental views, co-constructing what were effectively perso- stage (early motherhood) and lack of support (an nalized narratives of stress (both eventful and available confidant) come together to create a dis- ongoing) and its management – not the dubious tinctive risk niche. In the process, it implicates class as economy of checklists and scores – would prove the a moderator of institutional role – here, the quality of preferred tools of enquiry [131,149,160,189–191]. marital relationships. Subsequent research would go Brown and Harris’ community study of depression on to identify specific features of unusually strong in women living in the lower-middle class neighbour- contextualized stressors. For Brown and colleagues, hood of Camberwell in south London set what many these were experiences that occasioned loss, entrap- still consider the standard [192]. Their original inter- ment and humiliation [150,199]). In a different view schedule was later refined and pared down approach to narrative production, Dohrenwend and somewhat [193], which made for a more readily colleagues [149] developed an instrument that rates adapted instrument (e.g. see References [194] events by domain of experience, severity of threat and to [196]). degree to which a respondent exercised control. Using intensive in-depth interviews that sought to Unlike LEDS, which combined situational and ‘contextualize’ stress accounts, the original research biographical data to render a composite or ‘global’ team queried over 300 local women with respect to account of threat, the narratives yielded by the

TAKING INEQUALITY’S MEASURE 349 structured events probe and narratives rating method consistently coupled with both traumatic (life-threa- (SEPRATE) are stripped of biographic factors (their tening) and other stressful events, and may be more contribution can then be separately analysed) and associated with events involving violence [201]. reduced to journalistic accounts of what happened. Inconsistent findings for cumulative traumatic events (Dohrenwend [188] compares them to courtroom by ethnicity may reflect differential sampling: whites testimony.) Psychiatric outpatients interviewed with were rated higher in an HMO sample than blacks in a this instrument were nearly 14 times as likely to have household sample [201,202]. Both exclude institutio- one or more of such recent events than their commu- nalized and homeless persons – in each of thesegroups nity controls [200]. African-Americans are overrepresented [203] – but Like the aetiological studies, latter-day (and more the former excludes uninsured people as well. Better rigorous) assessments of the demographic distribution off, more stable working folk may be less likely to be of stressful life events (SLEs) have found it useful to exposed to stressful events and ongoing difficulties disaggregate the event class [186]. Lower SES is than their less fortunate ethnic counterparts. 27.3 UNEMPLOYMENT AND UNCERTAIN WORK As we have seen, recurring, sometimes widespread, loss is the depth and persistence of self-blame among unemployment is among the distinctive bequests of the unemployed, even when it was clear that external modernity, although the term itself, carrying some- factors are at fault [216–220]. To a large extent, this thing of its contemporary meaning, did not come into remains true today. common use until the late nineteenth century [204]. In Although a voluminous literature will be drawn unplanned economies, as the social relations of formal upon, this is far from a comprehensive review. Even work came to reside in (and be defined by) the market- with that archive at hand, this chapter labours under place [205], employment’s prominence as a sorting some restrictions. Most of the relevant research per- mechanism in the production and replication of tains to capitalist economies. Evidence of the disas- inequality grew. However, if rank and role within the trous health effects of wrecked or disestablished market mattered, to be located outside it altogether planned economies, as in post-Soviet Russia, are not was to court social inconsequence (or ‘redundancy’). covered here, but may be found elsewhere (e.g. see (Official neglect and economic misrecognition of one References [88], [221] and [222]). Moreover, it is gendered version of such inconsequence – housework paid, formal work in the private or public sectors, – has long been a complaint of feminist economists; with jobs distributed through labour markets, with see Reference [206]). Because paid work figures so which we are concerned. Housework, whether under- highly as a source of identity, sense of competence, taken solely or in conjunction with paid work, raises everyday structure, self-respect and social regard, loss distinctive difficulties of its own (see References [223] of work and sustained periods of joblessness are, to [225] and also [197]). The vast, unregulated, ill- unsurprisingly, damaging to health and well-being. documented and variegated informal labour market, This conclusion, among the earliest and most durable whose flexible regimes of work may well figure highly of social epidemiology, has been sustained over sev- in the more favourable outcome of schizophrenia in eral generations of research and, since the 1930s, has less-developed economies [56], falls outside the com- been shown to include mental health consequences as pass of attention here (but see References [226] well [207–211]. True, local context, social security and [227]). schemes and culture all matter in taking the measure Like the coroner statistics on suicide Durkheim was of the meaning and consequences of unemploy- forced to rely upon, official unemployment figures ment [212–215]. Yet surely among the most striking may badly underestimate the true extent of the pro- features of the demoralizing toll of unemployment in blem. In the US, for example, such figures typically the original reports of the social consequences of job omit ‘discouraged’ workers (who have quit looking

350 SOCIAL CONSEQUENCES AND RESPONSES for work) and those in part-time employment who levels of insecurity and anxiety, even when official would prefer full-time jobs [228]. ‘Disguised unemployment rates remain low [37,258–260], and unemployment’ or underemployment – workers toil- across very different labour markets. As alluded to ing at jobs beneath their normal occupational posi- earlier, some job-related health-relevant stress tions or that are economically inadequate [229–231] – appears to be event-specific – e.g. the anticipation of is a further (and poorly documented) problem. Hence, being laid off [261] – while another type of stress is contrast effects at the categorical level may well be built into the job insecurity that is part and parcel of diluted. Some analysts, mindful of unemployment’s increasingly flexible work arrangements [262]. categorical limitations, ask us to consider the debil- The corrosive effects of the latter on ability to plan, itating effects of lifelong patterns of uncertain and already apparent to analysts in the 1930s [263], unfulfilling labour that make up ‘careers of labour have become more pervasive since then. The increas- market disadvantage’ [211]. Even with these caveats, ingly tenuous psychological distinction between the toll of unemployment is impressive. Using a borderline jobs and none at all suggests that – like variety of outcomes, research designs and methodo- residential stability (as we will see shortly) – employ- logical tools, empirical tests of the relation between ment status might best be conceptualized as a unemployment and mental health have shown sub- continuum [264]. stantial impacts on depression, social isolation, Much harder to document are the collective effects psychosomatic complaints, hospitalization, suicide, on communities ‘when work disappears’ [215], or is alcohol abuse and domestic violence ([232–238]). so thoroughly transformed that the new regime effec- Many of these conditions may affect the intensity tively debars whole classes of formerly able (and still and/or appropriate targeting of a job-search, which willing) men and women from steady jobs [265–267]. can then prolong the period of joblessness and com- Nor are laid-off workers and their families the only pound its effects, in a progressively worsening casualties: not only is it the case that their households gyre [239]. Adding moral insult to health injury is and dependants also suffer, but local cohorts of labour- the surplus stigma felt by some of the unemployed market-entry-aged youth are also deprivedof the loose [240,241] and by those enrolled in certain work-relief extended networks (the ‘weak ties’) that figure so schemes [242]. Adverse impacts have been documen- highly in early job search and placement [268,269]. ted for the mental health of spouses [243–245] and Painstaking, longitudinal work may be required to children [246], as well as for the quality of family and document such effects (e.g. see References [270] marital relationships [247–250]. Especially persua- to [272]) and the subjective readings of the actors sive arguments are made possible with longitudinal involved [273]. Such portraitures may best be analyses of ‘no-fault’ job loss. Such studies are able to assembled using documentary or ethnographic meth- disentangle the plausibly co-animating effects of poor ods. Long-stay field study makes possible the patient health or dysfunctional behaviour and unemployment: recording of ‘backstage’ adjustments, the private when the event being assessed is a plant closing or a anxieties behind public bluster and the unheralded company folding, and those impacted are the laid-off sources of solace and shame. At the individual level, workers, the causal arrow of effect is clear (e.g. see the social disconsolations of unemployment – the loss References [251] to [255]. of structure in everyday life, the sudden absence of a However, as noted above, it would be a mistake to source ofmeaningandidentity,thesurrenderofregular think that a clear psychological divide separates work instrumental purpose, marital discord or slippage into and nonwork [211]. When work is soulless and dirty, ‘unmarriageability’ altogether, gnawing feelings of when wages are set deliberately low because it is worthlessness and the low fug of self-pity, alleviated assumed that labourers will steal while on the job, perhaps by heroic drinking [274–276] – all this may be the very act of labouring can serve to confirm a man’s captured and calibrated by nuanced inventories of degraded sense of self [256,257]. When labour mar- psychological distress. However, full appreciation of kets ‘churn’ and businesses place increasing impor- their collective impact may require the broader lens tance on flexibility, workers can experience high that a documentary approach or community study

TAKING INEQUALITY’S MEASURE 351 affords (e.g. see References [277] to [279]). Where and frequently durable impact of unemployment is epidemiologists happen to be on hand to take their that it soon impoverishes many of its casualties. measure, elevated rates of depression and suicide are However, the second, more insidious fact about not uncommon as well (e.g. see Reference [280]). unemployment is its tendency over time to leach away As this miscellany of effects suggests, unemploy- at the spirit: to restrict, baffle and discourage the ment’s effects are registered not only in well-being – unemployed agent [211]. It disables the moral self, the impressive range of morbidity and mortality costs cutting the ground out from under its capacity to peer only partially sketched above – but also in the ability into the future with any sense of being able to impact to plan and capacity to take action of consequence – it. In the words of Appadurai [167], it sabotages the the damage to agency in short [211]. The first, obvious jobless worker’s ‘capacity to aspire’. 27.4 DISPLACEMENT AND RESIDENTIAL INSTABILITY The dislocations associated with poverty range from closing of psychiatric institutions without ensuring ‘planned’ mass displacements to clear the way for the functional equivalents of asylums or supportive commercial development, to relocations engineered residences meant that many vulnerable people were in the interest of ‘urban renewal’, to eviction for shunted to liveson‘the institutional circuit’ [113,286]. failure to pay rent, to what are increasingly common But that was only part of the picture. Reservoirs of (and variously precipitated) bouts of homelessness. affordable housing that had once provided respite, Vulnerability to displacement is one of the signal shabby but cheap, outside the hospital had dried up. forms of powerlessness among the poor, especially Research studies and the experience of service pro- when compounded by minority racial or ethnic viders would demonstrate the complex relationship status. The trauma that follows upon ‘tearing up of homelessness and mental health, reflecting not only neighbourhoods’ [281] or wholesale relocation in the differential vulnerability to homelessness but distinc- wake of natural disaster [166] can remain an open tive pathways to that state (and the likelihood of communal wound. staying there) as well. Both forced migrations (beyond the scope of the Conceptual and methodological problems harry work here, but see References [282] and [283]) and research in this field, especially as it relates to its frank homelessness lie at the extremes of displace- mental health dimensions [287]. As with work, a ment. In the developed world, the US experience may straightforward categorical definition of homeless- be the best documented. By the late 1980s, significant ness has proven elusive, and for good reason. Any numbers of Americans – by some estimates 3% of the candidates put forward are typically tailored for spe- population over a five-year period [284] – had found cific purposes, often those of bureaucracies charged themselves unable to sustain a place of their own and with establishing eligibility criteria for aid. The social compelled instead to resort to public or informal science case is driven less by pragmatism than by sources of shelter. Most of the informal (or ‘shadow’) conceptual utility. Here the distinction between sheltering was invisible and went undocumented, the socially productive processes and relatively durable commonplace clemencies of kinship and friendship conditions is also pertinent [94]. Rather than two stepping in where the market and welfare state classes of people, we face a continuum of residential failed [285]. However, the storied annals of contem- instability: a gallimaufry of housing arrangements in a porary homelessness draw upon decidedly different matrix of stability, adequacy, affordability and form- sources. Press coverage of the sometime antic entrea- ality. Where within this matrix locally meaningful ties and evident suffering of the ‘street homeless’ clusters may be inscribed to distinguish housed from effectively sealed the simplistic association in the homeless is largely a political decision. Little debate public’s mind – not supported by research – between surrounds those who take up residence on the streets, homelessness and deinstitutionalization. True, the parks, transportation depots or subways, or who

352 SOCIAL CONSEQUENCES AND RESPONSES successfully broker entry to designated shelters. in-time assays of the population will be distorted (Merely applying for shelter triggers the application by overrepresentation of long-stayers (those more of prevailing eligibility criteria.) At the other end of likely to be caught by point-in-time methods). If the the continuum are those stably ‘doubled up’ situations factors associated with long stays in shelter include in which unrelated people (or people outside the sparse or depleted social networks or practical diffi- culturally prescribed model of family) share a dwell- culties in managing one’s household, for example, ing unit. In between these two extremes are informal one would expect persons with severe psychiatric religious-based shelters for the homeless, domestic and substance abuse disorders to be disproportio- violence sanctuaries, institutional options (such as nately among the ranks of ‘chronic’ homeless. Admin- hospitals or jails) where otherwise stranded people istrative records in some jurisdictions confirm that may be housed because more appropriate placements expectation [288–291] and the testimony of service and more conventional but precarious housing situa- expertise concur [292]. In programmes designed to tions including overcrowded dwellings and unstable serve the long-term homeless, medical comorbidities doubled-up arrangements are unavailable. are also common [293]. Duration of homelessness is Again, practice trumps category: people commonly also a strong predictor of mortality, even when move among these different housing conditions, so adjusted for age and disability [294], with HIV/AIDS that a temporary dwelling place should not be con- playing a strong role for some populations in some fused with a class of homeless people. In some US cities [295–297]. cities, a substantial number of homeless individuals As is true of risk and social context more gener- circulate frequently among the streets, short-stay ally [85], the factors that make for vulnerability to hotels, doubled-up arrangements and institutions displacement and durable homelessness are a combi- (including jails, hospitals and detox facilities). Home- nation of personal and structural [203,298,299]. As a lessness in this broader sense is characterized by national phenomenon (or ‘social fact’), homelessness instability across a variety of impermanent, often may best be understood as a structural problem: makeshift, living situations. Enquiry into the lived increasing numbers of poor people pursuing declining experience of displacement often confronts a moving numbers of low-cost housing units. A game of musical target. As with unemployment, research on home- chairs results, where some portion of the population is lessness tends to be more narrowly and expediently left without a home (or crowding another’s lap). By focused. Difficulties identifying homeless people who default, the losers are the most vulnerable, and that are doubled-up, or interviewing those in institutional vulnerability may be due to extreme poverty, age, facilities, or tracking those whose situations change race/ethnicity, disability, substance abuse or badly precipitously, have resulted in operationalized defini- frayed social networks. Again risk factors dispose, tions of homelessness that are keyed to sheltered but do not determine. Put differently, when there was a populations, certain users of services (e.g. soup sufficient stock of affordable housing available to kitchens) without a place of their own, and people people living at the edge of respectability, people living on the streets or in encampments. who found it difficult or distasteful to fit in were Methodological difficulties compound the concep- consigned instead to this ‘social and cultural limbo’, tual ones. Epidemiologists distinguish between point and reliably found their way there. Indistinguishable prevalence, or the number of people in a given con- in the main from their homeless counterparts decades dition at a particular point in time, and period pre- later, they made up the bulk of the residential valence, or the number of people experiencing that hotel population in American cities in the earlier condition during some stipulated period. If some twentieth century [300]. Even among the structurally people experience a condition only briefly, as with displaced today, bracing displays of resourcefulness homelessness, period prevalence may be many times and self-direction can make facile assumptions about greater than point prevalence. Further, if thosewho are pathology, disaffiliation and homelessness mislead- only briefly or episodically homeless differ from those ing. It may take ethnographic accounts to see this, who remain so for extended periods, then point- however [301,302].

TAKING INEQUALITY’S MEASURE 353 Reported rates of mental illness tend to be higher workers or political refugees make population gen- among homeless populations than among those eralizations ill-advised [290,303–305]. housed. This is not to say, for reasons outlined Gender differences may also apply, for reasons that above, that mental illness causes homelessness. (Nor, again illustrate the intersection of clinical and cultural indeed, is it any longer the case, in some US factors. Homeless men are less likely than homeless jurisdictions, that mental illness invariably prolongs women without children to show evidence of severe shelter stays, for reasons that illustrate the bureau- mental illness, a fact reflecting the operation of several cratic logic behind apparent risk profiles. People factors. As noted earlier, doubling up with others is the with certified disabilities are entitled to more sub- most common form of shadow shelter and the last stop stantial income support in the US; that fact, coupled before literal homelessness. Women are typically with a targeted policy of aggressive outreach, may more versatile and valued as household members. actually enhance the chances of a nonkin mediated They also do the ‘work of kinship’ – their investment exit from shelter if one carries a diagnosis of severe in (and expected return on) the social capital of family psychiatric disorder.) Additional methodological dif- ties is greater and more durable than that of men. ficulties – defining the population, sampling ade- Cultural thresholds of tolerance for dependency and quately, measuring appropriately, deciding between difficult behaviour seem to be set higher for women instruments and clinical judgement, avoiding con- than men, permitting them to stay with family and tamination of symptom profiles by lack of sleep and friends under circumstances where men would be unrelenting vigilance – further complicate the task of asked to leave [301,306]. Women who eventually untangling the association of mental illness and exceed thresholds may be expected to show substan- homelessness. Troubling behaviour that would ordi- tially more disturbed or disruptive behaviour than narily betoken psychiatric disorder may, in the case their male counterparts ejected earlier. (Note that of the street-dwelling homeless, reflect the influence the types of mental illness most common among of an unusually stressful environment, lack of access women are less often associated with threats or to washrooms (hence ‘poor hygiene’ and ‘a dishev- violent behaviour than those common among men.) elled appearance’) or strategic adaptations to the Finally, the population of officially ‘single’ homeless hazards of street life (certain bizarre behaviours may women may include a fair number of mothers who function as a defence against unwanted attention). have lost custody of their children due to mental Then, too, mental health problems vary in intensity illness [307]. and duration, so deciding whether lifetime preva- For all these reasons, reported rates of mental ill- lence or point-in-time assays are more appropriate ness among homeless populations have proven notor- will substantially affect results. Even if the assess- iously inconsistent. After culling reported results for ment takes place during an episode of homelessness, minimal methodological adequacy in study design it is difficult to say whether the measured psychiatric and analysis, one of the more sophisticated assess- disability is a cause, consequence or correlate of the ments of rates of current psychiatric disorder, sub- homelessness. Finally, in the US, provisions for stance abuse, and dual diagnosis [308] produced the homeless poor, the demographic characteristics ranges of prevalence estimates that look like of subpopulations, local driving forces and feeder Table 27.1. (The authors cautioned, however, that this institutions, controlling legal authority and thus pat- terns of homelessness are all extremely diverse. Table 27.1 Current mental health and substance use (Like gambling, it is safe to assume that it exists problems (urban adults aged 18 þ ) everywhere, but the social response to it radically Severe/ Any shapes its visibility, magnitude and place in the Any axis 1 persistent substance makeshift economies of the poor.) Regional differ- disorder axis 1 use Dual Dx a ences, temporal shifts, relative proportions of 0.28–0.71 0.15–0.26 0.30–0.73 0.12–0.26 families (intact and otherwise), unattached adults, tramping adolescents and undocumented transient a Severe mental disorder þ axis 1 substance use disorder.

354 SOCIAL CONSEQUENCES AND RESPONSES was an ‘exploratory’ assessment.) As a general rule, of group-wide analysis of variables across people) the better the methodological rigour, the lower the suggests, however, that there may be marked differ- estimate. ences in adaptive resiliency across subgroups Regardless of the reasons for their displacement – within homeless families [320]. Similarly, the foster-care graduate, runaway, ejected from house- much-relied-upon (and thus exhausted) social net- hold – homeless youth seem to be at especially high works of homeless families seen at the time of risk for mood disorders, suicide attempts and post- applying for shelter [321] had returned to configura- traumatic stress disorder (PTSD), especially those tions virtually indistinguishable from those of their who have spent time contending with the depredations housed counterparts when those same (and since of life on the street [309]. Behavioural problems and rehoused) families were interviewed five years substance use are more common still. Again, more later [322]. The same research suggests that the rigorous assessments tend to produce lower estimates long-term impact (55 months after shelter entry) on for some disorders, though not depression [310]. Here children and adolescents, using a variety of clinical considerable ambiguity clouds the causal picture. and social measures, was found to be minimal, with That survival on the street may necessitate a variety the possible exception of those children who had first of high-risk behaviours (sex work and favours, drug experienced homelessness at a very early age. Both use and trafficking) seems scarcely in doubt; not the formerly homeless and their poor but housed surprisingly, then, high rates of physical and sexual peers fared badly when scored against cognitive and victimization are commonly reported (e.g. see Refer- educational norms [323]. At follow-up, too, more ences [311] and [312]). recent life events (violence at home or in the neigh- Among homeless families, strikingly different bourhood) seemed more ingredient to mental health results are obtained. Indeed, the most common and than a history of homelessness [324]. consistent finding from second-generation research on If obtaining reliable estimates of disorder among this population in the US (from the early 1990s the homeless poor is tricky, teasing out causal infer- onward) was the documentation of elevated levels of ence from the tangled web of association between various problems among low-income children, displacement and pathology (or other risk factor) is whether homeless or housed, when compared against more difficult still. Certain ground rules apply: rates or normative data [313]; poverty was what they had in experiences must be shown to be higher among (some) common [314,315]. Current rates of disorders diag- homeless persons than among their merely poor coun- nosed using DSM-III-R criteria, for example, were terparts; the condition must precede initial experi- found to be identical in a Worcester sample of home- ences of homelessness; and plausible mechanisms can less and low-income children eight years and be proposed to explain how the disabilities associated older [316]. Inconsistency dogs other findings, with with that disorder can (under certain circumstances) coeval studies reporting, for example, that elevate risk of homelessness. Psychiatric disorders ‘externalizing’ disorders (e.g. displays of aggressive such as schizophrenia meet such criteria; so do early behaviour) are more common among homeless pre- childhooddisruptionslikefostercareplacement[325]. schoolers than low-income housed children, but However, even here, effective risk is socially ‘internalizing’ disorders (e.g. depression, anxiety) are embedded: in a tight rental housing market, a history not [317]; companion reports find the opposite [318]. of lax rent payments and/or unconventional appear- Reports of the impact of homelessness on school- ance or behaviour may render someone a less desir- related performance are divided by educational able tenant. In more forgiving rental markets, such reforms instituted as part of the Stuart B. McKinney behaviour may be overlooked or tolerated [326]. Homelessness Assistance Act (1987): those con- Landlords seeking to improve rental revenues provide ducted before the legislation found substantial defi- a plausible mechanism for selective displacement of cits; those after (e.g. see Reference [319]) generally ‘difficult’ tenants. (Relatedly, the disorder and threa- found no differences in absenteeism or academic tening atmosphere associated with warehouse-type performance [313]. Person-centred analysis (instead shelters can make people acutely tuned to menace

TAKING INEQUALITY’S MEASURE 355 unlikely to use such facilities, ‘choosing’ instead to it; signs should not be mistaken for underlying fend for themselves on the street.) Similarly, early process.) childhood events may exert their destabilizing effect Scrupulous documentation of more recent residen- over time in a variety of ways. Foster care, for tial histories would seem to offer one way to explore example, could signify developmental damage to the dynamics of homelessness, but such histories are coping abilities, the loss of kin-based support later difficult to reconstruct reliably. Ironically, first-person in life or the sudden withdrawal of public support at an accounts can overplay the role of disability [328], age when few are prepared to be self-sufficient – or all ignoring the fact that such conditions had long been of the above. present in one’s life without causing homelessness. Other conditions – depression and demoralization As noted earlier, the negotiation of life transitions especially – are as likely to be accompaniments or (either expected or unexpected) can be especially consequences of the trauma of displacement and tricky, and their difficulties are compounded when homelessness. As we have seen, studies are divided mental illness intrudes. The same seemed to be true, on the effects of homelessness on children, some early on, of transitions from out of homelessness. showing developmental delays and other damage, First-generation longitudinal studies showed that others not, with still others seeking additional expla- although many people with severe mental illness natory value in differential resiliency within that managed to leave the streets and shelters, such unas- population. Careful research practice (the inclusion sisted ‘exits’ tended to be short-lived, shallow and of comparison groups of poor but housed families) unstable, such that returns to homelessness were all shows that the general rule with families, however, is but ensured [329]. Short-term intensive case manage- that it is the (antecedent and enduring) poverty – ment services provided one promising targeted inter- along with its multiple depredations – not home- vention [330]. More recent assessments of the staying lessness per se that is at work. (As Rutter has else- power of supported housing placement, even when where argued [327]: It isn’t the ‘brokenness’ of the intervention takes place without interim transi- homes that is so damaging so much as the ‘family tional housing, have been much more optimistic (e.g. discord and conflict’ that precedes and precipitates see Reference [331]). 27.5 CONCLUSION: RETHINKING POVERTY If useful measures of social distinction must be accounts of migration and risk of psychosis (e.g. see theory-driven [94,332], the same is true for poverty Reference [342]). and embodiment [81,97,333–335]. Acutely tuned to With a few notable exceptions (e.g. see Refer- some developments in the social sciences (e.g. social ences [343] and [344]), more radical attempts to capital, chiefly of the communitarian variety [336]), reframe poverty and disadvantage have gone largely public health has been slow to make use of recent untapped by social psychiatry. Owing to its political efforts, coming largely from the contentious world of embrace by the European Union and emphasis on international aid, to reframe poverty. Social psychia- full participation as a claim of citizenship, ‘social try is no exception. This is not to discount striking exclusion’ [345] may be more readily folded within progress made in domains traditionally more the standard assays of deprivation or stigma (e.g. see conversant with the clinical or place-based epide- References [51], [89] and [346]). However, more miology – the lasting impact, compound and cumu- thoroughgoing efforts to rethink the social wages of lative risks of childhood adversity and its impact inequality would compel us to redraw the conven- over the life course (e.g. see References [337] tional metes and bounds of poverty, cast a sceptical to [339]); innovative approaches to urbanicity, eye on claims of irreversibility and take a hard look at with attention to both built and neighbourhood how public policy reproduces (whether inadvertently environments [98,340,341]; and iconic, late-modern or by design) social inequalities in health. Income,

356 SOCIAL CONSEQUENCES AND RESPONSES work and education are not ignored, but neither are troubled brief but also of an older fallible self, whose their limitations. Instead, these heterodox approaches rough brush with misfortune can mean: use metrics grounded in human development, make- shift livelihoods, the exercise of agency and an ... a new beginning, a raid on the inarticulate engaged, contentious view of culture as an ongoing With shabby equipment, always deteriorating. argument (not always voiced as such) about what it In the general mess of imprecision of feeling, means to be one of us (e.g. see UN Human Devel- Undisciplined squads of emotion. opment Reports 1990–2009 and References [347] (Eliot, East Coker) and [348]). The arguments advanced and the frame- Eavesdropping upon the exertions of such selves – works laid out can be distinctive, even (viewed from as they seek refuge, confront challenge, forge mean- the once-calm centres of functionalist sociology or ing, offer solace and slog through an unforgiving free-market economics) eccentric. Measurement pro- everyday, unbending if not proud – is one of the blems are admittedly formidable: practice and pro- offbeat practices of the new poverty studies. Social cess are harnessed alongside outcome and achieve- psychiatry would be well-advised to consult these ment; meaning vies with functioning (‘Q-squared’ alternative writs of suffering. research methods are a practical necessity). However, after all that has been rehearsed in this chapter about the force of structure, the pluralities of disadvantage REFERENCES and the many varieties of scuttled agency, their appeal is undeniable. Arguably, they represent the 1. Salter, F. R. (ed.) (1926) Some Early Tracts on Poor difficult next step in reconceiving ‘the social Relief, Methuen, London. world’ [94]. They range from calls to reintegrate the 2. Jusserand, J. J. (1920) English Wayfaring Life in the claims for recognition into older demands for redis- Middle Ages, Revised edition, Ernest Benn, London. tribution [23,349]; to research programmes bent on 3. Davis, N. Z. (1968) Poor relief, humanism, and heresy: re-theorizing approaches to well-being in developing the case of Lyon. Studies in Medieval and Renaissance countries [25], especially those that reappraise the History, 5, 217–275. realm of resources [350]; and to the cottage industry 4. Braudel, F. (1982) Civilization and Capitalism, that has grown up around Amartya Sen’s 15–18th Century. Vol. II. The Wheels of Commerce, ‘capabilities’ approach, with its emphasis on real Harper & Row, New York. opportunities to achieve locally valued ‘beings and 5. Lis, C. (1986) Social Change and the Labouring Poor: doings’ (e.g. see References [21], [22], [351] Antwerp, 1700–1860, Yale University Press, New and [352]). For the most part, they remain unknown, Haven, Connecticut. and their potential utility untested, within the ranks 6. Adams, T. M. (1990) Bureaucrats and Beggars, of social psychiatry. Suffice it to mention that Oxford University Press, New York. attempts to derive any implications they may 7. Sigerist, H. E. (1943) Civilization and Disease, hold for reconceptualizing aetiology and recovery University of Chicago, Chicago, Illinois. are in their infancy (e.g. see References [76], [353] 8. Rosen, G. (1947) What is social medicine? A genetic and [354]). history of the concept. Bulletin of the History of In all of this, too, it may not be too much to detect a Medicine, 21, 674–733. reawakened keenness for dynamic process in aetiol- 9. Rosen, G. (1973) Disease, debility and death, in The ogy. The huge shadows thrown by time and history, Victorian City: Images and Realities, vol. 2 (eds H. J. the swords of circumstance and chance, are well Dyos and M. Wolf), Routledge & Kegan Paul, London, caught in the new social psychiatry, in ways that pp. 625–667. both pay tribute to forebears and (provisionally, 10. Brockington, C. F. (1961) The history of public partially) reach out to fresh attempts to rethink the health, in The Theory and Practice of Public social. In the interstices of its dry texts, one may also Health (ed. W. Hobson), Oxford University Press, detect the faint presence, not only of resiliency’s London.

TAKING INEQUALITY’S MEASURE 357 11. Susser, M., Watson, W. and Hopper, K. (1985) Sociol- 28. Cockerham, W. (2007) Social Causes of Health and ogy in Medicine, 3rd edn, Oxford University Press, Disease, Polity, New York. New York. 29. Giddens, A. (1990) The Consequences of Modernity, 12. Porter, R. (1997) The Greatest Benefit to Mankind: A Stanford University Press, Stanford, California. Medical History of Humanity, Norton, New York. 30. Harvey, D. (1990) The Condition of Postmodernity, 13. Durkheim, E. (1951) Suicide, Free Press, Glencoe, Blackwell, London. Illinois, First published 1897. 31. Beck, U. (1992) Risk Society: Towards a New Mod- 14. Patel, V. and Kleinman, A. (2003) Poverty and com- ernity, Sage, Thousand Oaks, California. mon mental disorders in developing countries. Bulletin 32. Appadurai, A. (1996) Modernity at Large: Cultural of the World Health Organization, 81, 609–615. Dimensions of Globalization, University of Minnesota 15. Biswas-Diener, R. and Diener, E. (2001) Making the Press, Minneapolis, Minnesota. best of a bad situation: satisfaction in the slums of 33. Castells, M. (2000) The Rise of the Network Society, Calcutta. Social Indicators Research, 55, 329–352. 2nd edn, Wiley-Blackwell, New York. 16. Luthar, S. S. (ed.) (2003) Reslience and Vulnerability, 34. Latour, B. (1993) We Have Never Been Modern, Cambridge University Press, New York. Harvard University Press, Cambridge, Massachusetts. 17. Camfield, L. and McGregor, A. (2005) Resilience and 35. Stiglitz, J. E. (2002) Globalization and Its Discon- well-being in developing countries, in Handbook for tents, Norton, New York. Working with Children and Youth: Pathways to Resi- 36. Munck, R. (2004) Globalization and Social Exclusion, lience across Cultures and Context (ed. M. Ungar), Kumarian Press, Bloomfield, Connecticut. Sage, Thousand Oaks, California, pp. 189–210. 37. Hacker, J. S. (2006) The Great Risk Shift, Oxford 18. Collins, D., Morduch, J., Rutherford, S. and Ruthven, University Press, New York. O. (2009) Portfolios of the Poor, Princeton University 38. Ignatieff, M. (1984) The Needs of Strangers, Anchor, Press, Princeton, New Jersey. New York. 19. Wikan, U. (1996) The nun’s story. American Anthro- 39. Ong, A. (1999) Flexible Citizenship: The Cultural pologist, 98, 279–289. Logics of Multinationality, Duke University Press, 20. Canvin, K., Marttila, A., Burstrom, B. and Whitehead, Durham, North Carolina. M. (2009) Tales of the unexpected? Hidden resilience 40. Giddens, A. (1991) Modernity and Self-Identity, Stan- in poor households in Britain. Social Science and ford University Press, Stanford, California. Medicine, 69, 238–245. 41. Bauman, Z. (2007) Liquid Times: Living in an Age of 21. Sen, A. (1999) Development as Freedom, Anchor, Uncertainty, Polity Press, Oxford. New York. 42. McKenzie, K., Fearon, P. and Hutchinson, G. (2008) 22. Nussbaum, M. C. (2000) Women and Human Devel- Migration, ethnicity and psychosis, in Society and opment: The Capabilities Approach, Cambridge Psychosis(edsC.Morgan,K.McKenzieandP.Fearon), University Press, New York. Cambridge University Press, New York, pp. 143–159. 23. Fraser, N. and Honneth, A. (2003) Redistribution or 43. Wakefield, J. C. (1999) The measurement of mental Recognition? A Political–Philosophical Exchange, disorder, in A Handbook for the Study of Mental Verso, London. Health (eds A. V. Horwitz and T. L. Schied), Cam- 24. Sayer, A. (2005) The Moral Significance of Class, bridge University Press, New York, pp. 29–57. Cambridge University Press, New York. 44. Prince, M., Patel, V., Saxena, S. et al. (2007) No health 25. Gough, I. and McGregor, J. A. (eds) (2007) Wellbeing without mental health. Lancet, 370, 859–877. in Developing Countries: From Theory to Research, Cambridge University Press, New York. 45. Wang, P. S., Berglund, P., Olfson, M. et al. (2005) Failure and delay in initial treatment contact after first 26. Burds-Sharps, S., Lewis, K. and Martins, E. B. (2008) onset of mental disorders in the National Comorbidity The Measure of America: American Human Develop- Survey Replication. Archives of General Psychiatry, ment Report 2008–2009, SSRC/Columbia University 62, 603–613. Press, New York. 46. Pescosolido, B. A. and Boyer, C. (1999) How do 27. Farmer, P. (2003) Pathologies of Power, University of people come to use mental health services? Current California Press, Berkeley, California.

358 SOCIAL CONSEQUENCES AND RESPONSES knowledge and changing perspectives, in A Handbook 59. Hopper, K. (1991) Some old questions for the new for the Study of Mental Health (eds A. V. Horwitz and cross-cultural psychiatry. Medical Anthropology T. L. Scheid), Cambridge University Press, New York, Quarterly (New Series), 5, 299–330. pp. 392–411. 60. Hopper, K. (2008) Outcomes elsewhere: course of 47. Pescosolido, B. A. Gardner, C. B. and Lubell, K. M. psychosis in ‘other cultures’, in Society and (1998) How people get into mental health services: Psychosis (eds C. Morgan, K. McKenzie and P. stories of choice, coercion and ‘muddling through’ Fearon), Cambridge University Press, New York, pp. from ‘first-timers’. Social Science and Medicine, 46, 198–215. 275–286. 61. Mizruchi, E. H. (1987) Regulating Society: Beguines, 48. Link, B. G. and Phelan, J. (1999) The labeling theory Bohemians and Other Marginals, University of Chi- of mental disorder (II): the consequences of labeling, cago Press, Chicago, Illinois. in A Handbook for the Study of Mental Health (eds A. 62. Scull, A. T. (1977) Decarceration: Community Treat- V. Horwitz and T. L. Scheid), Cambridge University ment and the Deviant – A Radical View, Prentice-Hall, Press, New York, pp. 361–376. Englewood Cliffs, New Jersey. 49. Phelan, J. C. (2002) Genetic bases of mental illness – 63. Estroff, S. E. (1981) Making It Crazy, University of a cure for stigma? Trends in Neuroscience, 25, California Press, Berkeley, California. 430–431. 64. Brown, P. (1986) Mental Health Care and Social 50. Corrigan, P. W. (ed.) (2003) On the Stigma of Mental Policy, Routledge & Kegan Paul, London. Illness: Practical Strategies for Research and Social 65. Mechanic, D. and Rochefort, D. A. (1990) Deinstitu- Change, American Psychological Association, tionalization: an appraisal of reform. Annual Review of Washington, DC. Sociology, 16, 301–327. 51. Thornicroft, G. (2006) Shunned: Discrimination 66. Chamberlin, J. (1978) On Our Own, Hawthorn, New against People with Mental Illness, Oxford University York. Press, New York. 67. Kaufman, C. L. (1999) An introduction to the mental 52. Leff, J. (2008) Climate change in psychiatry: health consumer movement, in A Handbook for the periodic fluctuations or terminal trend, in Society Study of Mental Health (eds A. V. Horwitz and T. L. and Psychosis (eds C. Morgan, K. McKenzie and P. Schied), Cambridge University Press, New York, pp. Fearon), Cambridge University Press, New York, 493–507. pp. 11–22. 68. Estroff, S. E. (2003) Subject/subjectivities in dispute: 53. Dunham, H. W. (1965) Community psychiatry: the the poetics, politics and performance of first person newest therapeutic bandwagon. Archives of General narratives of people with schizophrenia, in Schizo- Psychiatry, 12, 303–313. phrenia, Culture and Subjectivity: The Edge of 54. World Health Organization (WHO) (1973) Report of Experience (eds J. H. Jenkins and R. Barrett), Cam- the International Pilot Study of Schizophrenia, vol. 1, bridge University Press, New York, pp. 282–302. World Health Organization, Geneva. 69. Hornstein, G. A. (2009) Agnes’ Jacket, Rodale, New 55. World Health Organization (WHO) (1979) Schizo- York. phrenia. An International Follow-Up Study, John 70. Robertson, A. (1998) Critical reflections on thepolitics Wiley & Sons, Ltd, Chichester. of need: implications for public health. Social Science 56. Warner, R. (1985) Recovery from Schizophrenia: Psy- and Medicine, 47, 1419–1430. chiatry and Political Economy, Routledge & Kegan 71. Parker, R. and Aggleton, P. (2003) HIV and AIDS- Paul, London. related stigma and discrimination: a conceptual frame- 57. Kleinman, A. (1988) Rethinking Psychiatry, Free work and implications for action. Social Science and Press, New York. Medicine, 57, 13–24. 58. Jablensky, A., Sartorius, N. and Ernberg, G. et al. 72. Cohen, C. I. (1993) The biomedicalization of psychia- (1992) Schizophrenia: manifestations, incidence in try: a critical overview. Community Mental Health course in different cultures. AWorld Health Organiza- Journal, 29, 509–521. tion ten-country study. Psychological Medicine. 73. Draine, J., Salzer, M. S., Culhane, D. P. and Hadley, Monograph Supplement, 20, 1–97. T. R. (2002) Role of social disadvantage in crime,

TAKING INEQUALITY’S MEASURE 359 joblessness, and homelessness among persons with ser- 88. Marmot, M. (2004) The Status Syndrome, Times ious mental illness. Psychiatric Services, 53, 565–573. Books, New York. 74. Borg, M., Sells, D., Topor, A. et al. (2005) What makes 89. World Health Organization Commission on Social a house a home: the role of material resources in Determinants of Health (WHO CSDH) (2008) Closing recoveryfrom severe mental illness. American Journal the Gap in a Generation: Health Equity through of Psychiatric Rehabilitation, 8, 243–256. Action on the Social Determinants of Health. Final 75. Wilton, R. (2004) Putting policy into practice? Poverty Report of the Commission on Social Determinants of and people with serious mental illness. Social Science Health, World Health Organization, Geneva. and Medicine, 58, 25–39. 90. Boydell, J., van Os, J. and McKenzie, K. (2001) 76. Warner, R. (2008) Social factors as a basis for treat- Incidence of schizophrenia in ethnic minorities in ment, in Society and Psychosis (eds C. Morgan, K. London: ecological study into interactions with envir- McKenzie and P. Fearon), Cambridge University onment. British Medical Journal, 323, 1336–1338. Press, New York, pp. 163–178. 91. Gephardt, M. A. (1997) Neighborhoods and commu- 77. Mental Health ‘Recovery’ Study Working Group nities as contexts for development, in Neighborhood (2009) Mental Health ‘Recovery’: Users and Refusers, Poverty, Volume I: Context and Consequences for Wellesley Institute, Toronto. Children (eds J. Brooks-Gunn, G. J. Duncan and J. L. Aber), Russell Sage Foundation, New York, pp. 1–43. 78. Ware, N. and Goldfinger, S. (1997) Poverty and reha- bilitation in severe psychiatric disorders. Psychiatric 92. Diez-Roux, A. (1998) Bringing context back into epi- Rehabilitation Journal, 21, 3–9. demiology: variables and fallacies in multilevel analy- sis. American Journal of Public Health, 88, 216–222. 79. Lovell, A. M. (1997) The city is my mother: narratives of schizophrenia and homelessness. American Anthro- 93. Diez-Roux, A. (2001) Investigating neighborhood and pologist, New Series, 9 (2), 355–368. area effects on health. American Journal of Public Health, 91, 1783–1789. 80. McGrath, J. J. (2007) The surprisingly rich contours of schizophrenia epidemiology. Archives of General Psy- 94. March, D., Morgan, C., Bresnahan, M. and Susser, E. chiatry, 64, 14–16. (2008)Conceptualizingthesocialworld,inSocietyand Psychosis(edsC.Morgan,K.McKenzieandP.Fearon), 81. Morgan, C., McKenzie, K. and Fearon, P. (eds) (2008) Cambridge University Press, New York, pp. 41–56. Society and Psychosis, Cambridge University Press, New York. 95. Macera, C. A., Armstead, C. A. and Anderson, N. A. (2001) Sociocultural influences on health, in Hand- 82. Eisenberg, L. (2004) Social psychiatry and the human book of Health Psychology (eds A. Baum, T. A. genome: contextualising heritability. British Journal Revenson and J. E. Singer), Lawrence Erlbaum, Mah- of Psychiatry, 184, 101–103. wah, New Jersey, pp. 427–440. 83. Muntaner, C., Lynch, J. W. and Davey Smith, G. 96. Bourdieu, P. (1977) Outline of a Theory of Practice, (2008) Social capital and the third way in public Cambridge University Press, New York. health, in Critical Perspectives in Public Health (eds J. Green and R. Labonte), Oxford University Press, 97. Kreiger, N. and DaveySmith,G. (2004) ‘Bodies count’ New York, pp. 35–46. and body counts: social epidemiology and embodying 84. Friedli, L. (2009) Mental Health, Resilience and inequality. Epidemiologic Reviews, 26, 92–103. Inequalities, WHO/Europe, Copenhagen. 98. Boydell, J. and McKenzie, K. (2008) Society, place and space, in Society and Psychosis (eds C. Morgan,K. 85. Link, B. G. and Phelan, J. (1995) Social conditions as fundamental causes of disease. Journal of Health and McKenzie and P. Fearon), Cambridge University Press, New York, pp. 77–94. Social Behavior, Extra Issue, 80–94. 99. Lutfey, K. and Freese, J. (2005) Toward some funda- 86. Link, B. G. and Phelan, J. (2005) Fundamental sources of health inequities, in Policy Challenges in Modern mentals of fundamental causality: socioeconomic sta- Health Care (eds D. Mechanic, L. B. Rogut, D. C. tus and health in the routine clinic visit for diabetes. Colby and J. R. Knickmam), Rutgers University Press, American Journal of Sociology, 110, 1326–1372. New Brunswick, New Jersey, pp. 71–84. 100. Elkind, H. B. (1938) Is there an epidemiology of mental disease? American Journal of Public Health, 87. Wilkinson, R. (1996) Unhealthy Societies: The Afflic- 28, 245–250. tions of Inequality, Routledge, London.

360 SOCIAL CONSEQUENCES AND RESPONSES 101. Linton, R. (1956) Culture and Mental Disorders,C.C. 119. Holmes, T. H. and Rahe, R. H. (1967) The social Thomas, Springfield, Massachusetts. readjustment rating scale. Journal of Psychosomatic 102. Hollingshead, A. B. and Redlich, F. (1958) Social Research, 11, 213–218. Class and Mental Health, John Wiley & Sons, Inc., 120. Dohrenwend, B. S. and Dohrenwend, B. P. (eds) New York. (1974) Stressful Life Events: Their Nature and 103. Murphy, H. (1961) Social change and mental Effects, John Wiley & Sons, Inc., New York. health. The Milbank Memorial Fund Quarterly, 39, 121. Lazarus, R. S. and Folkman, S. (1984) Stress, Apprai- 385–445. sal, and Coping, Springer, New York. 104. Leighton, D. C., Harding, J. S., Macklin, D. B. et al. 122. Grant, I., Sweetwood, H. L., Yager, J. and Gerst, M. (1963) The Character of Danger. Vol. III, The Stirling (1981) Quality of life events in relation to psychiatric County Study of Psychiatric Disorder and Sociocul- symptoms. Archives of General Psychiatry, 38, tural Environment, Basic Books, New York. 335–339. 105. Opler, M. K. (1967) Culture and Social Psychiatry, 123. Paykel, E. S. (1974) Life stress and psychiatric Atherton. disorder: applications of the clinical approach, in 106. Hammer, M. and Leacock, E. (1961) Source material Stressful Life Events: Their Nature and Effects on the epidemiology of illness, in Field Studies in the (eds B. S. Dohrenwend and B. P. Dohrenwend), John Mental Disorders (ed. J. Zubin), Grune and Stratton, Wiley & Sons, Inc., New York, pp. 135–149. New York, pp. 418–486. 124. Ross, C. E. and Mirowsky, J. (1979) A comparison of 107. Faris, R. and Dunham, H. (1939) Mental Disorders in life event weighting schemes: change, undesirability, Urban Areas, University of Chicago Press, Chicago, and effect-proportional indices. Journal of Health and Illinois. Social Behavior, 20, 166–177. 108. Lief, A. (ed.) (1948) The Commonsense Psychiatry of 125. Vinokur, A. and Selzer, M. L. (1975) Desirable versus Dr. Adolf Meyer, McGraw-Hill, New York. undesirable life events: their relationship to stress and mental distress. Journal of Personality and Social 109. Selye, H. (1936) A syndrome produced by diverse Psychology, 32, 329–337. nocuous agents. Nature, 138, 32. 126. Zautra, A. J. and Reich, J. W. (1983) Life events and 110. Selye, H. (1956) The Stress of Life, McGraw-Hill, New perceptions of life quality: developments in a two- York. factor approach. Journal of Community Psychology, 111. Dubos, R. (1965) Man Adapting, Yale University 11, 121–132. Press, New Haven, Connecticut. 127. Dohrenwend, B. P. (1979) Stressful life events and psy- 112. Selten, J. P. and Cantor-Graae, E. (2005) Social defeat: chopathology: some issues of theory and method, in risk factor for schizophrenia? British Journal of Psy- StressandMentalDisorder(edsJ.E.Barrett,R.M.Rose chiatry, 187, 101–102. and G. L. Klerman), Raven, New York, pp. 1–15. 113. Luhrmann, T. M. (2008) ‘The street will drive you 128. Pearlin, L. I. and Lieberman, M. A. (1979) Social crazy’: why homeless psychotic women in the institu- sources of emotional stress, in Research in Community tional circuit in the United States often say no to offers and Mental Health (ed. R. Simmons), JAI, Greenwich, of help. American Journal of Psychiatry, 165, 15–20. Connecticut, pp. 217–248. 114. Merton, R. K. (1938) Social structure and anomie. 129. Modell, J. (1989) Into One’s Own: From Youth to American Sociological Review, 3, 672–682. Adulthood in the United States, 1920–1975, Univer- 115. Fox, R. W. (1979) So Far Disordered in Mind, Uni- sity of California Press, Berkeley, California. versity of California Press, Berkeley, California. 130. Hogan, D. P. and Astone, N. M. (1986) The transition 116. Ingelby, D. (1980) Critical Psychiatry, Pantheon, New to adulthood. Annual Review of Sociology, 12, York. 109–130. 117. Spitzer, S. and Scull, A. T. (eds) (1985) Research 131. Brown, G. W. and Harris, T. O. (1978) Social Origins in Law, Deviance and Social Control, JAI Press, of Depression: A Study of Psychiatric Disorder in New York. Women, Free Press, New York. 118. Whitehead, M. and Dahlgren, G. (1991) What can be 132. Kanner, A. D., Coyne, J. C., Schaefer, C. and Lazarus, done about inequalities in health? Lancet, 338, R. S. (1981) Comparison of two modes of stress 1059–1063.

TAKING INEQUALITY’S MEASURE 361 measurement: daily hassles and uplifts versus major published online August 14, DOI: 10.1093/schbul/ life events. Journal of Behavioral Medicine, 4, 1–39. sbn103. 133. Gersten, J. C., Langner, T. S., Eisenberg, J. G. and 146. Turner, R. J. and Wheaton, B. (1995) Checklist mea- Simcha-Fagan, O. (1977) An evaluation of the etio- surement of life events, in Measuring Stress: A guide logical role of stressful life-change events in psycho- for Health and Social Scientists (eds S. Cohen, R. C. logical disorders. Journal of Health and Social Beha- Kessler and L. U. Gordon), Oxford University Press, vior, 18, 228–244. Oxford, pp. 40–85. 134. Rubin, L. (1977) Worlds of Pain, Basic Books, New 147. Gersten, J. C., Langner, T. S., Eisenberg, J. C. and York. Orzek, L. (1974) Child behavior and life events: 135. House, J. S. (1981) Work Stress and Social Support, undesirable change or change per se? in Stressful Life Addison Wesley, Reading, Massachusetts. Events: Their Nature and Effects (eds B. S. Dohren- wend and B. P. Dohrenwend), John Wiley& Sons, Inc., 136. Thoits, P. A. (1999) Sociological approaches to mental illness, in A Handbook for the Study of Mental Health New York, pp. 159–170. (eds. A. V. Horwitz and T. L. Schied), Cambridge 148. Hammen, C. (1991) Generation of stress in the course University Press, New York, pp. 121–138. of unipolar depression. Journal of Abnormal Psychol- ogy, 100, 555–561. 137. Wheaton, B. (1990) Life transitions, role histories, and mental health. American Sociological Review, 55, 149. Dohrenwend, B. P., Raphael, K. G., Schwartz, S. 209–223. et al. (1993) The structured event probe and narrative rating method (SEPRATE) for measuring stressful 138. Parcel, T. L. and Menaghan, E. G. (1990) Maternal working conditions and child verbal facility: studying life events, in Handbook of Stress: Theoretical and the intergenerational transmission of inequality from Clinical Aspects, 2nd edn (eds L. Golberger and mothers to young children. Social Psychology Quar- S. Breznitz), Free Press, New York, pp. 174–199. terly, 53, 132–147. 150. Harris, T. (2001) Recent developments in understand- ing the psychosocial aspects of depression. British 139. Anehensel, C. S., Pearlin, L. I., Mullan, J. T. et al. Medical Bulletin, 57, 17–32. (1995) Profiles in Caregiving: The Unexpected Career, Academic Press, New York. 151. Herman, D. B., Susser, E. S., Struening, E. L., Link, B.G. (1997) Adverse child experiences: Are they risk 140. Monroe, S. M. (2008) Modern approaches to concep- factors for adult homelessness? American Journal of tualizing and measuring human life stress. Annual Public Health, 87, 249–255. Review of Clinical Psychology, 4, 33–52. 152. Craig, T. K. J. and Hodson, S. (1998) Homeless 141. Brown, G. W. (1974) Meaning, measurement, and youth in London: I. Childhood antecedents and psy- stress of life events, in Stressful Life Events: Their chiatric disorder. Psychological Medicine, 28, Nature and Effects (eds B. S. Dohrenwend and B. P. 1379–1388. Dohrenwend), John Wiley & Sons, Inc., New York, pp. 217–243. 153. Bruce, M. L. (1990) Socioeconomic status and psy- chiatric disorders. Current Opinion in Psychiatry, 3, 142. Brown, G. W. (1989) Life event and measurement, in 696–699. Life Events and Illness (eds G. W. Brown and T. O. Harris), Guildford Press, London and New York, 154. Dohrenwend, B. P. (1990) Socioeconomic status pp. 3–45. (SES) and psychiatric disorders. Social Psychiatry and Psychiatric Epidemiology, 25, 41–47. 143. Pearlin, L. I. (1999) Stress and mental health: a con- ceptual overview, in A Handbook for the Study of 155. Grant, K. E., Compas, B. E., Thurm, A. E. et al. (2004) Mental Health (eds A. V. Horwitz and T. L. Scheid), Stressors and child and adolescent psychopathology: Cambridge University Press, New York, pp. 161–175. measurement issues and prospective effects. Journal 144. Langner, T. S. and Michael, S. T. (1963) The Midtown of Clinical Child and Adolescent Psychology, 33, Manhattan Study. vol. 2, Life Stress and Mental 412–425. Health, Free Press, New York. 156. Kessler, R. C.,McGonagle, K. A., Zhao, S. et al.(1994) Lifetime and 12-month prevalence of DSM-III-R 145. Bentall, R. P. and Ferneyhough, C. (2008) Social psychiatric disorders in the United States: results from predictors of psychotic experiences: specificity and the National Comorbidity Survey. Archives of General psychological mechanisms. Schizophrenia Bulletin, Psychiatry, 51, 8–19.

362 SOCIAL CONSEQUENCES AND RESPONSES 157. Kessler, R. C., Chiu, W. T., Demier, O. and Walters, 171. Eaton, W. W. (1985) The epidemiology of schizo- E. E. (2005) Prevalence, severity, and comorbidity phrenia. Epidemiological Reviews, 7, 105–126. of 12-month DSM-IV disorders in the National 172. Eaton, W. W. and Muntaner, C. (1999) Social strati- Comorbidity Survey Replication. Archives of General fication and mental disorder, in A Handbook for the Psychiatry, 62, 617–627. Study of Mental Health (eds A. V. Horwitz and T. L. 158. Dohrenwend, B. P. and Schwartz, S. (1995) Socio- Scheid), Cambridge University Press, New York, economic status and psychiatric disorders. Current pp. 259–283. Opinion in Psychiatry, 82, 138–141. 173. Dohrenwend, B. P. and Dohrenwend, B. S. (1969) 159. Thoits, P. A. (1981) Undesirable life events and psy- Social Status and Psychological Disorder: A Causal chophysiological distress: a problem of operational Inquiry, John Wiley & Sons, Inc., New York. confounding. American Sociological Review, 46, 174. Dohrenwend, B. P., Levav, I., Shrout, P. E. et al. (1992) 97–109. Socioeconomic status and psychiatric disorders: the 160. Hammen, C. (2005) Stress and depression. Annual causation-selection issue. Science, 255, 946–952. Review of Clinical Psychology, 1, 293–319. 175. Waters, M. (2000) Black Identities: West Indian 161. Monroe, S. M. and Harness, K. L. (2005) Life stress, Immigrant Dreams and American Realities, Harvard the ‘kindling’ hypothesis, and the recurrence of University Press, Cambridge, Massachusetts. depression: considerations from a life stress perspec- 176. Williams, D. R., Hector, M., Gonz alez, H. M. et al. tive. Psychological Review, 112, 417–445. (2007) Prevalence and distribution of major depressive 162. Olson, K. (2001) Distributive justice and the politics of disorder in African Americans, Caribbean blacks, and difference. Critical Horizons, 2, 5–32. non-Hispanic whites: results from the National Survey 163. Lareau, A. (2003) Unequal Childhoods, University of of American Life. Archives of General Psychiatry, 64 California Press, Berkeley, California. (3), 305–315. 164. Dressler, W. W., Dos Santos, J. E. and Balieiro, M. C. 177. Fraser, N. (2000) Rethinking recognition. New Left (1996) Studying diversity and sharing in culture: an Review, 3 (May–June), 107–120. example of lifestyle in Brazil. Journal of Anthropo- 178. Karlsen, S., Nazroo, J. Y., McKenzie, K., Bhui, K. and logical Research, 52, 331–353. Weich, S. (2005) Racism, psychosis and common 165. Dressler, W. W. and Bindon, J. R. (2000) The health mental disorder among ethnic minority groups in consequences of cultural consonance: cultural dimen- England. Psychological Medicine, 35, 1795–1803. sions of lifestyle, social support, and arterial blood 179. Tanur, J. (ed.) (1992) Questions about Questions: pressure in an African American community. Amer- Inquiries into the Cognitive Bases of Surveys, Russell ican Anthropologist, 102, 244–260. Sage, New York. 166. Erikson, K. (1976) Everything in Its Path, Simon & 180. Schwarz, N. (2007) Cognitive aspects of survey meth- Schuster, New York. odology. Applied Cognitive Psychology, 21, 277–287. 167. Appadurai, A. (2004) The capacity to aspire: culture 181. Dohrenwend, B. P., Link, B.G., Kern, R. et al. (1990) and the terms of recognition, in Culture and Public Measuring life events: the problem of variability Action (eds V. Rao and M. Walton), Stanford within event categories. Stress Medicine, 6, 179–187. University, Stanford, California. 182. Briggs, C. (1986) Learning How to Ask, Cambridge 168. Lutfey, K. and Freese, J. (2005) Toward some funda- University Press, New York. mentals of fundamental causality: socioeconomic sta- 183. Monroe, S. M. and McQuaid, J. R. (1994) Measuring tus and health in the routine clinic visit for diabetes. life stress and assessing its impact on mental health, in American Journal of Sociology, 110, 1326–1372. Stress and Mental Health: Contemporary Issues and 169. Dohrenwend,B.S.,Dohrenwend,B.P.,Dodson,M.and Prospects for the Future (eds W. R. Avison and I. H. Shrout,P.E.(1984)Symptoms,hassles,socialsupports Gotlib), Plenum, New York, pp. 43–73. and life events: the problem of confounded measures. 184. Southwick, S. M., Morgan III, C. A., Nicolau, A. L. Journal of Abnormal Psychology, 93, 222–230. and Charney, D. S. (1997) Consistency of memory for 170. Eaton, W. W. and Lasry, J.-C. (1978) Mental health and combat-related traumatic events in veterans of Opera- occupational mobility in a group of immigrants. Social tion Desert Storm. American Journal of Psychiatry, Science and Medicine, 12, 53–58. 154, 173–177.

TAKING INEQUALITY’S MEASURE 363 185. Kihlstrom, J. F., Eich, E., Sandbrand, D. and Tobias, B. outcome in recurrent depression. Journal of Abnormal A. (1999) Emotion and memory: implications for self- Psychology, 109, 787–791. report, in The Science of Self-Report: Implications for 197. Brown, G. W., Bhrolchain, M. N. and Harris, T. (1975) Research and Practice (eds A. A. Stone, C. A. Social class and psychiatric disturbance among Bachrach, J. B. Jobe et al.), Lawrence Erlbaum, women in an urban population. Sociology, 2, 225–254. Mahwah, New Jersey. 198. Kessler, R. C. and McLeod, J. D. (1984) Sex differ- 186. Hatch, L. and Dohrenwend, B. P. (2007) Distribution ences invulnerability to undesirable life events. Amer- of traumatic and other stressful life events by race/ ican Sociological Review, 49, 620–631. ethnicity, gender, SES and age: a review of the 199. Brown, G. W., Harris. T. O. and Hepworth, C. (1995) research. American Journal of Community Psychol- Loss, humiliation and entrapment among women ogy, 40, 313–332. developing depression: a patient and non-patient com- 187. Bifulco, A., Brown, G. W., Lillie, A. and Jarvis, J. parison. Psychological Medicine, 25, 7–21. (1997) Memories of childhood and neglect: corrobora- 200. Stueve, A., Dohrenwend, B. P. and Skodol, A. E. tion in a series of sisters. Journal of Child Psychology (1998) Relationships between stressful life events and and Psychiatry, 38, 365–374. episodes of major depression and nonaffective psy- 188. Dohrenwend, B. P. (2006) Inventorying stressful life chotic disorders. Selected results from a New York risk events as risk factors for psychopathology: toward factor study, in Adversity, Stress, and Psychopathol- resolution of the problem of intracategory variability. ogy (ed. B. P. Dohrenwend), Oxford University Press, Psychological Bulletin, 132, 477–495. New York, pp. 341–357. 189. Gorman, D. M. (1993) A review of studies comparing 201. Breslau, N., Kessler, R. C., Chilcoat, H. D. et al. (1998) checklist with interview methods of data collection in Trauma and posttraumatic stress disorder in the com- life event research. Behavioral Medicine, 19, 66–73. munity. Archives of General Psychiatry, 55, 626–632. 190. Thoits, P. A. (1995) Stress, coping, and social support 202. Breslau, N., Davis, G. C., Adreski, P. and Peterson, E. processes: Where are we? What next? Journal of (1991) Traumatic events and posttraumatic stress dis- Health and Social Behavior, 36, Extra Issue, 53–79. order in an urban population of young adults. Archives 191. Paykel, E. S. (2001) The evolution of life events of General Psychiatry, 48, 216–222. research in psychiatry. Journal of Affective Disorders, 203. Hopper, K. (2003) Reckoning with Homelessness, 62, 141–149. Cornell University Press, Ithaca, New York. 192. Brown, G. W. and Harris, T. O. (1978) Social Origins 204. Garraty, J. (1978) Unemployment in History, Harper, of Depression: A Study of Psychiatric Disorder in New York. Women, Free Press, New York. 205. Kumar, K. (1988) From work to employment to 193. Wethington, E., Brown, G. W. and Kessler, R. C. unemployment, in On Work: Historical, Comparative (1995) Interview measurement of stressful life events, and Theoretical Approaches (ed. R. E. Pahl), Basil in Measuring Stress: A Guide for Health and Social Blackwell, New York, pp. 138–164. Scientists (eds S. Cohen, R. C. Kessler and L. U. 206. Folbre, N. (2001) The Invisible Heart: Economics and Gordon), Oxford University Press, Oxford, pp. 59–79. Family Values, The New Press. 194. Harnish, O. D., Aseltine, R. H. and Gore, S. (2000) 207. Eisenberg, P. and Lazarsfeld, P. F. (1938) The psycho- Resolution of stressful experiences as an indicator of logical effects of unemployment. Psychological Bul- coping effectiveness in young adults: an event history letin, 35, 258–390. analysis. Journal of Health and Social Behavior, 41, 208. Elder, G. H. (1974) Children of the Great Depression, 121–136. University of Chicago Press, Chicago, Illinois. 195. Duggal, S., Malkoff-Schwartz, S., Birmaher, B. et al. 209. Perucci, R. and Perucci, C. (1990) Unemployment and (2000) Assessment of life stress in adolescents: self-report versus interview methods. Journal of the mental health: research and policy implications, in American Academy of Child and Adolescent Psychia- Research in Community and Mental Health: Mental try, 39, 445–451. Disorder in Social Context (ed. J. R. Greenley), JAI, Greenwich, Connecticut, pp. 237–264. 196. McQuaid, J. R., Monroe, S. M., Roberts, J. F. et al. 210. Bhugra, D. (1993) Unemployment, poverty and (2000) A comparison of two life stress assessment homelessness, in Principles of Social Psychiatry approaches: prospective prediction of treatment

364 SOCIAL CONSEQUENCES AND RESPONSES (eds D. Bhugra and J. Leff), Blackwell, Oxford, 226. Pahl, R. E. (ed.) (1988) On Work: Historical, Com- pp. 355–382. parative and Theoretical Approaches, Blackwell, 211. Fryer, D. (1995) Benefit Agency? Labour Market London. Disadvantage, Deprivation and Mental Health. The 227. Offe, C. and Heinze, R. G. (1992) Beyond Employ- C.S. Myers Lecture 1994. The Psychologist, June, ment: Time, Work and the Informal Economy, Temple 265–272. University Press, Philadelphia, Pennsylvania. 212. Fryer, D. and Ullah, P. (1987) Unemployed People: 228. Tilly, C. (1996) Half a Job, Temple University Press, Social and Psychological Perspectives, Open Univer- Philadelphia, Pennsylvania. sity Press, Berkshire. 229. Robinson, J. (1936) Disguised unemployment. 213. Engberson, G., Schuyt, K., Timmer, J. and van Waar- Economic Journal, 46, 225–237. den, F. (1993) Cultures of Unemployment, Westview 230. Kaufman, H. G. (1982) Professionals in Search of Press, Oxford. Work, John Wiley & Sons, Inc., New York. 214. Kronauer, M. and Vogel, B. (1993) Experiences with 231. Burris, B. H. (1983) The human effects of under- unemployment today: between temporary opportunity employment. Social Problems, 31, 96–110. and social exclusion. International Journal of Politi- 232. Dooley, D. and Catalano, R. (1980) Economic change cal Economy, 23, 70–94. as a cause of behavioral disorder. Psychological Bul- 215. Wilson, W. J. (1996) When Work Disappears: The letin, 87, 450–468. World of the New Urban Poor, Knopf, New York. 233. Jahoda, M. (1982) Employment and Unemployment: 216. Lewis, A. (1935) Neurosis and unemployment. A Social Psychological Analysis, Cambridge Univer- Lancet, 226, 293–297. sity Press, New York. 217. Bakke, E. W. (1940) Citizens Without Work, Yale 234. Warr, P. (1987) Work Unemployment and Mental University Press, New Haven, Connecticut. Health, Clarendon, Oxford. 218. Bakke, E. W. (1940) Unemployed Worker: A Study of 235. Whitehead, M. (1990) The health divide, in Inequal- the Task of Making a Living without a Job, Yale ities in Health (eds P. Townsend and N. Davidson), University Press, New Haven, Connecticut. Penguin, London, pp. 222–356. 219. Bremer, W. W. (1984) Depression Winters: New York 236. Dooley, D., Fielding, J. and Levi, L. (1996) Health and Social Workers and the New Deal, Temple University, unemployment. Annual Review of Public Health, 17, Philadelphia, Pennsylvania. 449–465. 220. McElvaine, R. S. (1984) The Great Depression, 237. Fryer, D. (1997) International perspectives on youth Crown, New York. unemployment and mental health: some central issues. 221. Cockerham, W. (1999) Health and Social Change in Journal of Adolescence, 20, 333–342. Russia and Eastern Europe, Routledge, New York. 238. Dooley, D., Prause, J. and Ham-Rowbottom, K. A. 222. Stuckler, D., King, L. and McKee, M. (2009) Mass (2000) Underemployment and depression: longitudi- privatization and the post-communist mortality crisis: nal relationships. Journal of Health and Social Beha- a cross-national analysis. The Lancet, 373, 399–407. vior, 41, 421–436. 223. Rosenfield, S. (1989) The effects of women’s employ- 239. Taris, T. W. (2002) Unemployment and mental health: ment: personal control and sex differences in mental a longitudinal perspective. International Journal of health. Journal of Health and Social Behavior, 30, Stress Management, 9, 43–57. 77–91. 240. Kelvin, P. and Jarrett, J. E. (1985) Unemployment: Its 224. Lennon, M. C. (1994) Women, work and well-being: Social Psychological Effects, Cambridge University the importance of work conditions. Journal of Health Press, Cambridge. and Social Behavior, 35, 235–247. 241. Leeflang, R. L. I., Klein-Hesselink, D. J. and Spruit, I. 225. Lennon, M. C. (1998) Domestic arrangements and P. (1992) Health effects of unemployment – I. Long- depression: an examination of household labor, in term unemployed men in a rural and an urban setting. Adversity, Stress, and Psychopathology (ed. B. P. Social Science and Medicine, 34, 341–350. Dohrenwend), Oxford University Press, New York, 242. Braithwaite, A. and Garcia, S. (1985) Depression in pp. 409–421. the young unemployed and those on Youth Opportu-

TAKING INEQUALITY’S MEASURE 365 nity Schemes. British Journal of Medical Psychology, 256. Liebow, E. (1967) Tally’s Corner, Little, Brown, 58, 67–74. Boston, Massachusetts. 243. McKee, L. and Bell, C. (1986) His unemployment, her 257. Rubin, L. B. (1976) Worlds of Pain: Life in the Work- problem: the domestic and marital consequences of ing Class Family, Basic, New York. male unemployment, in The Experience of Unemploy- 258. Mishel, L., Bernstein, J. and Schmitt, J. (1999) The ment (eds S. Allen, S. Waton, K. Purcell and S. Wood), State of Working America, 1998–1999, ILR Press, Macmillan, Basingstroke. Ithaca, New York. 244. Jones, F. and Fletcher, B. C. (1993) An empirical 259. Burgard, S. A., Brand, J. E. and House, J. S. (2009) study of occupational stress transmission in working Perceived job insecurity and worker health in the couples. Human Relations, 46, 881–903. United States. Social Science and Medicine, 69, 245. Liem, R. and Liem, J. H. (1988) Psychological effects 777–785. of unemployment on workers and their families. 260. Farber, H. S. (2005) What dowe know about job loss in Journal of Social Issues, 44, 87–105. the United States? Federal Reserve Bank of Chicago. 246. McLoyd, V. C. (1989) Socialization and development Economic Perspectives, 2Q, 13–27, www.chicagofed. in a changing economy: the effects of paternal job and org/publications/economicperspectives/ income loss on children. American Psychologist, 44, ep_2qtr2005_part2_farber.pdf. 293–302. 261. Cobb, S. and Kasl, S. V. (1977) Termination: The 247. Oakley, C. A. (1936) Some psychological problems of Consequences of Job Loss, U.S. DHEW, Cincinatti, a depressed area. Human Factor, 10, 393–404. Ohio. 248. Komarovsky, M. (1940) The Unemployed Man and 262. Hartley, J. F., Jacobson, D., Klandermans, B. and van His Family, Dryden, New York. Vuuren, T. (1991) Job Insecurity: Coping with Jobs at 249. Atkinson, T., Liem, R. and Liem, L. H. (1986) The Risk, Sage, London. social costs of unemployment: implications for social 263. Eisenberg, P. and Lazarsfeld, P. F. (1938) The psycho- support. Journal of Health and Social Behavior, 27, logical effects of unemployment. Psychological Bul- 317–331. letin, 35, 358–390. 250. Whelan, C. T. (1992) The role of income, life-style 264. Dooley, D. (2003) Unemployment, underemployment, deprivation and financial straits in mediating the and mental health: conceptualizing employment status impact of unemployment on psychological distress: as a continuum. American Journal of Community evidence from the Republic of Ireland. Journal of Psychology, 32, 9–20. Occupational and Organizational Psychology, 65, 265. Wilson, W. J. (1987) The Truly Disadvantaged, Uni- 331–344. versity of Chicago Press, New York. 251. Newman, K. S. (1988) Falling from Grace, Free Press, 266. Ehrenreich, B. (2001) Nickel and Dimed, Henry Holt, New York. New York. 252. Hamilton, V., Broman, C. L., Hoffman, W. and Renner, 267. Shipler, D. K. (2004) The Working Poor, Knopf, New D. (1990) Hard times and vulnerable people: initial York. effects of plant closings on autoworkers’ mental 268. Granovetter, M. S. (1973) The strength of weak ties. health. Journal of Health and Social Behavior, 31, American Journal of Sociology, 78, 1360–1380. 123–140. 269. Granovetter, M. S. (1983) The strength of weak ties: a 253. Viinam€ aki, H., Koskela, K., Niskanen, L. et al. (1993) network theory revisited. Sociological Theory, 1, Unemployment and mental well-being: a factory clo- 201–233. sure study in Finland. Acta Psychiatrica Scandinavia, 270. McLeod, J. (1995) Ain’t No Makin’ It: Aspirations 88, 429–433. and Attainment in a Low-Income Neighborhood, 254. Fenwick, R. and Tausig, M. (1994) The macroeco- Expanded edn, Westview Press, Boulder, Colorado. nomic context of job stress. Journal of Health and Social Behavior, 35, 266–282. 271. Royster, D.A. (2003) Race and the Invisible Hand: How White Networks Exclude Black Men from Blue- 255. Strully, K. W. (2009) Job loss and health in the U.S. Collar Jobs, University of California Press, Berkeley, labor market. Demography, 46, 221–246. California.

366 SOCIAL CONSEQUENCES AND RESPONSES 272. Newman, K. S. (1999) No Shame in My Game, Knopf, York City. American Journal of Public Health, 88, New York. 1651–1657. 273. Smith, S. (2007) Lone Pursuit: Distrust and Defensive 288. Culhane, D. P., Dejowski, E. F., Iban ˜ez, J. et al. (1994) Individualism among the Black Poor, Russell Sage, Public shelter admission rates in Philadelphia and New New York. York City: the implications of turnover for sheltered 274. Jahoda, M. and Rush, H. (1980) Work, Employment population counts. Housing Policy Debate, 5, and Unemployment. Occasional Papers Series, 107–140. Number 12, University of Sussex: Science Policy 289. Kuhn, R. and Culhane, D. P. (1998) Applying cluster Research Unit. analysis to test a typology of homelessness by pattern 275. Room, R. (1984) Alcohol and ethnography: a case of of shelter utilization. American Journal of Community problem deflation? Current Anthropology, 25, Psychology, 29, 207–232. 169–178. 290. Burt, M. R., Aron, L. Y. and Lee, E. (2001) 276. Dooley, D., Catalano, R. and Hough, R. (1992) Unem- Helping America’s Homeless: Emergency Shelter or ployment and alcohol disorders in 1910 and 1990: drift Affordable Housing? Urban Institute, Washington, DC. versus social causation. Journal of Occupational and 291. Kertesz, S. G., Larson, M. J., Horton, N. et al. (2005) Organizational Psychology, 65, 277–290. Homeless chronicity and health-related quality of life 277. Scheper-Hughes, N. (1979) Saints, Scholars and Schi- trajectories among adults with addictions. Medical zophrenics, University of California Press, Berkeley, Care, 43, 574–585. California. 292. Caton, C., Wilkins, C. and Anderson, J. (2007) People 278. Berger, J. (1992) Into Their Labours, Pantheon, New who experience long-term homelessness: characteris- York. tics and interventions, in Toward Understanding Homelessness, The 2007 National Symposium on 279. Bourdieu, P. (2008) The Bachelors’ Ball: The Crisis of Peasant Society in Bearn, University of Chicago Homelessness Research. Washington, DC: DHHS, DHUD (eds D. Dennis, G. Locke and J. Khadduri), Press, Chicago, Illinois. pp. 4-1–4-44. 280. Scheper-Hughes,N.(2001)Saints,ScholarsandSchizo- 293. Barrow, S., Soto, G. and Cordova, P. (2004) Final phrenics (Anniversary edition, updated and expanded), Report on the Evaluation of the Closer to Home University of California Press, Berkley, California. Initiative. Corporation for Supportive Housing, New 281. Fullilove, M. (2005) Root Shock, Ballantine, York. New York. 294. Barrow, S. M., Herman, D. B., Cordova, P. and Struen- 282. Ingleby, D. (ed.) (2004) Forced Migration and Mental ing, E. L. (1999) Mortality among homeless shelter Health, Springer, New York. residents in New York City. American Journal of 283. Schweitzer, R. and Steel, A. (2008) Researching refu- Public Health, 89, 529–534. gees: methodological and ethical considerations, in 295. Culhane, D. P., Gollub, E., Kuhn, R. and Shpaner, M. Doing Cross-Cultural Research (ed. P. Liamputtong), (2001) The co-occurrence of AIDS and home- Springer, New York, pp. 87–101. lessness: results from the integration of administrative 284. Link, B. G., Susser, E., Stueve, A. et al. (1994) Life- databases for AIDS surveillance and public shelter time and five-year prevalence of homelessness in the utilisation in Philadelphia. Journal of Epidemiology United States. American Journal of Public Health, 84, and Community Health, 55, 515–520. 1907–1912. 296. Robertson, M. J., Clark, R. A., Charlebois, E. D. et al. 285. Keyssar, A. (1986) Out of Work: The First Century of (2004) HIV seroprevalence among homeless and mar- Unemployment in Massachusetts, Cambridge Univer- ginally housed adults in San Francisco. American sity Press, New York. Journal of Public Health, 94, 1207–1217. 286. Hopper, K., Jost, J., Hay, T. et al. (1997) Homeless- 297. Kerker, B., Bainbridge, J., Li, W. et al. (2005) The ness, severe mental illness and the institutional circuit. Health of Homeless Adults in New York City: A Psychiatric Services, 48, 659–665. Report from the New York City Departments of Health 287. Shinn, M., Weitzman, B. C., Stojanovic, D. et al. and Mental Hygiene and Homeless Services. New (1998) Predictors of homelessness from shelter York City Departments of Health and Mental Hygiene request to housing stability among families in New and Homeless Services, New York.

TAKING INEQUALITY’S MEASURE 367 298. Koegel, P., Burnam, M. A. and Baumohl, J. (1996) The among homeless adolescents. Journal of the Amer- causes of homelessness, in Homelessness in America ican Academy of Child and Adolescent Psychiatry, (ed. J. Baumohl), Oryx, Phoenix, Arizona, pp. 24–33, 43, 325–331. 225–226. 312. Tyler, K., Whitbeck, L., Hoyt, D. and Cauce, A. (2004) 299. Sullivan, G., Burnam, A. and Koegel, P. (2000) Path- Risk factors for sexual victimization among male and ways to homelessness among the mentally ill. Social female homeless and runaway youth. Journal of Inter- Psychiatry and Psychiatric Epidemiology, 35, personal Violence, 19 (5), 503–520. 444–450. 313. Rog, D. J. and Buckner, J. C. (2007) Homeless families 300. Groth, P. (1994) Living Downtown, University of and children. Paper prepared for the National Sympo- California Press, Berkeley, California. sium on Homelessness Research, 1–2 March 2007, 301. Liebow, E. (1993) Tell Them Who I Am, Free Press, Washington, D.C. New York. 314. Masten, A. S., Miliotis, D., Graham-Bermann, S. A. 302. Duneier, M. (1999) Sidewalk, Farrar, Straus, Giroux, et al. (1993) Children in homeless families: risks to New York. mental health and development. Journal of Consulting and Clinical Psychology, 61, 335–343. 303. Baumohl, J. (ed.) (1996) Homelessness in America, Oryx, Phoenix, Arizona. 315. Schteingart, J. S., Molnar, J., Klein, T. P. et al. (1995) Homelessness and child functioning in the context of 304. Burt, M. R., Pearson, C. L. and Montgomery, A. E. (2006) Homelessness: Prevention, Strategies risk and protective factors moderating child outcomes. and Effectiveness, Nova Science, Hauppauge, Journal of Clinical Child Psychology, 24, 320–331. New York. 316. Buckner, J. C. and Bassuk, E. L. (1997) Mental dis- orders and service utilization among youths from 305. Shinn, M. and Weitzman, B. C. (eds) (1990) Urban homelessness. Journal of Social Issues, 46 (4), Special homeless and low-income housed families. Journal Issue. of the American Academy of Child and Adolescent Psychiatry, 36, 890–900. 306. Di Leonardo, M. (1987) The femaleworld of cards and 317. Bassuk, E. L., Weinreb, L. F., Dawson, R. et al. (1997) holidays: women, families, and the work of kinship. Determinants of behavior in homeless and low- Signs, 12, 440–453. income housed preschool children. Pediatrics, 100, 307. Barrow, S. M. and Lewinski, T. (2009) Contexts of 92–100. mother–child separations in homeless families. 318. Buckner, J. C., Bassuk, E. L., Weinreb, L. and Brooks, Analyses of Social Issues and Public Policy. Published M. (1999) Homelessness and its relation to the mental online: 8 April 2009, 8:52AM. DOI: 10.1111/j.1530- health and behavior of low-income school aged chil- 2415.2009.01171.x. dren. Developmental Psychology, 35, 246–257. 308. Lehman, A. F. and Cordray, D. S. (1993) Prevalence of 319. Buckner, J. C., Bassuk, E. L. and Weinreb, L. F. (2001) alcohol, drug, and mental disorders among the home- Predictors of academic achievement among homeless less: one more time. Contemporary Drug Problems, and low-income housed children. Journal of School 20, 355–381. Psychology, 39, 45–69. 309. Toro, P. A., Dworsky, A. and Fowler, P. J. (2007) 320. Huntington, N., Buckner, J. C. and Bassuk, E. C. Homeless Youth in the United States: Recent Research (2008) Adaptation in homeless children: an empirical Findings and Intervention Approaches, Paper prepared examination using cluster analysis. American Beha- for the National Symposium on Homelessness vioral Scientist, 51, 737–755. Research, 1–2 March 2007, Washington, DC. 321. Shinn, M., Knickman, J. R. and Weitzman, B. C. 310. Robertson, M. J. and Toro, P. A. (1999) Homeless youth: research, intervention and policy, in Practical (1991) Social relationships and vulnerability to Lessons: The 1998 National Symposium on Home- becoming homeless among poor families. American lessness Research, Washington, DC: DHUD and Psychologist, 46, 1180–1187. DHSS (eds L. B. Fosburg and D. L. Dennis), 322. Toohey, S. M., Shinn, M. and Weitzman, B. C. (2004) pp. 3-1–3-32. Social networks and homelessness among women heads of household. American Journal of Community 311. Stewart, A. J., Steiman, M., Cauce, A. M. et al. (2004) Psychology, 33, 7–20. Victimization and posttraumatic stress disorder

368 SOCIAL CONSEQUENCES AND RESPONSES 323. Shinn, M., Schteingart, J. S., Chioke-Williams, N. P. 336. Moore, S., Shiell, A., Hawe, P. and Haines, V. A. et al. (2008) Long-term associations of homelessness (2005) The privileging of communitarian ideas: cita- with children’s well-being. American Behavioral tion practices and the translation of social capital into Scientist, 51, 789–809. public health research. American Journal of Public 324. Buckner, J. C., Beardslee, W. R. and Bassuk, E. L. Health, 95, 1330–1337. (2004) Exposure to violence and low income chil- 337. Davey Smith, G., Hart, C., Blane, D. et al. (1997) dren’s mental health: direct, moderated, and mediated Lifetime socio-economic position and mortality. relations. American Journal of Orthopsychiatry, 74, Prospective observational study. British Medical Jour- 413–423. nal, 314, 547–552. 325. Susser, E. S., Line, S. P. and Conover, S. A. (1991) Risk 338. Fisher, H. and Craig, T. (2008) Childhood adversity factors for homelessness among patients admitted to a and psychosis, in Society and Psychosis (eds C. state mental hospital. American Journal of Psychiatry, Morgan, K. McKenzie and P. Fearon), Cambridge 148, 1659–1664. University Press, New York, pp. 95–111. 326. Hopper, K. and Baumohl, J. (1994) Held in abeyance: 339. Myin-Germeys, I. and van Os, J. (2008) Adult adver- rethinking homelessness and advocacy. American sity: Do early environment and genotype create lasting Behavioral Scientist, 37, 522–552. vulnerabilities for adult social adversity in psychosis? 327. Rutter, M. (2005) How the environment affects in Society and Psychosis (eds C. Morgan, K. McKen- mental health. British Journal of Psychiatry, 186, zie, and P. Fearon), Cambridge University Press, New 4–6. York, pp. 127–142. 328. Koegel, P. (1992) Through a different lens: an anthro- 340. Sampson, R. J., Morenoff, J. D. and Gannon-Rowley, pological perspective on the homeless mentally ill. T. (2002) Assessing ‘neighborhood effects’: social Culture, Medicine, and Psychiatry, 16, 1–22. processes and new directions in research. Annual Review of Sociology, 28, 443–478. 329. Sosin, M., Piliavin, I. and Westerfelt, H. (1990) Toward a longitudinal analysis of homelessness. 341. Kwate, N. O. A. (2008) Fried chicken and fresh apples: Journal of Social Issues, 46, 157–174. racial segregation as a fundamental cause of fast food density in black neighborhoods. Health and Place, 14, 330. Susser, E., Valencia, E., Conover, S. et al. (1997) 32–44. Preventing recurrent homelessness among mentally ill men: a ‘critical time’ intervention after discharge 342. Cantor-Graae, E. and Selten, J. P. (2005) Schizophre- from a shelter. American Journal of Public Health, nia and migration: a meta-analysis and review. 87, 256–262. American Journal of Psychiatry, 162, 12–24. 331. Tsemberis, S., Gulcur, L. and Nakae, M. (2004) Hous- 343. Rogers, A. and Pilgrim, D. (2003) Mental Health and ing first, consumer choice, and harm reduction for Inequality, Palgrave Macmillan, New York. homeless individuals. American Journal of Public 344. Friedl, L. (2009) Mental Health, Resilience and Health, 94, 651–656. Inequalities, World Health Organization Europe, 332. Gallobardes, B., Shaw, M., Lawlor, D. et al. (2006) WHO Regional Office for Europe, Copenhagen. Indicators of social position (Part 1). Journal of 345. Silver, H. (2006) Social exclusion, in Blackwell Epidemiology and Community Health, 60, 7–12. Encyclopedia of Sociology (ed. G. Ritzer), Blackwell, 333. Taylor, S. E. and Repetti, R. L. (1997) Health London, pp. 4411–4413. psychology: What is an unhealthy environment and 346. Sayce, L. (2000) From Psychiatric Patient to Citizen: how does it get under the skin? Annual Review of Overcoming Discrimination and Social Exclusion, Psychology, 48, 411–447. Palgrave Macmillan, London. 334. Lynch, J. W. and Kaplan, G. A. (2000) Socioeconomic 347. Radoki, C. (1999) A capital assets framework for position, in Social Epidemiology (eds L. F. Berkman analysing household livelihood strategies: implica- and I. Kawachi), Oxford University Press, New York, tions for policy. Development Policy Review, 17, pp. 13–35. 315–342. 335. Grusky, D. B. and Ravi Kanbur, R. (eds) (2006) 348. Rao, V. and Walton, M. (eds) (2004) Culture and Poverty and Inequality, Stanford University Press, Public Action, Stanford University Press, Stanford, Palo Alto, California. California.

TAKING INEQUALITY’S MEASURE 369 349. Fraser, N. and Olson, K. (2008) Adding Insult to 352. Comin, F., Qizilbash, M. and Alkire, S. (eds) (2008) Injury: Debating Redistribution, Recognition and The Capability Approach: Concepts, Measures Representation, Verso, London. and Applications, Cambridge University Press, 350. White,S.andEllison,M.(2007)Wellbeing,livelihoods New York. and resources in social practice, in Wellbeing in Deve- 353. Hopper, K. (2007) Rethinking social recovery in loping Countries (eds I. Gough and J. A. McGregor), schizophrenia: what a capabilities approach might Cambridge University Press, New York, pp. 157–175. offer. Social Science and Medicine, 65, 868–879. 351. Alkire, S. (2002) Valuing Freedoms: Sen’s Capabil- 354. Ware, N., Hopper, K., Tugenberg, T. et al. (2007) ities Approach and Poverty Reduction, Oxford Connectedness and citizenship: redefining social University Press, Oxford. integration. Psychiatric Services, 58, 469–474.



28 Health economics and psychiatry: the pursuit of efficiency 1 Martin Knapp and Jennifer Beecham 2 1 Personal Social Services Research Unit, London School of Economics, London and Centre for the Economics of Mental Health, King's College London, London, UK 2 Personal Social Services Research Unit, University of Kent, Canterbury and Personal Social Services Research Unit, London School of Economics, London, UK 28.1 BETTER SERVICES, LOWER COSTS Developments in social psychiatry as practised health and well-being of more individuals from the obviously do not stem solely from changes in the same pot of resources. skills or preferences of psychiatrists. Indeed, in their gloomier moments, psychiatrists may believe that ‘Efficiency’ is a term with unfortunate connotations. most of their actions and reactions are influenced Although it appears to some to be self-evidently more by the exigencies of social, economic, demo- a good thing, like ‘sincerity’ or ‘honesty’, just as graphic, ideological or political imperatives than by with the pursuit of these latter virtues, there can be the medical and social needs of the people they treat or unattractive consequences. ... People do not like [efficiency] carried to the extreme where it dominates support or by accepted best practices. Many factors all other considerations, and especially ... where the and forces combine to influence the ways that mental main purposes are the humane and just treatment of health systems are organized, the services and treat- people, and often of people who are having a hard ments they offer and the people who actually receive time of it anyway ([1], p. 1). those supports. One of those forces is the pressure to improve value for money, efficiency or cost effective- What, then, is meant by efficiency? How can it be ness. The consequences of this emphasis on efficiency studied? These questions are addressed in this chapter. are not all welcome, but it would be foolish to reject We first briefly consider the broader relevance of an this emphasis, partly because it is never likely to go economics perspective, and then outline a theoretical away and partly because – after all – a more efficient framework that will facilitate understanding of mental health system is one that will improve the the concept of efficiency. We next describe the most Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

372 SOCIAL CONSEQUENCES AND RESPONSES commonly used approaches to the analysis of effi- providesomeillustrations.Thechapterconcludeswith ciency (particularly cost-effectiveness analysis) and somecomments onwhatneeds tobe doneinthefuture. 28.2 THE CONTRIBUTION OF ECONOMICS Readers of this book will most often come into . What are the workforce requirements of a particular contact with economics via media reports of the state service model? of the national economy, or through their everyday purchases of goods and services in the High Street, or . What are the relative costs and effectiveness of when getting their monthly salary payment. In these alternative courses of action? ways they are exposed to economics the topic.This must be distinguished from economics the disci- . Why do the costs of ostensibly similar services vary pline [2]. Consequently, general dissatisfaction with so markedly? the direction or philosophy of macroeconomic policy . Do market forces raise efficiency? Do they enhance for a country as a whole or the level of taxation is not consumer choice? sufficient justification for criticizing the principles of microeconomic theory or their application in prac- . What are the economic consequences of an empha- tical analysis. One does not have to agree with the sis on ‘personalization’ (through personal budgets, economic policies of the government of the day and for example)? what they mean for price inflation, interest rates or unemployment to advocate the sensible pursuit of . To what extent can we recoup the costs of care from efficiency in a health system. Economists working in charges to service users? the mental health sphere are concerned partly with the economic topics of health care – expenditure . Is ‘allocation according to need’ substantially dif- limits, costs, the supply of qualified labour, market ferent from ‘allocation according to ability to pay’? forces, purchaser–provider contracts and so on – but more often with the application of the discipline of . How can we fairly allocate resources between dif- economics to practise and policy issues concerning ferent areas of the country? health and health care. We focus on the latter in this chapter. . Is there an economic case for investment in pre- Consequently, economics tackles a wide range of ventive strategies? issues, and by concentrating on efficiency in this chapter we do not mean to suppose that other areas In addressing and attempting to answer any of these are less important. Among the social or public policy questions economics cannot replace the judgements of questions that economics has successfully addressed, decision makers, but it ought to be able to inform those often in collaboration with other disciplines, are the judgements. It can help the decision maker formulate following: the practice or policy questions sensibly and logically, and then provide a range of answers from which the . What is the true cost of care? decision maker may choose, but it is obviously not the economist’s role to determine policies. Exploitation . What are the outcomes of services and treatments of the interplay of economic appraisal, political prio- and how do we measure them? rities, clinical expertise and – of course – theviews and preferences of the people affected is usually going to . What is the relationship between resources and be the most sensible way to proceed in a mental health outcomes? system.

HEALTH ECONOMICS AND PSYCHIATRY: THE PURSUIT OF EFFICIENCY 373 28.3 DEFINING EFFICIENCY The pursuit of efficiency is predicated on the assump- economic or underlying demographic characteristics tion – hardly contentious – that resources are scarce to have an influence. Hence access to, payment for or relative to the demands placed upon them. There will impact of evidence-based treatment are not fairly never be enough resources to meet all of society’s distributed by gender, ethnicity, age, language, reli- needs or wants, not because members of society are gion, income, socioeconomic group or place of resi- greedy, prodigal or selfish, but because need is relative dence. The challenge when discussing equity is in and not absolute. In provision of care it seems that deciding what is meant by ‘fair’. It would be seen by almost every time service boundaries are pushed most people as ‘unfair’ if everyone in the population forward to reach more individuals or families or to had exactly the same amount of support or treatment serve those people better who are already supported, from the mental health system, because most people so more needs are revealed and more demands sti- have no need for such support, while a few people have mulated. Scarcity suggests that it would be prudent to very considerable needs. An equitable allocation of make careful use of available resources, and econom- mental health resources would therefore usually mean ics has developed as the study of allocation under better access to treatment for people with greater conditions of scarcity. Efficiency is simply one criter- needs, or ensuring that those with the lowest ability ion for guiding or operationalizing the careful use of to pay for their treatment are charged lower amounts resources in the face of scarcity. than those with higher incomes. These are the two Efficiency refers to a situation in which the alloca- most commonly discussed aspects of equity: (i) tion of a given level of resources maximizes the whether individual financial contributions are linked outcomes of a system of care (or component part to ability to pay, indeed whether there needs to be a thereof) or which minimizes the resource require- redistributive effort so that individuals with lower ments to achieve specified ends. (We will define terms incomes contribute proportionately lower amounts, like outcomes and resource requirements in a and (ii) whether access to evidence-based treatment is moment.) Absolute efficiency is hard to conceptualize linked to type and level of need. for activities with such varied objectives as mental Thus, the scarcity of resources necessitates a choice health care. However, while it may be impossible to that should be based at least on the twin criteria of say what configuration represents the most efficient efficiency and equity, and decision makers may also utilization of resources, it ought to be possible to say want to introduce additional criteria such as auton- whether a policy or practice change results in a more omy, liberty and diversity. In this chapter the focus is efficient utilization. Relative efficiency is an accessi- on efficiency. ble concept in principle and practice. Defined in even these rather general terms, a change in clinical prac- tice or resource allocation that results in greater 28.3.1 A theoretical framework efficiency, without damaging clinical or other objec- tives, is undeniably a good thing. The crucial qualifier Most clinical research in the mental health field here is that efficiency is not, of course, the only includes an examination of effectiveness or outcomes, criterion for good resource allocation. based upon a description of the objectives of the The equity criterion refers to fairness in the dis- intervention being studied. These objectives are of tribution of outcomes, access and payments across two types. Final objectives are ends in themselves, individuals or parts of a society. Most mental health even though there will be causal connections between systems are inequitable: they do not distribute these them, while intermediate objectives are couched outcomes or responsibilities for payment fairly, in terms of means to those ends. Providing mental but instead (wittingly or unwittingly) allow social, health services where they are needed is really an

374 SOCIAL CONSEQUENCES AND RESPONSES intermediate objective, and so too is offering a sup- sense to try to define a cost. (Because these nonre- portive, high-quality environment in a residential source inputs do not have a cost, but exert an influence facility, for these are means to achieve various desired on outcomes, the discussion or study of efficiency changes in, for example, the mental and physical needs to include them. Some of the nonresource inputs health of individuals. The latter are final objectives are correlated with, and perhaps determined by, the in so far as they are desired for their own sake. Other resource inputs, but some are determined outside the objectives might include improving individual com- care system, and these need either to be comprehen- petence and independence in the activities of daily sively purged from an evaluation by rigorous design or living, promoting the well-being of relatives or infor- built into the analyses.) mal carers, and protecting the rights of other members Thebasicassumptionwearemakinghereisthatfinal of society. Deficiencies in respect of these objectives outcomes and intermediate outputs are determined by are generally taken to be needs, and changes along the the level and modes of combination of the many and dimensions spanned by these objectives are usually various resource and nonresource inputs. This imme- defined as outcomes (improvements, for example, are diately suggests that analogies can be drawn between equivalent to reductions in need). This suggests a the provision of mental health services and the terms, distinction between intermediate outputs and final concepts and tools of economics. What we have outcomes, depending on the objectives under discus- described is what the economist would recognize as sion or study. a model of production, where individual health and The objectives of mental health services are well-beingaretheproductsoroutputs.Thisproduction usually discussed by reference to hypothesized influ- ofwelfareperspectiveisnotsuggestingthatthesupport ences on outcome. Much applied research in psy- andtreatmentofpeoplewithmentalhealth problemsis chiatry is designed to test the impact on users of one mechanistic; nor does it seek to reduce the myriad or more service, drug or social situation, frequently interrelationships between outcomes and treatment using an experimental or carefully executed quasi- characteristics to simple summary formulae. experimental design to control for extraneous influ- The influence and relevance of any one factor on ences. The range of factors with a potential influence individual outcomes depends on a combination of on outcomes is broad. The personal characteristics, factors, the sequence in which they appear or are experiences and circumstances of individuals with experienced, and the marginality of the stimulus that mental health needs will be of particular importance, theybringtothecaresetting.Theproductionofwelfare but there are also the various social and family approach thus allows us to organize or locate research characteristics such as the quality of a care environ- on, or discussions of efficiency within, both a theore- ment and the availability of support and participant tical framework and a clinical or service delivery opportunities. context. We have described it in more detail elsewhere We can introduce a more specific terminology here. for other social and health care services (see, for Resource inputs can be defined as the tangible example,[3]and[4]),andinamentalhealthcontext[5]. resources such as staff, physical capital (including The point to stress is that the framework requires no housing), provisions and other consumable items that leapoffaith,nodeparturefromtheimplicit‘models’of go to create service packages and help achieve the service delivery underpinning mental health practice desirable outcomes. Associated with each of them is a and research. With this theoretical framework we can cost, this being a shorthand term and summary mea- now define efficiency with more precision. sure for all these resource inputs. Nonresource inputs, by contrast, are those determinants of outcomes that are neither physical nor tangible: they are embodied in 28.3.2 Definitions of efficiency the personalities, activities, attitudes and experiences of the principal actors in the mental health care system Effectiveness, as the term is conventionally employed, or process. For the nonresource inputs itdoes not make refers to a simple improvement in outcome following

HEALTH ECONOMICS AND PSYCHIATRY: THE PURSUIT OF EFFICIENCY 375 the introduction of an additional unit of input. An and efficiency are of equal importance in defining effective production process is simply a process that target efficiency. Two dimensions can be distin- produces or achieves something desirable. A neces- guished: horizontal target efficiency, the extent to sary but not sufficient condition for efficiency is which those people believed to be in need of support effectiveness. A mental health service or activity or treatment actually receive it, and vertical target cannot be efficient if it is not effective. A closely efficiency, the extent to which the available related concept is productivity, the capacity to pro- resources are received by those in need. duce, which can be simply defined as the ratio of Efficiency in whatever form is therefore a state- outcome to input. As we move from effectiveness and ment about, or measure of, the achievements of a productivity to efficiency we need to distinguish a service or policy and the resources and other factors number of component definitions. The terms that combine to secure them. Some of those influ- employed for them are not unique, even though the ential factors are difficult to measure or impossible concepts are widely employed. A process is techni- to cost – including the nonresource inputs intro- cally efficient when it produces maximum outcomes duced earlier – but they should not be overlooked. from given inputs. Price efficiency is attained when the For example, an evaluation of hospital and commu- various inputs are employed in such proportions as to nity locations as alternative places of treatment for a produce a given level of outcome at minimum cost. A particular group of individuals would either need to technically efficient production process that is also measure personal preferences, the social milieu of price efficient can be called cost-effective. different accommodation or day settings, staff atti- A cost-effective technique might not be ‘socially tudes and so on, or would need to employ a research efficient’. Cost effectiveness indicates only the most design, such as an appropriately pragmatic rando- sensible among alternative ways of doing some- mized controlled trial, which adjusted for their thing; it does not tell us whether we should be influences. The nonresource inputs are crucial inter- doing that thing in the first place. It does not, for vening variables between the resource inputs or example, tell us whether we are better allocating costs and the outcomes or effects of mental health money to fighting global warming or improving services. mental health services; nor must it ever pretend that We now turn our attention to the study of effi- it can. Full social efficiency is achieved when net ciency in practice, describing and illustrating the social benefits (the overall benefits to society minus most common of the ‘efficiency analyses’ used the total costs to society) are maximized. The move today: cost benefit, cost effectiveness, cost utility from effectiveness through to social efficiency thus and cost function analyses. These analyses generate proceeds logically: effectiveness is a necessary information needed for a variety of decisions, which prerequisite for technical efficiency; cost effective- can be summarized in five deceptively simple ness is defined with reference to the set of all questions: technically efficient and price efficient activities or techniques; and social efficiency builds on these . What intervention (treatment, drug, programme of lower-level concepts. support or policy approach) is more or most appro- One further concept has attracted attention. priate in given circumstances? Target efficiency is the efficiency with which resources are allocated to and among those for . When should the intervention be provided? whom receipt has been judged the most cost-effec- tive method of intervention [6]. It measures the . Where should it be provided? extent to which a particular principle of equity (embodied in allocation according to need or alloca- . To whom should it be provided? tion according to some individually based measure of cost effectiveness) is met in practice. Thus equity . How should it be provided?

376 SOCIAL CONSEQUENCES AND RESPONSES 28.4 EFFICIENCY IN PRACTICE: MEASURES AND ANALYSES In the fifteen years since the first edition of this book, well-validated instruments can provide empirical the main techniques to appraise efficiency have measures. become far more familiar to mental health care policy Thus far, the discussion of outcome measurement makers and practitioners. In principle these techni- will be familiar to noneconomists. The two optional ques – which we can gather together under the general further stages may be less familiar, and are generally heading of cost effectiveness (although later we will found only when economists are involved and are want to separate them) – are simple. The costs of an linked to the type of efficiency analysis undertaken. intervention that fall to all members of society and its One of these further stages is to place monetary values effects or outcomes are measured and set alongside on the outcomes; the other is to amalgamate the each other. If the outcomes are sufficiently good to outcome data into a single measure. Neither is without warrant the costs, then the project or service will look complexity or controversy, for the information needed attractive. Of course, usually two or more interven- is difficult and costly to obtain, and laden with value tions are vying for selection, and the comparative judgements (although not necessarily any more than outcome gains and costs of the two will need to be for other aspects or approaches to evaluation). examined. There are advantages of moving to these two Common to each technique of efficiency analysis optional stages. First, if outcomes are measured in are outcomes and costs. What are these two central monetary terms they can be directly compared with components of the production of welfare approach? costs: Does treatment X generate more benefits than it costs? Is service innovation Y worth doing? The second reduces the multidimensional changes result- 28.4.1 Outcomes ing from an intervention to a single indicator. Both make the data easier to manage and the findings easier Summarizing the discussion above, outcomes can be to understand, but these desirable characteristics are conceptualized and measured at two levels. Final also the principal drawbacks. Conflation to a single outcomes reflect the changes in individual well-being measure and reduction to monetary measures wastes (including health) along various dimensions, together information, perhaps losing indicators along some of with the impacts on relatives and other significant the many dimensions on which mental ill-health may actors. By contrast, intermediate outputs are defined have an impact. It may also unwittingly disguise the and measured in terms of the volume and quality of value judgements and methodological assumptions care, the use of which generates final outcomes. Other that have been made to create the measure. chapters discuss measurement of outcomes for psy- The Euro-Qol or EQ-5D is a commonly used single chiatric services; here we rehearse the approach com- measure that has at its base a measure of people’s monly found in economic evaluations. descriptions of their health [7]. The Health Utilities Outcome measurement proceeds through two obli- Index and the index developed from the Positive and gatory and two optional stages. Having identified the Negative Symptoms Scale, a recently developed schi- ‘well-being’ dimensions – perhaps symptoms, func- zophrenia-specific measure, are other examples [8,9]. tioning and quality of life – along which outcomes are The quality adjusted life year (QALY), derived from to be explored, the first stage is to develop or select only one of these instruments, conflates health status scales for each dimension. At the second stage the task and longevity, aiming to assess the quality of life is to assess the impact of the intervention along each enjoyed in each extra life year gained from better dimension over a period of time. Reference to treatment. QALYs are becoming more commonly the objectives of the intervention will help identify used in evaluations of new psychiatric drugs but the dimensions of well-being and judicious use of more slowly gaining ground in evaluations of mental

HEALTH ECONOMICS AND PSYCHIATRY: THE PURSUIT OF EFFICIENCY 377 health services and supports (see, for example, patient outcomes. Of course, cost data are used for References [10] to [14]). many purposes, including cost-of-illness studies Outcomes measure movement towards objectives. (which aggregate the overall resource impact of an They need to be net measures of change so observa- illness), pricing services for contracts or describing tions at one point in time are rarely sufficient. Out- expenditure flows. However, efficiency analyses comes should measure the differential effect of an require this fourth rule, and sole reliance on cost intervention – the ‘additional well-being’ that it gen- findings in the absence of outcome data is almost erates over and above what would have happened never going to be adequate. Conversely, neglecting without the intervention. This leads to four practical costs when making policy decisions, or when under- decisions. What kind of comparison is needed in order taking the evaluations that inform them, should also to measure the differential or net effect of care on prompt questions about adequacy and, indeed, scien- individuals? How often should this differential effect tific validity. be measured? For how long is it necessary to monitor Illustrations of the application of these cost rules these effects? How are errors and distortions asso- come from evaluations that address questions about ciated with measurement to be minimized? These are the locus of care. While most people have moved from all standard (if tricky) questions of research design long-stay psychiatric hospitals to community loca- that are considered at length in the evaluation and tions in England, there is still a large amount of service social statistics literatures. Outcome evaluation by relocation to be done in some other parts of Eur- economists is therefore no different in intent, method, ope [17]. Long-stay hospitals provide, for example, design or constraint from outcome evaluation by shelter, food, care, treatment and daytime activities all social psychiatrists. on one site. Yet as individuals move to the community, these supports often come from a range of different providers, including families. Estimating comprehen- 28.4.2 Costs sive costs in the community means the supports actually used need to be identified, and costs com- On the opposite side of the production of welfare pared with (say) more institutional settings using a relationship are the resource inputs, which can be similar scope (like-with-like comparisons). The same summarized by their costs. These have known or issues arise when comparing inpatient psychiatric hypothesized influences on the achievement of the treatment with treatment in a day hospital or at an intermediate outputs and final outcomes of interven- outpatient clinic [18,19]. A further example comes tions. Elsewhere we have discussed and illustrated a from children’s services. Intensive community-based recommended set of rules for estimating the costs of multisystemic therapy in foster care homes is cur- mental health services [15,16]. First, costs should be rently being piloted and evaluated in 18 local autho- measured comprehensively to range over all relevant rities in England as an alternative to custodial care for supports and components of a treatment or care very challenging young people, predicated on ‘package’. Second, the variations in costs that will encouraging cost and outcome findings from the be revealed between patients, facilities or areas of the US [20]. country when using this comprehensive approach With these understandings of outcomes and costs, should be examined for their policy and practice we can now introduce the main tools of efficiency insights. Third, like-with-like comparisons should be analysis employed within mental health economics. made; this may require some standardization for the We start with cost-effectiveness analysis, setting out influences of extraneous factors (through design or the stages through which an actual study would pro- statistical manipulation) to ensure that comparable ceed. Descriptions of some other analytical techni- samples of patients or facilities are studied. ques then follow. Drummond et al. [21] provide an The final rule is that cost information should be excellent thorough account of cost-effectiveness integrated where possible with information on analysis.

378 SOCIAL CONSEQUENCES AND RESPONSES 28.4.3 Cost-effectiveness analysis another then it is clearly the more cost-effective of the two. If, on the other hand, one intervention is more There are essentially six stages in a cost-effectiveness effective but also more costly, then the decision maker analysis (CEA): must decide whether the better outcomes are worth the higher costs. An example will help to explain the 1. Define the alternative interventions to be examined challenge. Cognitive stimulation therapy (CST) was in the analysis. found to generate better outcomes for older peoplewith dementia resident in care homes than care as usual, and 2. List the costs and effects (outcomes) to be costs for the two groups that were similar [11]. Under measured. these conditions, where the intervention is achieving better outcomes for the same amount of money, CST 3. Quantify and value the costs and effects. represents a more efficient use of resources. To help in deciding whether to invest in a new intervention, 4. Compare the costs and effects. economists use cost-effectiveness acceptability curves (CEACs) to provide additional and visually appealing 5. Explore the robustness of the findings using sensi- information on the probability of a particular interven- tivity analyses. tion being more efficient in given circumstances. In the CST study, the CEAC showed that if decision makers 6. Examine the distributional implications. were willing to pay £100, for example, for a one-point additional gain onthe cognition or quality of life scales, These stages represent a simplification of checklists the probability of CST being cost-effectivewould be 55 used to appraise the extent to which an economic and 85%, respectively. evaluation conforms to an ‘ideal type’, perhaps for The decision rules underlying such analyses are not publication or inclusion in an evidence-based always easy to apply in practice, particularly in studies guideline [22,23]. where some outcome dimensions register improve- At the first stage the exact nature of the policy or ments and others indicate deterioration. It is not the treatment options needs to be made explicit so that the task of the researcher to advocate a particular policy or research question is clear. Once identified, this ques- treatment option. Rather, it is their task to point to the tion determines the characteristics of the study; some various efficiency and other consequences and to leave policy or practice issues do not lend themselves to a decisions to the politician, clinical manager or budget straightforward evaluative design. The second stage is holder. to identify all likely costs and outcomes, drawing The costs and effects calculated for the various attention to those that might prove beyond the tech- options are likely to be subject to some error, and a nology, ethics or politics of measurement. It is impor- good CEA would include sensitivity analyses, exam- tant to be aware of all factors necessary to implement, ining the implications of different assumptions or likely to result from, the options being appraised, regarding the estimation of costs, effects and so on. even if they do not later figure in the empirical Modelling studies are not an ideal way to estimate evaluation. efficiency but provide useful information on the likely The next stages are to quantify and value the costs implications of a particular course of action. Such and outcomes in the manner described above. In cost- studies rarely collect primary data, but collate infor- effectiveness analyses there is no attempt to place mation from a number of sources (mainly previous monetary values on the outcomes whereas – as we studies and routine data collections) and employ describebelow–cost-benefitanalysisdoesattemptthis. assumptions based on expert opinion where data are In comparing costs and effects, the efficiency rule not available. Sensitivity analyses are important and wouldbetocomparethedifferenceincostsbetweenthe can generate a range of possible results, obtained by twointerventionsrelativetothedifferenceinoutcomes. varying the assumptions about costs and outcomes Ifone interventionis more effective and less costly than that have been incorporated into the model.

HEALTH ECONOMICS AND PSYCHIATRY: THE PURSUIT OF EFFICIENCY 379 Thismodellingapproachwasusedinastudyofearly outcomes and the latter does not. The CBA technique intervention (EI) services, which have become a cen- canbeusedtosaywhetherornotthebenefitsofaproject tralcomponentofacomprehensivecommunitymental or procedure outweigh the costs; this cannot be health system in England. Findings suggest that using achieved with a CEA because the resources (costs) EI reduces costs by about a third over 12 months and their impacts (effectiveness) are measured in dif- compared to standard care coordinated by a commu- ferent units. The inherent ease with which CBA results nity mental health team (CMHT) [24]. This model can be understood is a major advantage for this tech- blended data from research, audits, routine collections nique,butthedisadvantagesincludethedifficultiesand and expert advice. One of the aims of EI is to reduce additional value judgements associated with the inpatient admissions so the assumptions about the attachment of monetary values to outcomes. probability of initial admission and readmission were One recent study typifies the complications around potentially key and were investigated in the sensitivity the meaning of a CBA. This influential study explored analyses. Neither a 50% increase nor decrease in the the potential of increasing NHS expenditure on cog- probability of an initial admission had an impact on nitive behavioural therapy to alleviate symptoms of cost,butforthosereceivingstandardcare,reducingthe depression, which in turn would reduce the costs of readmissionrateby50%ledtothecostadvantageforEI employment-related difficulties [30]. Styled as a disappearing. Dramatic turnarounds from sensitivity CBA, it used data from previous epidemiological, analysesarerare,butthecostandoutcomeassumptions clinical and economic research to model the economic should nevertheless be checked. payoffs and showed that a nation-wide investment in Finally, the distributional or equity consequences of psychological therapists in England would pay for the different options should be addressed. In part, itself many times over. Herein lays the problem. In its these are ‘who’ questions. To whom should care be ideal form a CBA includes placing a monetary value provided? Many studies have found that costs are on final outcomes, i.e. changes in the welfare of influenced by personal characteristics and needs, participants. However, this study identified what is including age, gender, ethnicity, duration and severity effectively an intermediate output – return to work – of symptoms, personal and social functioning. In turn as the end-point of interest, and the proportion of these cost variations could imply differential effi- work-days gained from cognitive behavioural therapy ciency. Recent work, for example, has highlighted for depression can be valued as monetary gains to the the considerable age discrimination in mental health national economy. This study provided the evidence services and social care [25,26], in access to mental base that led directly to the £180 million Improving health resources in different areas of the coun- Access to Psychological Therapies (IAPT). try [27,28] and in levels of expenditure [29]. Too One other type of study is often mislabelled as a much reliance on efficiency may mean that distribu- CBA. These studies define ‘benefit’ as expenditure tional findings, and equity more generally, might not avoided or cost saved, which either implies that out- receive sufficient policy attention. comes are irrelevant or makes the strong assumption of identical outcome implications of the alternatives under consideration. These cost-offset calculations 28.4.4 Cost-benefit analysis can be informative but they do not assess efficiency. A CEA is designed to examine the technical and price efficiency of one or more mental health procedures; in 28.4.5 Cost-utility analysis contrast,acost-benefitanalysis(CBA)examinessocial efficiency. (Some economists would see CBA as a Cost-utility analysis (CUA) is the label attached to a special case of CEA; others would disagree. We do particular data configuration: it is a cost-effectiveness not need to delve into that debate here.) The stages of a analysis conducted with outcome measured in terms cost-benefit analysis exactly mirror those of a CEA of individuals’ ‘utility’ – a measure that captures except that the former places monetary values on the individuals’ preferences for health states that are

380 SOCIAL CONSEQUENCES AND RESPONSES differentiated by life expectancy and quality of life. 28.4.6 Cost functions A common metric is the cost per QALY gained for different procedures. (We introduced the QALY, qual- In most other fields where economists are active it is ity-adjustedlifeyear,earlierinthe chapter.)One advan- rarely possible to set up a controlled study. For exam- tageofthisapproachisthatthesameutilitymeasurecan ple, it is very hard to persuade companies to be ran- be used across different health problems. In turn this domly allocatedto differentmarket regimes.Similarly holds the potential to compare results across treatments national economies are unique, disallowing direct and disorders. The National Institute for Health and comparisons. Thus, applied economics research in Clinical Excellence (NICE) in England and Wales these other areas has developed a range of multivariate advocates the use of QALYs and assesses the social statistical techniques to get around the problems of efficiencyofinterventionsforalltheirclinicalguidelines less-than-ideal study designs. With sufficient data, the against a standardized value of decision-makers’ will- techniques allow exploration of cost differences ingness to pay for an additional QALY gain [23]. between units; in social psychiatry these units might Despite its increasing popularity there remain three beNHSTrusts,hospitals,areas,years,individualusers challenges to estimating QALYs in mental health and so on. A ‘statistical cost function’ is an attempt to contexts. First, although the approach provides the ‘explain’ observed cost variations. It is the estimated clinical, managerial or financial decision maker with a relationshipbetweenthecostofprovidingaservice,the set of precise-lookingstatistics, these may disguise the outcomes,resourcepricesandotherfactorsthathavean complexity of service provision for people whose hypothesizedinfluenceoncost,suchastheneedsofthe service needs are many and wide-ranging. Second, individuals using the service, the arrangement and by using only the single outcome dimension, the organization of care, and the broader local context of effects of the interventions on the many areas on care[32,33].Thesimpleproductionofawelfareframe- which mental health problems and their treatments work set out above provides a theoretical structure may impact – positively or negatively – may be within which to suggest these potential influences and missed. Third, there are worries that the available to interpret their links to costs. generic utility-generating instruments, the best known The cost function takes a rather different approach being the EQ-5D [7], are insufficiently sensitive to the from CEA, CBA and CUA. It is partly a statistical changes in symptoms and functioning typically technique, and has most commonly been estimated recorded in mental health studies. This particular with cross-sectional data for a sample of ‘production measure has been found not to work well in evalua- units’ that are known or assumed to have reasonably tions that include people with psychosis, personality similar objectives and to employ reasonably similar disorder, child and adolescent problems and demen- production techniques. There is no engineer’s tia [31]. We would certainly not suggest that blueprint to guide the study of cost variations, and QALYs should be excluded from economic evalua- the model does not suggest that ‘production’ is rou- tions of social psychiatry interventions, butto use such tinized or standardized, but ‘behavioural cost measures unthinkingly or in the absence of other functions’ have been estimated to good effect for measures will risk missing information valuable to mental health facilities, for administrative areas such providers and commissioners as they try to respond to as municipalities and for individual users of the needs of their population within limited resources. services [25,29,34–37]. 28.5 THE PURSUIT OF EFFICIENCY Efficiency is still not a concept that is widely under- experiencing distressing symptoms or of families stood, and partly as a result it is not a concept that is carrying a heavy burden of unpaid care. Similarly, warmly welcomed by professionals charged with requests for and utilization of cost information are still meeting the clinical and wider needs of individuals regarded with some suspicion. However, awareness of

HEALTH ECONOMICS AND PSYCHIATRY: THE PURSUIT OF EFFICIENCY 381 the need to understand what resources are used in (say) clinical trials of medications include a cost-effective- the delivery of treatment or the implementation of a ness evaluation, and increasingly so too do evaluations new policy is considerably more widespread today of psychological therapies and service arrangements. than it was twenty or even ten years ago. That same commitment by the research community Legislation, national service frameworks, clinical and the bodies that support it through funding does guidelines, regulatory frameworks, auditing and the sometimes appear to be lacking in the policy com- introduction of the ‘internal market’ in Britain’s munity, where grand decisions have been taken to National Health Service have all reinforced the mes- reorganize or restructure the health system (broadly sage that resources are not limitless. They have also defined) in an evidence-free way. emphasized that careful (which is now usually inter- Two further challenges are linked to the compo- preted to mean ‘evidence-based’) decisions have to be nent elements of any efficiency analysis or discus- taken about how to deploy those resources. As will be sion: the measurement of outcomes and the measure- crystal clear from our arguments in this chapter, we do ment of costs. Outcome measures that attempt to pick not find this trend to greater efficiency emphasis to be up changes in symptoms, personal functioning, qual- regrettable provided that it is implemented advisedly ity of life and so on are many and generally of a good and managed wisely. For example, misinterpretation calibre. The relatively recent emphasis on addition- of what efficiency means, or inadequate measurement ally measuring a generic outcome such as utility of the core components of outcomes and costs, or (most readily seen in the quality-adjusted life year) na€ ıve interpretation of the evidence gathered can each still has a little way to go before decision makers in mean that the pursuit of efficiency does not achieve its mental health systems will be convinced that they objectives. This is why the pursuit of efficiency in a have available to them a tool that will deliver reliable, mental health system still has some distance to travel. sensitive, valid and meaningful indicators of achieve- What, then, are the challenges for the future? One is ment. On the cost side, available methods and results to ensure that the concept of efficiency is widely and are generally more satisfactory, although it is still too properly understood by all of those individuals common for empirical research to focus narrowly on charged with taking decisions within the mental health the costs falling to health services and ignore the system. These could be major countrywide decisions costs relating to social care, housing and other ser- about some new national policy initiative, or very vices frequently used by people with mental health local decisions about the expansion of a particular needs. community initiative, or practice decisions about As these challenges suggest, there are tasks for treatment modalities for people with particular con- economists and other researchers, in particular to figurations of need. Every one of those decisions needs improve measures so that they better reflect the cir- to have one eye on the likely or expected outcomes cumstances and experiences of people with mental gauged in terms of improvements in health and quality health problems. It also behoves policy makers, man- of life, but the other eye needs to be on the resources agers and practitioners at all levels to ensure that they expended in the achievement of those outcomes. demand and use economic evidence in their everyday Another challenge for the future is to ensure that the work.If the overriding aim of the mental health system accumulation of new evidence on the effectiveness of is to meet more needs and further improve the health treatments, community arrangements, policy initia- and well-being of its users, then efficiency must be a tives and so on is accompanied by the collection of key objective, because – basically – it means striving information on the resources expended. Today most to achieve more from available resources. ACKNOWLEDGEMENTS This work was undertaken within the Personal Social theDepartmentofHealth,althoughallresponsibilityfor Services Research Unit’s core programme, funded by the contents rests with the authors alone.

382 SOCIAL CONSEQUENCES AND RESPONSES REFERENCES 14. NICE (2004) Schizophrenia: Full National Guideline on Core Interventions in Primary and Secondary Care, 1. Williams, A. and Anderson, R. (1975) Efficiency in the The British Psychological Society and Gaskell, Social Services, Basil Blackwell, Oxford. London. 2. Culyer, A. (1981) Economics, social policy and social 15. Knapp, M. and Beecham, J. (1990) Costing mental administration: the interplay between topics and health services. Psychological Medicine, 20, 893–908. disciplines. Journal of Social Policy, 10, 311–329. 16. Beecham, J. and Knapp, M. (2001) Costing psychiatric 3. Davies, B. and Knapp, M. (1988) The production of interventions, in Measuring Mental Health Needs, 2nd welfare approach: evidence and argument from the edn(ed.G.Thornicroft),Gaskell,London,pp.200–224. PSSRU. British Journal of Social Work, 18 (suppl.), 17. Mansell, J., Knapp, M., Beadle-Brown, J. and 1–11. Beecham, J. (2007) Deinstitutionalisation and Commu- 4. Knapp, M. (1984) The Economics of Social Care, nity Living – Outcomes andCosts: Report of a European Macmillan, London. Study, Tizard Centre, University of Kent, Canterbury. 5. Knapp, M. and McDaid, D. (2007) Mental Health 18. Beecham, J., Sleed, M., Knapp, M. et al. (2006) The Policy and Practice Across Europe (eds M. Knapp, costs and effectiveness of two psychosocial treatment D. McDaid, E. Mossialos and G. Thornicroft), Open programmes for personality disorder: a controlled University Press, Maidenhead. study. European Psychiatry, 21, 102–109. 6. Bebbington, A. and Davies, B. (1983) Equity and 19. Priebe, S., Briscoe, J., Jones, G. et al. (2006) Effec- efficiency in the allocation of the personal social ser- tiveness and costs of acute day hospital treatment as vices. Journal of Social Policy, 12, 309–330. compared with conventional in-patient care: a rando- mised controlled trial. British Journal of Psychiatry, 7. EuroQol Group (1990) EuroQol – a new facility for the measurement of health-related quality of life. Health 188, 243–149. Policy, 16, 199–208. 20. Aos, S., Phipps, P., Barnoski, R. and Lieb, R. (2001) The ComparativeCostsandBenefitsofProgramstoReduce 8. Horsman, J., Furlong, W., Feeny, D. and Torrance, G. Crime, Washington State Institute for Public Policy. (2003) The Health Utilities Index (HUI): concepts, measurement properties and applications. Health and 21. Drummond, M., Sculpher, M., Torrance, G. et al. Quality of Life Outcomes, 1 (5), 54. (2005) Methods for the Economic Evaluation of Health Care Programmes, Oxford University Press, Oxford. 9. Lenert, L., Sturly, A., Rapaport, M. et al. (2004) Public preferences for health states with schizophrenia and a 22. Drummond, M. and Jefferson, T. (1996) Guidelines for mapping function to estimate utilities from positive and authors and peer reviewers of economic submissions to negative syndrome scale scores. Schizophrenia the BMJ. British Medical Journal, 313, 275–283. Research, 71, 155–165. 23. NICE (2008) Guide to the Methods of Technology 10. Byford, S., Barratt, B., Roberts, C. et al. (2007) Eco- Appraisal, National Institute for Health and Clinical nomic evaluation s of a randomized controlled trail for Excellence, London. anorexia nervosa in adolescents. British Journal of 24. McCrone, P., Knapp, M. and Dhanasiri, S. (2009) Psychiatry, 191, 436–440. Economic impact of services for first-episode psycho- 11. Knapp, M., Thorgrimsen, L., Patel, A. et al. (2006) sis: a decision model approach. Early Intervention in Cognitive stimulations therapy for people with demen- Psychiatry, 3, (4), 266–273. tia: cost-effectiveness analysis. British Journal of 25. Beecham, J., Knapp, M., Fern andez, J.-L. et al. (2008) Psychiatry, 188, 574–580. Age Discrimination in Mental Health Services, PSSRU 12. McCrone, P., Knapp, M., Proudfoot, J. et al. (2004) Discussion Paper 2536. Cost-effectiveness of computerized cognitive beha- 26. Forder, J. (2008) The Costs of Addressing Age Dis- viour therapy for anxiety and depression in primary crimination in Social Care, PSSRU Discussion Paper care: randomized controlled trail. British Journal of 2538. Psychiatry, 185, 55–62. 27. Beecham, J. (2005) Access to mental health supports in 13. NICE (2003) Depression: Management of Depression England: crisis resolution teams and day services. in Primary and Secondary Care, Guideline No. 23, The International Journal of Law and Psychiatry, 28, British Psychological Society and Gaskell, London. 574–587.