Content Server 1

Health and Social Costs of Unemployment Research and Policy Considerations RAMSAY LIEM PAULA RAYMAN ABSTRACT: This article selectively reviews recent studies of the social and private costs of unemployment. Although the experience of joblessness differs considerably according to a wide variety of circumstances, it is apparent that prolonged unemployment is commonly a serious threat to health and the broad quality of life. These costs, furthermore, are borne not only by individual workers, but also by their families and communities. The implications of this research for economic and social policy are discussed and contrasted with policy development premised on a benign view of unemployment. Priorities for future research concerning the impacts of unemployment and a declining economy are also proposed, placing special emphasis on the interdisciplinary character of these concerns.

Unemployment is a major issue in economic and social policy. It is generally assumed that, since the Great Depression, the most devastating consequences of unemployment have been minimized by social welfare programs. Some have even argued that most contemporary joblessness is frictiohal; that it is the inevitable by-product of a fully productive and open economy in which some people are always in transition, entering and leaving the labor market by choice. This is a benigrTview of joblessness. Such a conceptualization is reflected in a "full employment" economic policy premised upon a base level of unemployment. The problem of unemployment, however, is considerably more complex. With the enormous increase in the of;ficial unemployment statistics over the last two years, public policy debates are reexamining issues related to unemployment. A diverse literature, with contributions from the behavioral, medical, and social sciences, has accumulated that can help us better understand some of the social and health consequences of unemployment. Collectively, these studies do not portray job loss as a source of dramatic, overwhelming stress and disorganization for everyone who ex1116 • OCTOBER 1982 • AMERICAN PSYCHOLOGIST

Boston College Brandeis University

periehces it. However, there is good evidence that losing one's job can increase health risks, exacerbate chronic and latent disorders, alter usual patterns of, health^seeking behavior, and exact numerous other social and interpersonal costs. This commentary does not review all of this research. We try, initead, to illustrate the range of work in this area, outline possible policy directions compatible with the collective findings, and recommend priorities for continuing research on this deepening social problem.

Research on Unemployment Research on unemployment and related concerns has been reported for nearly two decades. However, the line of study initiated oh a large scale by Harvey Brenner (1973, 1976) has clearly been the catalyst for centering attention on this area of work. This research strategy tries to estimate the strain produced by unemployment by looking at how indicators of state or national economic performance relate to measures of strain (e.g., hospital admissions for psychiatric impairment, deaths related to cardiovascular disease). Brenner has found that such indicators-of strain as first admissions to psychiatric hospitals, rates of infant mortality, deaths from cardiovascular and alcohol-related diseases, and suicide rates increase sharply during periods of economic decline. The most powerful We would like to thank Thomas Atkinson, Joan Lieni, and Robert Reid for their careful reading of an earlier draft of this article Some of the research reported in this article was supported by the Center for Work and Mental Health, the National Institute of Mental Health, Grant MH 31316 to Joan Liem and Ramsay Liem and Grant MH 33251 to Paula Rayrrian and Barry Bluestone. Requests for reprints should be sent to Ramsay Liem, Department of Psychology, Boston College, Chestnut Hill, Massachusetts 02167.

Vol. 37, No. 10, 1116-1123 Copyright 1982 by the American Psychological Association, Inc. 0003-066X/82/3710-1116$Oa75

economic indicator in these analyses consistently has been the^ unemployment rate (or some derivative such as the number of persons employed in the labor force). Brenner has adopted an essentially exploratory attitude toward these analyses, relying on replication with different populations, areas, and indicators of strain to establish the reliability of the findings. In spite of methodological criticisms of this work (Marshall & Punch, 1979), reverse patterns of association between economic change and strain for some groups in the population (Brenner, 1979), and differences in the interpretation of main findings (Catalano & Dooley, 1979), these data challenge the benign view of joblessness. "Ecological" studies of economic change and well-being continue to make contributions to the literature on unemployment. By ecological we simply mean the strategy of examining aggregate eco, ndmic and health indicators as a source of information and hypotheses about how economic events can influence health status. Two especially noteworthy observations have been the most recent product of such activity. First, Brenner (1979) and Eyer (1977) have found that short, rapid upturns ,in the economy also predict increased rates of mortality from a variety of causes and psychiatric distress. These data are consistent with a hypothesis at the center, of the study of stressful life events (Dohrenwend & Dohrenwend, 1974, 1981); this hypothesis states that demands for change and readjustment per se, rather than desirability of such change, are the source of stress in a social event. These new findings by Brenner and Eyer do not negate the stress of unemployment. The decline in well-being and the increase in social costs that accompany increasing unemployment rates remain the stronger association. Furthermore, a short period of rapid upturn in the economy could also mark a time when fallout from the previous downturn is experienced (e.g., reemployment at a lower scale, continuing joblessness when others are returning to work). A second important development in this line of research comes from the work of Catalano, Dooley, and Jackson (1981). They have introduced a third variable in the basic time-series design—psychiatric symptoms surveyed on a monthly basis for a representative sample of the study population. Brenner has argued that economic decline and increased unemployment raise the demand for mental health services by provoking greater psychological impairment in the population. Catalano et al. reasoned that if this process occurs then eco-

nomic change should increase symptoms in the overall population (which would in turn raise the demand for mental health services). However, on two occasions these investigators have found that estimates of symptomatology for whole communities do not appear to respond to local unemployment rates in ways that might account for increased psychiatric admission. Catalano et al. therefore offer an alternative explanation for the process linking economic change and hospital admissions. They suggest that downturns in the economy create a climate in which some families become less tolerant of existing impairment in a family member or face the loss of resources needed to care for such a person in the home. Thus, demand for formal services increases. However, it is not at all clear that a representative sample of a population should exhibit changes in symptoms in response to area-wide economic trends. For this to happen we would have to assume, for example, that monthly changes in the employment index for the city of Boston somehow initiate stress reactions among a large percentage of the 750,000 residents of that city. Although we do believe that the "spread of effect" from joblessness is substantial, it is unlikely that indicators of strain based on the entire community population will be heavily influenced by small changes in economic activity—unless, of course> the majority of the community is directly experiencing unemployment, such as in Jahoda, Lazarsfeld, and Zeisel's (1971) Marienthal. It is also important to understand individualand group-level experiences that mediate the system-level findings of ecological studies. To do so, however, we need methods of research appropriate to these smaller units of analysis. Although the body of work is limited, there are several distinct types of contributions from the psychological, sociological, and epidemiological literatures that address, the problem of unemployment on personal arid group levels. One research strategy, which we will not review in detail, is the study of stressful life events (for a comprehensive introduction, see Dohrenwend & Dohrenwend, 1974, 1981). Scale development for determining recent stress experience has generated an enormous quantity of research indicating that cumulative social stressors play a role in precipitating and predisposing individuals to impaired physical and emotional health. Some evidence exists in these studies that negative job-related events (and especially unemployment) are commonly among the stronger predictors of health strain (Coates, Moyer, & WellAMERICAN PSYCHOLOGIST • OCTOBER 1982 • 1117

man, 1969). The basic thrust of this line of re- views were conducted over a one-year period. All search,.however, emphasizes the cumulative stress families had at least one child under the age of 18 of all the events to which a person is exposed over living in the home. Each "unemployed" family a period of time. Consequently, the observation was matched pairwise with a control (i.e., emthat job-related events are especially problematic ployed) family on such characteristics as occupahas not been pursued with any special interest tion of husband, work status of wife, family lifewithin this research tradition. cycle stage, and locality. One of the few detailed studies focusing directly The research has a number of objectives conon the worker facing unemployment is the work cerning the process of family arid personal response of Kasl and Cobb (1979). This two-year panel study to unemployment. We will, however, comment on followed a group of 100 married, male, blue-collar just two preliminary observations from our interworkers displaced by two plant closings, compar- views. ing it to a sample of controls on a variety of inBeing without work, for husbands, was strongly dicators of health and economic strain. Data from associated (at one and four months after loss of job) this research have been cited as evidence that plant with higher levels of psychiatric symptoms relative closings can cost workers their physical and emo- to control husbands. This difference was also sigtional health as well as undermine their economic nificantly greater at the fourth month than at the security. The findings are complex, however. time of the first interview. Interestingly, unemThese investigators caution against drawing too ployed men who became reemployed by the fourth strong a conclusion from these data that unem- month initially reported symptom levels compaployment inevitably impairs health. Although this rable to those of continuously unemployed men, outcome appears to be true for physical ailments but appeared even less stressed than controls (con-like coronary disease, dyspepsia, and hypertension, tinuously employed) following their return to emotional health is not uniformly undermined. work. In other words, for those unemployed and Responses for unemployed men generally indicate then reemployed within four to five months, the greater stress than what is reported by working strain evident in psychiatric symptoms was initially men. However, elevated depression, anxiety, and no different from that of workers who remained somaticism appear to occur only as brief, initial without a job throughout this period. Following a responses for some workers (Kasl, 1979), whereas return to work, however, there was almost a comfor other workers the emotional strain does not pensatory effect, as indicated by the drop in sympabate even when unemployment ends (Kasl & toms to a level below that of controls. This pattern Cobb, 1979). One might even contend that for was common for 'five of nine symptom clusters some of these persons psychiatric symptoms are measured in this research (depression, anxiety, hoschronic or characterological and predict length of tility, paranoia, and somaticism). It was reflected unemployment rather than respond to it. Our view in the experience of both blue-collar and whiteis that health costs of job loss are attenuated in this collar workers, although significantly more so for , sample by virtue of a number of circumstances the former group. The reemployed group is critical in this analysis like the regional employment climate at the time of these plant closings. What remains indetermi- because its collective experience over time is the nate is how typical these findings are for general basis for our relatively strong inference that ununemployment in the population. We shall return employment causes rather than responds to emoto this kind of problem following a brief review tional strain. These data stand in contrast to some of research programs with which each of us is as- of the patterns reported by Kasl and Cobb. In support of our findings we should also note that a very sociated. similar pattern of results is reported by Warr (Note 1) from two panels of psychiatric data for young THE WORK AND UNEMPLOYMENT PROJECT English adults. This English study also includes a The work and unemployment project (Liem & sample of young men who became unemployed Liem, 1979) is a panel study involving approxi- during the study period, and these respondents inmately 40 blue- and 40 white-collar families fol- dicated a significant rise in general symplowing the involuntary loss of jobs by the husbands. ' tomatology. The second observation we wish to make from The most common reason for unemployment was layoff due to cutbacks, although some workers ex- our research relates to the psychiatric status of perienced plant shutdowns. Four intensive inter- wives. No significant differences in symptoms were 1118* OCTOBER 1982 • AMERICAN PSYCHOLOGIST

reported by spouses of employed and unemployed men during the first interview other than greater depression among spouses from families facing job loss. By the second interview, though, wives in unemployed families were significantly more depressed, anxious, phobic, and sensitive about their interpersonal relationships than spouses from control families. Wives of reemployed men were indistinguishable from their controls at this time. Our hypothesis regarding this pattern of effects among wives is that the stress of the husband's job loss for wives was mediated by the impact of unemployment on the family system. These wives did not experience the personal loss of a job and the attending guilt and responsibility. They were, however, exposed to a changed family environment, reflecting, among other things, changes in mood and behavior of family members—especially the husband. We have assumed all along that a structural interdependence exists between the family and workplace (Kantor, 1977) that is threatened by the event of unemployment. Adaptive changes are inevitable within the family, and they become one potential source of stress for all family members. This scenario is supported by convergent and independent observations by both spouses that only the husbands' role performance and supportiveness in the family is diminished during the first months of job loss. Wives' role performance and supportiveness deteriorate only after several more months of joblessness. Measures of family cohesion, organization, and conflict also indicate that unemployed families become significantly more stressed than controls by the fourth month of unemployment. We believe that wives' increasing symptoms, as their husbands continue to be without work, are largely responsive to these strained relationships and increased disorganization in the family. From the perspective of traditional models of the etiology of mental illness, this process illustrates how social context factors can play a role in the development of impairment and how the immediate social context mediates the person's ties to the wider social environment. These findings also bear out the notion that unemployment does not simply happen to individuals. The vast structured and less formal interdependencies that describe present-day social relations are the channels through which families, ^extended kin networks, and whole communities are potentially affected by the joblessness of individuals. The most striking finding from this research relative to those reported by Kasl and Cobb is the

clear indication that emotional strain was a direct consequence of work loss in this group of workers.1 As we noted before, one explanation for this difference may be found in the particular circumstances of the unemployment in both studies. Although information is not available to clearly identify the major differences, we suspect there are many. For example, the median length of unemployment during the first year in our sample was approximately 18 weeks, as opposed to about 5 weeks in the plant-closing research. Workers studied by Kasl and Cobb appear to have been separated from work during a generally healthy period in the wider economy, whereas our respondents' job loss preceded by about a half year the start of the worst recession in the. Northeast since World War II. Additionally, Kasl and Cobb noted that plant closings create a type of unemployment in which the likelihood of self-blame is probably minimized, at least in the short run. Unemployment in these two studies, then, is probably not the same experience, an obvious conclusion that nevertheless could easily be overlooked if one's attention were solely restricted to individuals without jobs. ' The enduring issue for studies of individuals facing job loss is that, just as competing hypotheses generated by ecological studies can be guided toward resolution by studying the personal experience of job loss, accurate interpretation of microlevel research depends on knowledge of the economic and social contexts of job loss. Rayman and Bluestone (Note 2), for example, found important differences in the job loss experience according to workers' union affiliation and whether they worked in prime or subcontract firms. A variety of methodologies are required to study unemployment in all these aspects—approaches that cannot be found in any single discipline. A similar conclusion was reached at a recent, interdisciplinary conference on the economy and mental health (Ferman & Gordus, 1979), where it was urged that economic and stress variables be studied simultaneously at several conceptually meaningful points of analysis. 1 We should also note our belief that absolute levels of emotional strain reported by our respondents underestimate the intensity of distress actually experienced by many persons in like circumstances. Persons who refused to participate in this research and those who dropped out at some point commonly excused themselves because of the hardships of their unemployment. Another indication of real strain among the unemployed is three times as much marital separation in this group as opposed to controls during the study period (seven cases). One suicide also occurred involving an unemployed husband who participated in a pilot interview.

AMERICAN PSYCHOLOGIST • OCTOBER 1982 • 1119

HARTFORD PROJECT

the job loss experience and elicited information The Hartford project (Rayman & Bluestone, Note regarding coping responses. While final analysis of the Hartford Project data 2) is the first interdisciplinary unemployment is in progress, a number of salient findings can be study to integrate investigation of a specific inshared. Foremost is the observation that the overdustry and its labor force, the response of a specific whelming majority of respondents who expericommunity to periods of bust and boom in this enced job loss in this industry sometime during a, industry, and joblessness among individual work10-year period indicated related periods of serious ers. An explicit effort was made to describe much physical or emotional strain as well as financial of the relevant context of job loss as a necessary hardship. High blood pressure, alcoholism, inpart of the assessment of the costs of unemploycreased smoking, insomnia, neurasthenia, and ment. The project developed a systematic fourworry and anxiety were among the more compart approach to link the private costs of job loss monly reported forms of strain. for individuals and their families to the social imMiddle-aged heads of households with young pact of unemployment on the community. The dependents experienced more intense stress effects research thus provided an opportunity to explore than younger, single workers. The type of job loss the "rippling" effect of unemployment from the experienced by minority workers was most apt to unemployed individual to the "unemployed fambe discharge, and they had the hardest time findily" to the "unemployed community." ing a new job in the industry. Female aircraft The four-stage approach began with a study of workers who held blue-collar jobs were much more the aircraft industry, including its segmentation likely than their male counterparts to be thrust out into prime and subcontractor firms, union and nonof the primary sector industry and into more marunion shops, and its internal labor market strucginal, unskilled jobs that meant lower pay and ture. Interviews were conducted with representafewer benefits. Unionized workers were out of tives of 15 aircraft firms to determine industry work longer than nonunionized workers, but they policies concerning labor recruitment and labor were more likely to be recalled to their former displacement. The political economy of the Hartford (Connecticut) metropolitan area was also stud- places of work. Each of these circumstances of unemployment ied to understand what role the aircraft industry influenced the specific experience of job loss for played in its history. these workers, with implications for both the duLabor force questions were then addressed. ration and intensity of stress reactions. Although Based on a comprehensive analysis of the Longi2 it is not possible to model all the relevant factors tudinal Employer-Employee Data (LEED) File that dictated the precise nature of these workers' we estimated some of the main characteristics of reactions to unemployment, these data clearly the total labor force in this industry in the Hartford demonstrate the range of specific conditions that area. One fourth of all jobs in this region are in can influence the course of the job loss experience. manufacturing, and approximately 87% of these A final and most significant contextual feature are in the aircraft industry. Thus, when the bust of unemployment in this study is the fact that most and boom periods endemic to military defense of the job loss occurred during periods of overall work hit the industry, the impact on the combusts in the aircraft industry. Being out of work munity is widespread. (We also learned that 16% meant virtually no opportunity to reenter the inof the workforce were women and 9% were midustry until it moved into a boom cycle. And benorities.) Overall, this is an aging population of cause the vast majority of manufacturing jobs in workers with fewer and fewer younger persons the Hartford area are in the aircraft industry, entering the industry. reemployment in the primary sector was essenStages three and four of the project involved tially closed off. We suspect that these conditions mailing a questionnaire to industry workers and are a major part of the reason why so many of the conducting personal interviews with 80 workers selected from the 206 respondents to the question2 The LEED file is maintained by the Social Security Adnaire. We used the LEED file data to create ministration and contains a random sample of approximately weights for different classes of workers in the sam- 1.5 million individual employee records. Each record is a social work history of an individual for the period 1957-1975. ple to enhance the representativeness of the data. security Age, race, and sex are available for each employee as well as Both the questionnaires and the interviews assessed all SSA-covered employers and specific counties of employment the financial, physical, and mental health costs of for each person. 1120 • OCTOBER 1982 • AMERICAN PSYCHOLOGIST

workers we reached experienced real health as well as economic costs during their unemployment. UNEMPLOYMENT AND SOCIAL POLICY

An important determinant of public policy regarding unemployment is whether one assumes that job loss provokes ill health, uncovers illness, or has negligible health consequences. If we look to current research on job loss for guidance in considering these alternatives, we might conclude that none of them is the superior choice and that no judgment can be made. On the other hand, policies relevant to the problem of unemployment are constantly being formulated and implemented. These policies do take a stand on this question. In our view a common assumption about unemployment is that in most cases joblessness is not a serious threat to personal health, family well-being, or the overall quality of life in our communities. Consequently, it is not possible to withhold judgment about the health and social costs of unemployment. Inaction is tantamount to passive adoption of such prevailing views. As our review of this literature has implied, we believe that existing research findings (enriched by many personal accounts of job loss) indicate that unemployment does constitute a serious problem for affected individuals, families, and communities. At the same time, it is not a uniform experience. Different scenarios of economic, social, and personal dynamics can give job loss different meanings. Nevertheless, the primary assumption informing overall policy on unemployment should be that job loss brings costs to personal health and the quality of family and community life. This premise, together with the social ethic of "the right to work," should figure centrally in the formulation of unemployment policy. The practical implications of this policy orientation can be expressed in the familiar terms of primary and secondary prevention. The most fundamental implication for primary prevention involves radically reshaping our sense of unemployment as not simply a problem in economics but as a factor in human and social welfare. To view job loss in this way is to recognize a legitimate interest of the helping professions in the formulation of broad economic policy that influences the amount and distribution of unemployment in the population. Similarly, where policies result in heightened joblessness, there must be accountability not only

for the economic consequences to those affected but also for the personal and social costs. This principle is contained in more encompassing economic policy alternatives such as those suggested by Sclar (1982). The basic assumption in these reformulations of responsible economic policy is that the decisions of companies and firms in a locality affect not only the economic prosperity of the area but also the broad quality of life and health of the community. Much as the quality of the physical environment has gained some legitimacy as a criterion for economic decision making, the public interest defined in these broadest of terms must be made a serious concern of those who determine economic priorities. There are some indications that this general policy orientation has gained some acceptance among advanced industrial nations in Europe. Britain, for example, requires a minimum period of notification before workers can be laid off. Sweden has developed elaborate job retraining programs, financial support packages for ailing companies, and pension insurance for older employees (for a summary of these kinds of programs, see Kelly & Webb, 1980). In the United States it is apparent that the first step is to broaden the terms of debate concerning economic policy. Regarding the design of specific supportive interventions that can soften the impact of unemployment, there are very few research data to direct our efforts. One reason for this condition is that most research thus far has been oriented to the outcomes or consequences of unemployment rather than to the mediating processes. We know very little about what kinds of coping strategies, informal supports, and formal services are best suited to different circumstances of unemployment. What is apparent, however, is that there is an educative function to be served. What we are learning about the relationship of the local, state, or national economy to individual health should be shared widely among professionals responsible for social service programs, health practitioners, community organizers, and religious leaders. One lesson we learned from the Hartford project was that unemployed workers and their families turned primarily to close relatives for help. They were generally unfamiliar with community agencies that could be of service to them, and we found that social service providers, unions, and companies were uniformly unresponsive to the needs of this group (Rayman, 1982). One response to this kind of situation has been developed by Akabas, Weiner, and Sommer (1973). Representatives of AMERICAN PSYCHOLOGIST • OCTOBER 1982 • 1121

labor, business, and the helping professions were engaged in dialogue around problems of common concern, like the aftermath of unemployment. The program is the process of interaction among these groups. It provides the vehicle for the development of concrete policies relevant to local needs and circumstances and mutually supported by the interested parties. There are, of course, other vital concerns that deserve attention in a comprehensive formulation of policy that addresses the problem of unemployment. For example, in our research workers typically expressed considerable anxiety about the loss of medical benefits during periods of unemployment. One of the most important policy recommendations coming directly from our respondents would thus be to create a system of medical insurance not tied solely to current employment. It is certain that continuing research in this area will reveal more of the inadequaciesin our current response to those without work. We have tried to indicate some of the general lines of policy development that we believe are consistent with the existing body of work on job loss.

employment. One especially important need is for research on several groups generally excluded from formal studies of the costs of job loss. Women and minorities shoulder vastly disproportionate shares of this nation's joblessness, and yet we know least about their encounters with unemployment. Most of the limited formal research on the health consequences of losing one's job, in fact, centers on the experience of relatively short-term dislocation among generally stably employed white men. Our own research projects will provide some new data for female and minority job losers, and new research by Rosen (Note 3) studies unemployment among women in the garment industry. Nonetheless, the experience of these groups remains woefully underrepresented in the informational resources available for national policy development. Both of our research programs also demonstrate the need for further research on the short- and long-term effects of unemployment on children. Workers expressed great concern about their inability to be good providers for their children. They also described various signs of stress in the children themselves—moodiness at home, new problems in school, and strained relationships with RESEARCH PRIORITIES peers. Besides these more immediate kinds of efSeveral brief comments on future research prior- fects, there is also the question of the effect of a ities are an appropriate conclusion to this review family history of job loss on a child's view of the of unemployment research and related policy im- world of work. When parents are unemployed, plications. First, and most generally, all levels of what happens to a child's sense of the place of social organization are implicated in the problem work in his or her life as a source of basic material of job loss and its consequences. Unemployment security, a primary mode of self-expression, and is therefore inescapably a problem for interdis- a vehicle of interrelatedness'to others? And what ciplinary research requiring conceptual tools that are the costs to society? These are some of the issues can link distinctly economic, sociological, and psy- for a research focus on children. chological processes. Herein lies the potential theo^ Separate from the question of who should be retical yield from the study of unemployment. studied is the common need in all research on job A corollary of the interdisciplinary stance for loss to place increasing attention on mediating proresearch in this area is the utility of multimethod cesses rather than on outcomes alone. We need to designs. The Hartford project, for example, used know how the costs of unemployment are incurred, survey questionnaires and intensive interviews to what material and social factors moderate their collect data. Interview questions about the costs of effects, and what conditions exacerbate them. Conjob loss elicited a much broader range of responses crete answers to these questions for different popthan did the same set of questions asked in a survey ulations of the jobless have a direct bearing on our format and addressed to the same workers. Al- capacity to develop meaningful programs of supthough survey procedures are economical and port and services to dislocated workers, their fam- • commonly used in policy-related research, it is ilies, and their communities. apparent that sound policy decisions on this quesFinally, there is an important clarification to be tion are riot best served by excessive reliance on made regarding the economic variables we address this, or probably any other, single methodology. in this research. The focus of these comments has Several more focused objectives should also be been the specific problem of unemployment. Some . made priorities for continuing research on un- of the research we have reviewed, however, con1122 • OCTOBER 1982 • AMERICAN PSYCHOLOGIST

cerns a considerably larger range of potential stressors than job loss alone. The research by Brenner, for example, measures broad change in the economic sector, only one component of which is the rate of unemployment. Although some have inferred that joblessness is a significant mechanism through which economic downturns exert their influence on demands for health services, other competing and simultaneously occurring processes may well be relevant (e.g., inflation, increasing orithe-job stresses). It is important not to confuse explaining how economic activity in general versus unemployment in particular affects health and social well-being. Accounting for the process whereby broad-scale economic functioning affects the quality of life is clearly the overarching task. One important aspect of this problem is to grasp fully the seriousness of the impact of unemployment on the worker and the wider population. REFERENCE NOTES 1. Warr, P. Studies of psychological well-being. Paper presented at the British Psychological Society Symposium on Unemployment, London, 1980. 2. Rayman, P., & Bluestone, B. The private and social response to job loss: A metropolitan study. Final report of research sponsored by the Center for Work and Mental Health, National Institute of Mental Health, 1982. 3. Rosen, E. "Hobson's choice": Employment and unemployment among women factory workers in New England. Final report of research prepared for the Employment and Training Administration, U.S. Department of Labor, 1981. REFERENCES Akabas, S., Weiner, H., & Sommer, J. Mental health care in the world of work. New York: Association Press, 1973. Brenner, M. H. Mental illness and the economy. Cambridge, Mass.: Haryard University Press, 1973. Brenner, M. H. Estimating the social costs of national economic policy: Implications for mental and physical health, and criminal violence. Report prepared for the Joint Economic Committee of Congress. Washington, D.C.: U.S. Government Printing Office, 1976.

Brenner, M. H. Influence of the social environment on psychopathology: The historic perspective. In J. E. Barrett (Ed.), Stress and mental disorder. New York: Raven Press, 1979. Catalano, R., & Dooley, D. Does economic change provoke or uncover behavioral disorder? A preliminary test. In L. Ferman & J. Gordus (Eds.), Mental health and the economy. Kalamazoo, Mich.: Upjohn Institute, 1979. Catalano, R., Dooley, D., & Jackson, R. Economic predictors of admissions to mental health facilities in a non-metropolitan community. Journal of Health and Social Behavior, 1981, 22, 284-298. Coates, D., Moyer, S., & Wellman, B. Yorklea study: Symptoms, problems, and life events. Canadian Journal of Public Health, 1969, 60, 471-481. Dohrenwend, B. S., & Dohrenwend, B. P. (Eds.). Stressful life events: Their nature and effects. New York: Wiley, 1974. Dohrenwend, B. S., & Dohrenwend, B. P. (Eds.). Stressful life events and their contexts. New York: Watson, 1981. Eyer, ]: Prosperity as a cause of death. International Journal of Health Services, 1977,, 7, 125-150. Ferman, L., & Gordus, J. (Eds.). Mental health and the economy. Kalamazoo, Mich.: Upjohn Institute, 1979. Jahoda, M., Lazarsfeld, P., & Zeisel, H. Marienthal. Chicago: Aldine Press, 1971. Kantor, R. M. Work and family in the United States: A critical review and agenda for research and policy. New York: Russell Sage, 1977. Kasl, S. Changes in mental health status associated with job loss and retirement. In J. Barrett (Ed.), Stress and mental disorder. New York: Raven Press, 1979. Kasl, S., & Cobb, S. Some mental health consequences of plant closing and job loss. In L. Ferman & J. Gordus (Eds.), Mental health and the economy. Kalamazoo, Mich.: Upjohn Institute, 1979. Kelly, E., & Webb, L. (Eds.). Plant closings. Washington, D.C.: Conference on Alternative State and Local Policy, 1980. Liem, R., & Liem, J. Social support.and stress: Some general issues and their application to the problem of unemployment. In L. Ferman & J. Gordus (Eds.), Mental health and the economy. Kalamazoo, Mich.: Upjohn Institute, 1979. Marshall, J. P., & Punch, D. P. Mental illness and the economy: A critique and partial replication. Journal of Health and Social Behavior, 1979, 20, 282-289. Rayman, P. The world of not working: An evaluation of urban social science response to unemployment. Journal of the Health and Human Resources Administration, 1982,4,319333. Sclar, E. Social cost minimization: A national-policy approach to the problems of distressed economic regions. In S. Redburn & T. Buss (Eds.), Public policies for distressed communities. Lexington, Mass.: Heath, 1982.

AMERICAN PSYCHOLOGIST • OCTOBER 1982 • 1123