Reflections on Marienthal and after.
The most obvious advance lies in the area of methods, particularly in the modern rigour of study design and data analysis. Marienthal was not designed in advance but developed during field work. The investigators wanted to know 'everything' related to the impact of unemployment on the community and its members. While they made some lucky hits, 'everything' was certainly not included. In the analysis of the data, intuitive interpretation played a significant role, statistics were limited to percentages worked out with a slide rule.
An argument could be made that the results of the study were on occasions too easily taken as valid across space, time and methods. If modern research confirms some of the earlier findings, in spite of the enormous social and economic changes that have taken place in the meantime, this had to be established anew. Now there is virtual unanimity in the research community that the vast majority of the unemployed, then as now, experience psychological impairment.
What is new and exciting in current research are the questions which are being asked, the controversies that have arisen, the concepts and search for explanations that guide contemporary work.
A charitable look at Marienthal may reveal that there are hidden in that largely descriptive account some data relevant to some modern questions, but the point is that the questions then were not explicitly posed and did certainly not receive clear answers. Question-asking is, perhaps, the most crucial step in any research project because the formulation of a research question determines the categories of possible answers. New questions advance knowledge; precisely formulated questions yield more precise answers.
The major new questions are about causality in the well-established correlation between unemployment and deficits in positive mental health. This correlation epitomizes a central question of all social science: determination from the inside-out or outside in? In unemployment research two types of causality have been investigated. One is the drift hypothesis, i.e. are there psychological deficits in those who become unemployed or does unemployment produce or strengthen such deficits? While it needs further research with different samples, in different localities, under different economic conditions, using different criteria of mental health, the tentative answer so far is that the impairment found among so many of the unemployed is for some the result of previous personality attributes, for others the result of becoming unemployed.
The other causal question is whether it is the economic hardship in unemployment or being without a job that accounts for the established correlation. Once again, both factors seem to be involved; their relative strength remains to be established. Within the broad framework of searching for causality many other new questions are being investigated, using standardized instruments for better comparability, longitudinal design, controls and computer-aided advanced statistics.
Marienthal remained, largely because of the political events at the time of publication, without criticism. The year was 1933, Hitler had just obtained power and the German publisher cautiously insisted on not putting the names of the Jewish authors on the book-cover at a time of book burnings. Thus the immediate impact of Marienthal was minimal. Its reputation began to grow with its German re-publication in 1960 and subsequent translations, when critique seemed purposeless.
In contrast, there is no shortage of criticism of modern research. Controversies have arisen over some findings, but particularly over some basic questions. In general, controversy at the forefront of a field of study is a sign of health; it sharpens the mind, energizes efforts on all sides and highlights crucial research questions. Of course there can also be destructive controversy leading to arguments ad bominem and assertions of a position, never mind the evidence; unemployment research has fortunately been largely free from this type of controversy. The healthy one, however, has flourished for some time, mostly around the second causality question: poverty or joblessness? Both positions have been glorified with the word 'theory' (but see below) and labelled agency versus deprivation or latent consequences 'theory.' After prolonged discussion it emerges that these are not theories in the strict sense of the term and that they are not diametrically opposed. The two positions differ in the emphasis on various tacit assumptions underlying approaches to social research. Agency 'theory' emphasizes that human beings are active organisms whose activity level is restricted by unemployment. The deprivation 'theory' emphasizes the habitual use people make of social institutions in meeting some psychological need. Put this bluntly there is here obviously no unbridgeable conflict; neither denies the relevance of the other's assumption, only its weight. But lest someone concludes that it was all a storm in a teacup, consider the consequences of either position: The agency approach has led to greater emphasis on the study of poverty in unemployment, making economic hardship a central explicator of the psychological impairment, a matter that had too often been ignored in psychological studies. The deprivation 'theory' leads to more sociological considerations of employment as a dominant social institution: exclusions from it leaves psychological needs unmet. Both positions use concepts that transcend psychological terminology--economics and sociology--and are thus steps toward an interdisciplinarity, often demanded for unemployment research but hardly ever implemented.
It is conceivable that a study could contain both positions, though it has not yet been done, by comparing the mental health of the poor amongst the unemployed with that of those who are poor for other reasons and of the unemployed who are not poor. If their impairment is comparable, both positions would receive support.
And yet--this would not add up to a theory of unemployment in the strict sense of the term as it is used in experimental psychology and other experimental sciences. There theories can be tested, found wanting, improved or validated and predict under the clause 'other things being equal.' But it is that clause which never holds in social research on unemployment and other social issues. In an ever-changing world other things are never equal. This is why theories in the social sciences--many have been proposed--have a peculiar fate: they stimulate, they provide important concepts, but the stipulated relation between explicator and explicandum does not stand the test of time and circumstances. So Marx's theory of the progressive polarization of social classes had been falsified, but his concept of alienation from work remains with powerful relevance, as does Weber's work ethic and Durkheim's anomie. The multitude of modern psychological theories relevant to unemployment research shares the same fate; learned helplessness, exchange-value, cognitive dissonance, attribution of causes and other concepts, self-theories and others provide immensely useful concepts for unemployment research, guiding question-asking and interpretations. The formalized theories, however, stipulating hypothetico-deductive relationships, do not play a constructive role in unemployment research. In every single study 'if A then B' has to be qualified beyond recognition, even if these qualifications are hidden in group averages and unconcern with generalizations. No single theory about the psychological impact of unemployment exists; none is likely to emerge. But there is much good thought and an increasing concern with the utilization of concepts developed in the experimental sciences.
Marienthal was atheoretical in an even more radical fashion, relying on hunches rather than formulated concepts, as a result of its peculiarly open-ended approach, which was both its major asset and its major fault. Because it was problem-, not method-centred, triangulation, quantitative and qualitative approaches were used, economic and psychological data juxtaposed. This is all to the good, and perhaps too often missing in modern research. When a central result is supported by a variety of methods one's confidence in it is increased. But what about instances where triangulation yields different results? Such a case actually occurred in Marienthal, bearing on the inside-out or outside-in issue; since at the time no way of dealing with it was seen, it was silently ignored. This is the case: having established that the response to unemployment was systematically related to the amount of unemployment allowance received, the life histories of the 'unbroken,' and the 'apathetic' were studied. It emerged that the former has always been 'unbroken,' i.e. actively shaping their fate, the latter had passively accepted what came their way. Economic or psychological determinism? The issue was not raised, let alone resolved. Notwithstanding this and other examples of insufficient conceptualizations, Marienthal can serve even now a constructive function by demonstrating an approach that is essential in every study of a relatively unexplored area: descriptive fieldwork. Where this is the major methodological approach it must, of course, be executed by the main investigators and not be delegated to less involved and less skilled assistants, as is the case in most survey research where the main investigator is cut off from the raw data. Direct contact with the people and the situation under study is the best source for discovery and new ideas, the best safeguard against overlooking the unexpected. Given the diversity of conditions under which unemployment is experienced it might be useful if modern, conceptually superior studies paid more attention to the latent facts (not just consequences) in every social situation, not to replace but to supplement their sophisticated techniques. To make visible in its complexity the otherwise invisible is surely one of the tasks of social research.
This is not to say, of course, that fieldwork is always the best method in modern research on unemployment, only the most neglected. Depending on the central research questions other approaches are indicated, yielding more precise results. They could, however, in their presentation, aim to combine some of the assets of fieldwork with their analyses. In some reports on modern studies it is hard to realize that the unemployed are human beings; they are dissolved into variables suitable for entry into regression analyses and ANOVAS. These estimable statistical techniques have their uses, of course. But if simple percentage distributions preceded them, such reintroduction of people into research reports would gain them not only a wider readership; it would also alert the authors to how much more remains to be explained than generalized tendencies.
There is, alas, little hope that unemployment will cease to be a significant blot on the social scene in the foreseeable future, but there is good hope that social scientists will continue to make its economic and human costs subject to their best thought and sophistication, advancing their disciplines as they are forcing attention on human welfare.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Special Issue: Marienthal and Beyond: 20th Century Research on Unemployment and Mental Health; social research on unemployment|
|Publication:||Journal of Occupational and Organizational Psychology|
|Date:||Dec 1, 1992|
|Previous Article:||Towards a social psychology of the labour market: or why we need to understand the labour market before we can understand unemployment.|
|Next Article:||Relationships of work stressors with aggression, withdrawal, theft and substance use: an exploratory study.|