Adolescence is usually defined as a period from puberty until full adult status has been attained. In modern societies this period appears to be prolonged, not only due to a longer period of education, but because of the increased length of time of unemployment afterwards. One of the consequences of unemployment is economic dependence, which makes it difficult for an adolescent to synchronize and to take full responsibility for the social roles expected at the end of adolescence, such as profession, spouse, and parent (Hultsch & Plemons, 1979).
Empirical research on the effects of unemployment undertaken in the 1930s, during and after the Great Depression, was mostly concerned with the material consequences (Feather, 1989). In the 1980s, when unemployment again became an issue, the focus of the studies changed to the psychological impact. The effects revealed by these studies included depression, lowered self-esteem, apathy, and self-doubt (Bachman, O'Malley, & Johnston, 1978; Feather, 1989). For example, Feather and Bond (1983) found that unemployed university graduates reported more depressive symptoms when compared with the employed sample. They also had lower self-esteem and were less organized and less purposeful in their use of time. Some other studies (Finlay-Jones & Eckhard, 1981; Kuzmanovic, 1985) indicated that between 30 to 40% of unemployed adolescents experience severe emotional problems.
The main purpose of the present study was to examine the relationship between the length of unemployment and level of self-esteem by unemployed Croatian graduate students. Most of the studies which have examined this relationship were carried out in countries where no such long-lasting social and economic crises exist as has recently been the case with Croatia. Given the poor economic conditions throughout the country and the large number of unemployed, it is possible that Croatian youths are less inclined to attribute failure to find a job to his or her own efforts and/or other personal qualities than do youth in more prosperous countries. Thus, the length of unemployment might have no effect on their self-esteem.
The second aim was to examine the role of social support in coping with unemployment, especially if it is long-lasting since it can be conceived as a stress-inducing condition. "Stress is viewed as a product of certain life events or conditions that can lead to a variety of consequences - coping efforts and defense strategies, feelings of distress, altered quality of functioning in major life roles, and psychological and physiological symptoms" (Lieberman, 1982, p. 766). Several studies have examined the role that social support from informal networks - family and friends - can play in stress reduction (Lieberman, 1982; Lempers & Clark-Lempers, 1990; Krause, 1990). It is generally assumed that social support can work as a stress buffer and thus modify the effects of stressful events that erode self-esteem and the sense of self-competence. This study examines that assumption.
The sample consisted of 98 unemployed university graduates (67 female and 31 male). Mean age was 27 years and 3 months. All respondents were registered at the Unemployment Registry Office in Zadar, Croatia.
In the first part of the questionnaire used, several items measured demographic variables.
Self-esteem. The subjects completed Rosenberg's Self-esteem Scale (1965), consisting of 10 items (e.g., "I feel I can't do anything right") rated on a 4-point scale. This scale has been extensively employed in self-esteem research, and its psychometric properties for the Croatian sample have been found satisfactory in a previous study (Bezinovic, 1988). Internal consistency, Chronbach's alpha, in the present sample was also high, .72.
Life satisfaction. The general feelings of happiness and satisfaction with one's life was assessed by the Life-Satisfaction Scale (Bezinovic, 1988), which consists of 7 items (e. …