Expectancy-value theory and unemployment effects.
Lewin, K., Dembo, T., Festinger, L. & Sears, P.S. (1944). Level of aspiration. In J. McV. Hunt (Ed.), Personality and the Behavior Disorders, vol. 1, pp. 333-378. New York: Ronald.
Locke, E. & Latham, G. P. (1990). A Theory of Goal Setting and Task Performance. Englewood Cliffs, NJ: Prentice-Hall.
Madden, T. J., Ellen, P. S. & Ajzen, I. (1992). A comparison of the theory of planned behavior and the theory This article describes how expectancy-value theory can be applied to the analysis of two key topics in unemployment research, the impact of unemployment on psychological well-being and job-seeking behaviour. The main concepts used in the expectancy-value theory are summarized and the theoretical framework is then related to the two topics and illustrated with research examples. The approach is seen to be more directly applicable to job-seeking behaviour than to psychological well-being, although it may also clarify the latter topic given additional assumptions. It is concluded that expectancy-value theory is one of a number of possible approaches whose application could provide new ways of examining unemployment effects. It needs to be supplemented by other approaches from industrial/organizational psychology, from general psychology, and from other social sciences.
Research into the psychological impact of unemployment has clearly indicated that unemployment has negative consequences for the psychological well-being of most individuals, that some people cope with the unemployment experience better than others, and that the effects of unemployment are determined by a complex set of variables that can exacerbate or reduce its impact (e.g. Feather, 1990b; Fryer, 1988; Warr, 1987). The recent literature on unemployment effects is now extensive and it is a literature to which psychologists have made major contributions. However, the psychologist does not have exclusive rights to the analysis of unemployment effects. A comprehensive approach to questions about unemployment will involve contributions from different disciplines. Some important questions, such as the effectiveness of political action in producing policy changes and the development and evaluation of social and economic measures designed to improve employment prospects, are more in the domain of the sociologist, the political scientist, and the economist than the psychologist.
What then is the special role of the psychologist in research into the impact of unemployment? The interests and skills of the psychologist can be applied to the sophisticated design of research that provides reliable answers about the effects that unemployment can have on the experience and behaviour of individuals and social groups such as the family. Research of this kind brings to light the influence of different variables such as financial stress, social support, length of unemployment, individual and demographic characteristics, and coping styles. Also important, however, is the development of research that relates to current psychological theories and the possible invention and refinement of new theoretical approaches that can be applied to understanding the effects of unemployment at the psychological level of enquiry.
It would be foolish to expect that one can account for the psychological impact of unemployment in terms of a single, all-embracing theory. There are a number of different theoretical approaches that can be applied, depending on the questions that are asked. Some theories come from the industrial/organizational area. These theories have the virtue of being specifically relevant to the work situation and they have contributed in important ways to the analysis of the effects of employment and unemployment on mental health, psychological well-being, and behaviour. Theories such as Jahoda's latent function analysis (Jahoda, 1982) and Warr's vitamin model (Warr, 1987) are of this type.
A number of other psychological approaches have been developed outside of the industrial/organizational area (Feather, 1990b). These approaches include recent contributions to the analysis of the self-concept (e.g. Higgins, 1987; Markus & Nurius, 1986; Marsh & O'Neill, 1984), stress and coping theory (e.g. Lazarus & Folkman, 1984), learned helplessness theory (e.g. Abramson, Metalsky & Alloy, 1989; Peterson & Seligman, 1984), expectancy-value theory (e.g. Feather, 1982c, 1990a, 1992a,b), attribution theory (e.g. Hewstone, 1989; Weiner, 1986), social cognitive and self-efficacy theory (e.g. Bandura, 1986), and theories about development across the life-span (e.g. Baltes, 1987). The list is probably not complete but it does sample some relevant theoretical approaches that have emerged in recent years.
These more general approaches can be applied to the analysis of particular questions about the effects of unemployment, such as the degree to which an unemployed person feels helpless and lacking in personal control, how much effort an unemployed person expends in looking for a job, how unemployment influences a person's level of self-esteem, how a person might cope with the strain that usually accompanies unemployment, how beliefs about the causes of unemployment may influence affective reactions, and how reactions to unemployment may vary across the life-cycle.
My main intention in this article is to describe the general framework of one of these approaches, expectancy-value theory, and to indicate how this type of theory might be usefully applied in the area of unemployment research.
Within the framework of expectancy-value theory a person's actions are related to the expectations that the person holds and the subjective values (or valences) that are associated with alternative instrumental actions and their possible outcomes (Feather, 1982c, 1990a, 1992b). The subjective values (or valences) may be positive or negative, signifying attractive or aversive events or outcomes. For example, a person may perceive that travelling a long distance to his or her job is aversive but that working at the job itself is attractive. The expectations encompass beliefs about whether a particular action can be performed to some required standard that defines a successful outcome and also beliefs about the various positive and negative consequences that may follow the outcome. An unemployed person, for example, may have a strong expectation that he or she can perform well at a job interview, thereby meeting the main requirements for successful performance (e.g. by addressing the questions asked at the interview and by presenting self in a favourable light). The unemployed person may also hold the expectation that succeeding at the job interview will have positive consequences, the major one being getting the job. This outcome may in turn be linked to other consequences that have positive or negative subjective value (e.g. getting a salary, having to do a lot of shift work).
The analysis of actions in terms of expectations and subjective values has a long history in psychology. This history has been described by Feather (1959a) and by Atkinson (1982) who both refer to contributions that came from the Lewinian analysis of level of aspiration or goal-setting behaviour (Lewin, Dembo, Festinger & Sears, 1944), Tolman's analysis of purposive behaviour and the principles of performance (Tolman, 1938, 1955), the subjectively expected utility (SEU) analysis of decision making (Edwards, 1954), social learning and clinical psychology (Rotter, 1954), and their own research in the analysis of achievement motivation (Atkinson, 1957) and object preference (Feather, 1959b). It is interesting to note that while the Marienthal research on unemployment was underway (Jahoda, Lazarsfeld & Zeisel, 1933), Lewin (1936, 1938) was developing his theoretical analysis of the concepts of force, valence, need, and psychological distance, and Tolman (1932) was elaborating a cognitive theory of learning in which expectations and demand characteristics were central concepts. These early contributions were subsequently elaborated and, together with theoretical inputs from decision theory and other areas, laid the conceptual foundations for a motivational approach of considerable generality.
The expectancy-value approach has been refined and liberalized in various ways over the intervening years (e.g. Brown & Veroff, 1986; Feather, 1982c, 1990a; Heckhausen, Schmalt & Schneider, 1985; Kuhl, 1987; Weiner, 1986). It has also been applied to the work situation in the guise of instrumentality theory (e.g. Mitchell, 1982; Vroom, 1964). Concepts and forms of analysis employed recently in social cognitive theory (Bandura, 1986) and goal theory (Locke & Latham, 1990) resemble similar ideas in earlier statements of expectancy-value theory.
The detailed elaboration of expectancy-value theory includes the analysis of the determinants of expectations and subjective values (or valences) and assumptions about how these variables combine to determine action tendencies (Feather, 1982c, 1986, 1990a, 1992b). The variables are assumed to be linked to the cognitive-affective system. The model of the person is that of an active agent, appraising and construing situations in terms of available alternative courses of action and assessing the likelihood that actions can be performed and that these actions will lead to affectively toned outcomes and consequences.
My analysis assumes that needs and values have the power to induce subjective values (or valences) on objects and events (Feather, 1990a, 1992b). They can be assumed to influence a person's subjective definition of a situation so that objects, activities, and states of affairs within the immediate situation become linked to the affective system and are seen as having positive valence (they become attractive) or negative valence (they become aversive), to use Lewin's (1936) terminology. But the valences also depend upon the perceived characteristics of the goal region itself (e.g. the job) and perhaps on other factors as well. We are still some way from having a comprehensive theory of valences (Feather, 1990a).
The antecedents or determinants of expectations and subjective values (or valences) can be illustrated by reference to a person's attempts to obtain a particular job. Expectations about getting the job would be related to such factors as the person's beliefs about the amount of competition for the job, whether or not the person believes that he or she has the necessary qualifications and skills for the job, whether or not the person believes that he or she can satisfy entry requirements such as an interview, the person's knowledge about the success or otherwise of his or her job applications in the past, where the person believes he or she stands relative to other applicants for the position, and by personality dispositions and more transient moods and affective states (e.g. anxiety) that might shift expectations upwards or downwards.
The subjective value (or valence) of a particular job would depend upon the perceived characteristics of the job itself and the person's needs and values. A job might be seen as attractive, for example, because it provides many of the environmental features that Warr (1987) lists in his vitamin model that are associated with good employment. That is, it may provide a secure income, opportunities for skill utilization, the exercise of control, variety, clarity, a defined goal structure, contacts with other people, and so on. The specific features of the job may in turn fit personal needs and values that are important for the individual and dominant at the time. For example, the job may be seen as attractive because it meets a person's strong need to achieve or has the potential to satisfy the high value a person places on helping others. Other jobs may be perceived as aversive because they have negative aspects such as a low salary, do not provide opportunities for the exercise of skill and control, lack variety and clarity, do not provide goals or regular social contacts, and so on. They may also fail to meet the important needs and values of the individual.
Expectations and subjective values (or valences) are assumed to combine to determine a person's motivation to act in a particular direction. A person's motivation to try for a particular job, for example, would depend upon that person's expectation about how likely it is that he or she will be appointed as a result of an interview or other activity, and on the degree to which the person sees actions leading to the job and the job itself as attractive or aversive. In any given situation there may be a number of competing motivational or action tendencies that relate to alternative goals and activities. For example, an unemployed person may have the choice of watching an attractive TV programme at home or travelling to a job interview. The behaviour that occurs would depend upon which action tendency is dominant at the time. Strong expectations or strong valences alone do not necessarily lead to a strong motivational tendency to act in a particular direction. For example, a person may perceive that being a physician would be a very attractive occupation but may not actively pursue that alternative because of low expectations. Instead, the person may seek a job that is less attractive but which is within the range of his or her realistic expectations. It is the combination of expectations and valences that is important.
It is possible to draw links between expectancy-value theory and psychological well-being if one extends the approach to take account of causal attributions and other subjective variables. The subjective values or valences attached to instrumental actions and to goals are linked to the affective system. Indeed, valences may be conceptualized as involving anticipated positive or negative affects associated with actions and their possible outcomes (i.e. anticipated satisfaction or dissatisfaction; Feather, 1986, 1990a). Hence one would expect that the occurrence or non-occurrence of positive or negative events and outcomes would have affective consequences for the individual (see Feather, 1982a, p. 70). For example, an unemployed person who obtains a positively valued job would feel happy; failure to obtain the job would be associated with sadness and frustration. These are obvious and predictable outcome-dependent affects that probably occur without much conscious thought or reflection.
There are other more specific feelings and emotions that may follow outcomes when these outcomes are unexpected, negative, or important. Weiner (1986) argues that the specific emotions that are experienced will depend upon particular kinds of causal attributions. His model would imply that a person who has lost a job would feel unhappy and frustrated and would undertake a causal search to determine why the outcome occurred, especially if the job was important for the person and the loss of it was unexpected. The job loss may be attributed to a particular cause. Thus, the person may ascribe it to a personal deficiency because he or she has a consistent record of losing jobs in the past while most others have managed to stay in secure employment. Another person may attribute the loss of the job to poor economic conditions because it is common knowledge that many workers have also lost their jobs in the prevailing economic climate despite the fact that they are competent workers.
Weiner (1986) assumes that the causal attributions that the person makes have consequences depending upon how the causes are located within a dimensional space that involves the dimensions of locus (internal, external), stability (stable, unstable), and controllability (controllable, uncontrollable). He argues that this location will influence both expectation change and affective reactions. For example, his approach would imply that attribution of the job loss to stable causes (e.g. lack of ability or a chronic economic recession) may determine reduced expectancies of finding a job in the future whereas attribution to unstable causes (e.g. bad luck or a change in the economy that is deemed to be temporary) may have only minor effects on a person's expectancies; an attributed cause of unemployment that is seen to be under the control of others (e.g. the economic mismanagement of government) may elicit anger; a causal attribution for unemployment that is seen by the person as self-related and uncontrollable (e.g. job loss because of lack of ability or some physical defect) may elicit shame and embarrassment; a cause that is seen by the person as involving an aspect of self that is controllable (e.g. job loss because of dishonest behaviour or lack of effort) may elicit guilt; feelings of hopelessness may occur when expectations are very low; diminished self-esteem may occur when the job loss is attributed to internal causes. It is apparent that the affective consequences may be complex depending upon the dimensional location of the attributed causes.
Weiner's (1986) attributional approach can be seen as a variant of expectancy-value theory because he assumes that expectancy and affect together determine action. For example, the model would predict that an unemployed person who had a low expectation of finding a job and who experienced negative affect such as depression would be less likely to look for a job than a person who was more hopeful and whose self-esteem was higher. However, Weiner (1986) does not include the subjective values or valences of potential actions and outcomes as variables in his approach. His emphasis is on how individuals attribute causality for positive or negative outcomes, on how these causal attributions determine changes in expectancies, and on the relations between causal attributions and positive and negative emotions. Thus, his analysis reflects the influence of Heider (1958) as well as his background in expectancy-value theory.
Emotional reactions, however, may be related not only to a complex attributional base but also to a motivational base involving needs and values (Feather, 1992a). The linkages between needs, values, and affect have often been ignored by attributional theorists in their emphasis on causal explanations. However, it is commonly recognized that the fulfilment or frustration of important needs and values may generate positive or negative affect respectively. Jahoda's (1982) discussion of the manifest and latent functions of employment and their relation to basic needs and psychological well-being involves this assumption. In the present context the assumption implies that the negative affect experienced by a person who fails to get a job will depend not only on how the person attributes causality for the event but also on the degree to which the person needs a job and values employment (Feather, 1990b; Feather & Davenport, 1981). An unemployed person may report feeling unhappy and dissatisfied with his or her lot even though it is reasonable to blame the unemployment on external conditions such as an economic recession. The negative affect relates to the failure to attain a positively valued goal, employment, and the frustration of underlying needs and values that make this goal attractive.
How others react to an unemployed person also depends on how they attribute causality for the unemployment and on their own needs and values. The recent attributional analysis of reactions to stigmas by Weiner, Perry & Magnusson (1988) emphasizes the degree to which the onset of the stigma is seen to be controllable or uncontrollable. In the present context this analysis implies that others may feel more sorry for an unemployed person when the unemployment is perceived as an unintended negative outcome beyond the person's control than when the unemployed person is seen as responsible for his or her condition because of lack of effort or some other controllable cause. In line with my previous argument, how people react to unemployment would also be influenced by their own needs and values (Feather, 1985). For example, those who support the work ethic would feel more sorry when they know that an unemployed person has tried hard to get a job than when they know that the person has not exerted much effort in the job search. It is likely that they would also consider the former person to deserve his or her negative condition (unemployment) less, whereas they might consider that some justice has been done in regard to the latter person. This expected difference follows from a recent conceptual analysis that relates beliefs about deservingness to the degree to which a person is seen to be responsible for a positively or negatively valued outcome and to whether the outcome is seen to follow positively or negatively valued behaviour (Feather, 1992a). The unemployed person's judgement of his or her own deservingness may also be associated with affective reactions (e.g. guilt) that are relevant to psychological well-being.
It is important to recognize that psychological well-being is more than affect experienced at any given moment of time. Psychological well-being is usually discussed as extending beyond the moment and having some continuity over time. For example, an unemployed person continues to feel depressed or angry or lacking in self-esteem following consistent failure to find a job. In terms of the analysis presented so far, the continuity of affective reactions could reflect both the persistence of stable patterns of causal attributions for consistently experienced positive or negative outcomes and the continuing fulfilment or frustration of important needs and values. For example, the unemployed person may believe that the causes of his or her chronic unemployment remain the same and he or she may continue to see employment as important for the fulfilment of self-related needs and values despite frustrated attempts to find a job. Hence, the emotions that are experienced retain their particular quality and stability because major determining conditions are themselves relatively stable.
The continuity of psychological well-being is also considered in theories that derive from learned helplessness theory (e.g. Abramson, Seligman & Teasdale, 1978). These approaches enable one to account for the persistence and generality of negative affective states and other depressive symptomatology in terms of a set of contributory causes that include stable and global causal attributions for negative life-events. One of the most recent statements assigns hopelessness a pivotal role as a proximal and sufficient cause of hopelessness depression (Abramson et al., 1989). Hopelessness is conceptualized as involving both negative expectations about the occurrence of highly valued outcomes and expectations of helplessness about changing the likelihood of occurrence of these outcomes. The variable that Feather & O'Brien (1987) called control-optimism may be interpreted as a measure of hopefulness relating to employment prospects, at the other pole from hopelessness but less global in its application when compared with the variable discussed by Abramson et al. (1989). The latter authors also refer in their model to the importance of the negative life-event, a variable that could be identified with the degree to which the life-event is perceived to be attractive or aversive (i.e. has positive or negative subjective value or valence for the individual). But they present importance as one of the contributing causes that are assumed to influence hopelessness, rather than as a separate direct influence on hopelessness depression.
There are other recent analyses of psychological well-being that relate to expectations and the subjective values of outcomes (Feather, 1990b). For example, Bandura (1982) discusses the interactive effects of self-percepts of efficacy and response-outcome expectations on behaviour and affective reactions. In his analysis, resignation and apathy are assumed to occur when self-efficacy is low and the environment is perceived to be unresponsive. But when self-efficacy is low and the environment is seen to reward responses, self-devaluation and despondency may follow. His analysis implies that psychological reactions to unemployment will vary depending upon the particular combination of outcome and efficacy beliefs. If unemployed individuals have high self-efficacy, believing that they have the necessary abilities and skills to qualify for a job, and if they also believe that getting a job will be followed by positive consequences, then they should actively continue to seek employment. But if these individuals have little confidence in their capabilities and believe that it is futile for them to seek employment, then they will become resigned and apathetic (if they believe their environment is unrewarding) or suffer low self-esteem and despondency (if they believe that their environment does reward appropriate behaviours).
The studies in the Flinders programme of research on unemployment effects have included depressive affect about unemployment and more generalized depression as dependent variables, as well as measures of apathy, self-esteem, life satisfaction, stress symptoms, and other affective variables (e.g. Feather & Barber, 1983; Feather & Davenport, 1981; Feather & O'Brien, 1986a,b). Some of these studies have been explicitly set within the expectancy-value framework (e.g. Feather & Davenport, 1981) and causal attributions for unemployment have been investigated in other studies (e.g. Feather & O'Brien, 1986a,b). It is clear, however, that the analysis of psychological well-being can be only partially understood in terms of the general expectancy-value approach and that other variables need to be introduced to account for the nature of the unemployment experience. Some of these variables will relate to the way a person construes or appraises the situation of unemployment in terms of social comparison and perceived causality, implicating either the self or external agents. Other variables will relate to factors such as social support and financial support that would be expected to moderate or buffer the effects of unemployment. Other variables will concern the availability of substitute activities that can reduce the impact of not having a job.
At a much more general level of analysis, the evidence from research on structure and purpose in the use of time can be interpreted as indicating that the psychological well-being of individuals depends on the degree to which they see themselves as having a goal structure in their daily lives. Our studies have consistently shown that psychological well-being is positively related to time structure and that unemployed people tend to report that their use of time is less structured or purposive than employed people (Bond & Feather, 1988; Feather, 1990b; Feather & Bond, 1983; Rowley & Feather, 1987). One would expect to find less structure and purpose in the use of time under conditions where a person is unable to fulfil needs and values that relate to important goals such as getting a job, where alternative goals cannot substitute for lack of a job, and where the unemployed person's environment is restricted in the positive goals and opportunities that are available. The conjunction of low expectations and very attractive but unrealizable goals (such as getting a job) and the lack of acceptable and available substitutes may ultimately lead to perceptions that daily life is without structure and purpose and to a diminished sense of psychological well-being.
Finally, in the present section I have focused on possible links between psychological well-being and expectancy-value concepts, supplementing the analysis with other variables where called for. It is important to remember that there are many other theoretical approaches that have been applied to the analysis of psychological well-being (Feather, 1990b). These include contributions from the industrial/organizational area as well as models from general psychology concerned with stress and coping and with changes in the self-concept. One should also note that the concept of psychological well-being is multidimensional and extends beyond affective well-being to include other aspects of mental health (e.g. Warr, 1987, 1990). A complete theoretical analysis of psychological well-being will have to recognize the multidimensional nature of the concept.
The expectancy-value approach can be readily applied to the analysis of job-seeking behaviour among the unemployed. Feather & O'Brien (1987) investigated job-seeking behaviour in two samples of young unemployed people. One was a sample of 131 respondents (52 males, 77 females, two who did not specify gender) who were unemployed in 1981; the second was a sample of 320 respondents (140 males, 179 females, one who did not specify gender) who were unemployed in 1982. These respondents had been enrolled in year 10, year 11, and year 12 classes in 15 state high schools in metropolitan Adelaide. They had left school and had not gone on to further study. The mean age of the respondents was 17.48 years for the first sample and 17.64 years for the second sample. All of the respondents reported that they were full-time unemployed.
Feather & O'Brien (1987) obtained measures of control-optimism, job valence, social support from parents, duration of unemployment (number of weeks spent looking for a job), the number of job applications, work ethic values, and the frequency of job-seeking behaviour defined by each respondent's answer to the question, 'How frequently do you look for a job?' Respondents answered this question by checking one of six categories that ranged from 'Not looking for a job at all' (1) to 'Daily' (6).
The measures of control-optimism and job valence were based on the results of a factor analysis that justified combining items so as to form scales that measured each variable (Feather & O'Brien, 1987). The control-optimism variable emerged as a factor from this analysis and it involved the following five items with high loadings on the factor: 'How helpless do you feel about your unemployment?' (completely in control/very helpless); 'How confident are you about finding the job you really want in the near future?' (not at all confident/very confident); 'How confident are you about finding any kind of job at all in the near future?' (not at all confident/very confident); 'In the future, if your unemployment continues, will the cause (of your unemployment) still be present?' (will never again be present/will always be present); and 'Can you do anything to change the cause of your unemployment?' (I can easily change the cause of my unemployment/totally beyond my power to change the cause of my unemployment). Respondents answered these questions by using seven-point rating scales with end-labels as shown for each question. The responses were combined to form a control-optimism score for each respondent, after reverse coding the items concerned with helplessness, stability of the perceived cause of unemployment, and personal uncontrollability of the perceived cause of unemployment. Hence subjects with higher scores on the control-optimism variable were those who were less likely to report feeling helpless about being unemployed, more likely to feel confident about finding the job they wanted in the near future or any other type of job, and less likely to see the cause of their unemployment as stable and unchangeable. Thus the control-optimism measure linked confidence about finding a job with beliefs about how much control the unemployed person had over outcomes.
Job valence also emerged as a separate factor from the factor analysis (Feather & O'Brien, 1987). It involved the following three items with high loadings on the factor: 'How much do you feel that you need a job?' (don't really need one/desperate to have a job); 'When you think about being unemployed, or the possibility of being unemployed, how does it make you feel?' (really glad/really depressed); and 'How depressed do you feel about your unemployment?' (not at all depressed/very depressed). Respondents answered these questions using seven-point rating scales with end-labels as shown for each question. The job valence measure for each respondent was obtained by summing his or her scores for the three items. Thus, those respondents with higher scores on the job valence variable were more likely to report that they needed a job and that they felt disappointed and depressed about not having one. This measure of valence is consistent with Lewin's (1938, pp. 106-107) usage of the term in its emphasis on psychological needs and tension systems that influence the perceived attractiveness or aversiveness of regions of the psychological environment.
Work ethic values were measured by using the 19-item scale developed by Mirels & Garrett (1971). Items in this scale were modelled on Weber's (1976) classic analysis of the work ethic and they emphasized the virtues of industriousness, asceticism, and individualism (Feather, 1984). Values may be conceived as general beliefs held by individuals about desirable or undesirable goals and ways of behaving (Feather, 1990a, 1992b). They are assumed to transcend more specific attitudes towards objects and situations but to influence the form these attitudes take. They provide standards or criteria that people use to evaluate actions and outcomes, to justify opinions and conduct, to plan and guide behaviour, to decide between different alternatives, to engage in social influence, and to present self to others. As indicated previously, values are assumed to be linked to the affective system and, along with other variables, they are assumed to influence a person's construal of a situation. Objects and events within the perceived environment may then be seen as attractive (positively valent) or aversive (negatively valent) depending on which values are activated (Feather, 1990a, 1992b).
Feather & O'Brien (1987) hypothesized that the frequency of job-seeking behaviour would be positively related to both a person's expectation of finding employment and to the extent to which employment was seen as attractive or positively valent. They also predicted that the measures of expectation and valence in combination would provide better prediction of job-seeking behaviour than either measure alone. Their results provided partial support for these hypotheses. There were statistically significant positive correlations between the frequency of job-seeking behaviour and job valence in the two samples but the correlations between job-seeking behaviour and control-optimism were non-significant.
The results also showed that job valence was positively related to work ethic values and that control-optimism was negatively related both to the length of time subjects had been unemployed and to the number of unsuccessful job applications. These results are consistent with the assumption that a person's values influence the specific valences attached to objects and events and with the assumption that a major determinant of expectations is a person's own past record of success and failure. Repeated unsuccessful attempts to obtain a job would have the effect of reducing a person's expectation of success.
A further study that involved a sample of 42 unemployed older men in metropolitan Adelaide provided similar results (Feather, 1990b, pp. 234-237). In this case four different measures of job-seeking behaviour were employed and the correlations were obtained between these variables and variables that were similar to the control-optimism and job valence measures just described. Again, the results showed that job-seeking behaviour was positively related to job valence but not significantly related to control-optimism. As in the Feather & O'Brien (1987) study, lower levels of control-optimism were associated with longer periods of unemployment duration.
The results of both studies also showed statistically significant positive correlations between job-seeking behaviour and negative affective reactions. Those subjects whose feelings about being unemployed were most negative were also those who were most likely to be looking for a job (Feather, 1990b, p. 236; Feather & O'Brien, 1987). It would be wrong to conclude, however, that negative affect caused or influenced job-seeking behaviour. While that may be the case, it is also possible that the negative affect was an outcome of repeated unsuccessful attempts to get a job and became more intense for higher levels of job-search activity. The results of the Feather (1990b) study also showed that those unemployed men who were satisfied with their present life were less likely to look for jobs than those who were dissatisfied with their present life.
It is evident that these results were only partially consistent with expectancy-value theory. The measure of valence predicted job-seeking behaviour but the measure of control-optimism did not. In some ways this is a surprising result because, in one of the earliest studies in the Flinders programme, Feather & Davenport (1981) found that measures of confidence about finding a job, need for a job, and disappointment or depressive affect about being unemployed predicted the effort that subjects reported they expended in looking for a job, although effort was more strongly related to the valence variables. Feather & Davenport (1981) also found that their measures of job need and depressive affect, components of the job valence index used in the Feather & O'Brien (1987) and Feather (1990b) studies, were positively correlated with a measure of employment value. Thus, again there was evidence to support the assumption that values can function like needs to induce valences.
The failure to find a positive relation between the control-optimism variable and job-seeking behaviour in the Feather & O'Brien (1987) and Feather (1990b) studies is also surprising given the fact that studies in other social contexts (e.g. joining a social movement organization, selecting an academic major) have shown that expectations play a part in predicting behavioural intentions (Feather, 1988, 1990a; Feather & Newton, 1982). There is also plenty of evidence from laboratory studies to show that expectations predict actions such as choice, performance, and persistence (e.g. Bandura, 1986; Feather, 1982a, 1989).
Why was this negative result obtained? There are a number of possible explanations. Predictions about behaviour become more difficult when one moves outside of well-controlled laboratory settings to real-life contexts. Other variables are then introduced that may have effects and that therefore have to be considered in any theoretical analysis. In particular, it is important to recognize that the forces that motivate an unemployed person to continue to look for a job involve not only the person's own motivations but external pressures as well. For example, a young person with low expectations of success may continue to seek employment because of parental pressure or because welfare benefits will only be paid if there is evidence of current job-seeking activity. These examples remind us that actions occur not only in relation to individual motivations but also as a response to situational constraints, social norms, task requirements, and other imposed conditions. Lewin (1951) recognized this point in his concept of power fields and induced forces that, together with driving forces (linked to valences) and restraining forces (linked to barriers), were assumed to influence the direction of action. Fishbein & Ajzen (1975) considered external pressures in relation to normative variables and motivation to comply in their theory of reasoned action. They assumed that behavioural intentions were influenced by these variables as well as by the attitude held by the individual towards the specific behaviour.
It is also possible that the measures that we used in our studies were pitched at too general a level. It may be preferable in future to focus on more specific measures of valences and expectations that take account of particular job characteristics (e.g. the salary associated with a job; opportunities for control, variety, skill-utilization, and social contacts) and the specific activities that are involved in looking for a job (e.g. making a satisfactory job application or meeting other entry requirements). It would also be an advantage in future studies to go beyond the use of self-report measures that may be associated with response biases (e.g. pressures on respondents to be consistent in their answers) and to obtain behavioural data that involve less reliance on self-report.
Future applications of the expectancy-value framework to the analysis of job-seeking behaviour might therefore extend the approach used by Feather & O'Brien (1987) by using more specific measures of expectations and valences and by including variables that relate to normative requirements and social pressures. Some other investigations of job-seeking behaviour based on the theory of reasoned action (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975) have taken this step (e.g. Caplan, Vinokur, Price & van Ryn, 1989; Vinokur & Caplan, 1987; Ullah & Banks, 1985).
A further variable that might be added is the degree to which an individual believes that the outcomes that follow behaviour can be related to internal controllable causes as opposed to the operation of external forces, fate, or luck. It will be evident from the previous discussion of Weiner's (1986) attributional approach and learned helplessness theory (Abramson et al., 1978, 1989) that perceived controllability has emerged in various guises as an important variable in current motivational theory (see also Feather, 1982a, pp. 78-80; O'Brien, 1984). Another move in this direction is the theory of planned behaviour (Ajzen, 1985; Ajzen & Madden, 1986) that expands the theory of reasoned action by including a perceived behavioural control component relating to beliefs regarding the possession of requisite resources and opportunities for performing a given behaviour. Some recent evidence suggests that the inclusion of perceived behavioural control enhances the prediction of behavioural intention and behaviour beyond that achieved by the theory of reasoned action (Madden, Ellen & Ajzen, 1992). Future studies of job-seeking behaviour might usefully include this variable though, like control-optimism which it resembles, it may turn out to be too general in its definition, mixing variables that refer to personal resources with variables that concern environmental obstacles and opportunities. These variables would be expected to influence expectations of success as well as perceptions of control. The question of how to conceptualize controllability, whether of causes or outcomes, continues to be an important theoretical issue.
Further refinement in the application of expectancy-value theory would also involve detailed specification of the current goal structure of the unemployed person. A basic idea in this approach is that individuals are involved in environments that provide alternative goals and courses of action. In Lewin's (1936) terminology, the psychological environment contains regions of positive and negative valence that are coordinated to force fields, and it also involves a structured set of regions that define the activities that are perceived to be necessary if a person is to move towards an attractive outcome or away from an aversive outcome. For example, the alternatives facing an unemployed person may involve travelling a long distance for a job interview, staying at home and watching a football match on TV, having a drink with friends, or displaying enough job-search activity to qualify for the dole cheque. These alternatives are associated with action tendencies that, in traditional expectancy-value theory, depend upon the strength of expectations and valences and that change as expectations and valences change. Action tendencies can also be assumed to build up and to dissipate over time according to assumptions developed by Atkinson & Birch (1970) in their dynamics of action model.
What is crucial in future research, however, is the need to specify the basic set of action tendencies instigated at any one time and the dominant action tendency that can be assumed to issue into behaviour. Future studies of job-seeking behaviour therefore require careful specification of the alternative activities and goals that exist for unemployed people in their daily lives. If there are many positive alternatives to having a job then one would expect to find less evidence of job-seeking than if the alternatives are generally negative. The finding noted previously that unemployed men who were more satisfied with their lives were less likely to look for jobs than those who were dissatisfied is consistent with this argument.
It may be argued that the expectancy-value approach is overly intellectual and rational in its approach, but that would be an oversimplification of the position. As noted previously, the approach does allow for the role of affect. Needs, values, and valences are assumed to be linked to the affective system. Moreover, it is recognized in the approach that the cognitive expectations that individuals hold may be influenced by motivational and affective variables (Feather, 1982c, 1990a). This assumption is consistent with a lot of evidence that indicates that social judgements can be influenced by affective states (e.g. Forgas, in press). For example, one would expect that an unemployed person's expectation of getting a job would be affected by strong needs and also by feelings of helplessness, disappointment, and other negative emotional states, especially if the expectation is not well articulated or not firmly based in relation to situational cues or past experience.
Possible interrelations between expectations and valences have been discussed in another context, as well as the conditions that affect the relative weights assigned to expectations and valences in determining motivational or action tendencies (Feather, 1990a). It is possible that expectations and valences may have different weights, depending upon individuals and situations (see also Kuhl, 1986). For example, where a situation is structured at a wishful level, individuals may attend more to the subjective values or valences of possible events and either discount the subjective probabilities or construe them in unrealistic ways in line with their wishes. These effects may occur in the case of some unemployed individuals whose motivation to obtain employment is so strong that they deny reality and continue seeking their desired job against all odds.
It is also important to recognize that the expectancy-value approach has been used to account for instrumental behaviours that can be understood in terms of means-end structures that involve beliefs about possible outcomes and subjective values or valences that are linked to the cognitive-affective system. However, many behaviours occur without much thought about expected consequences. There are behaviours such as habitual actions and the routine use of heuristics that involve a minimum use of conscious reflection and that fall outside the boundary conditions of the expectancy-value approach (see also Feather, 1982b, pp. 397-400).
Finally, it may also be necessary in the future to take account of the renewed interest in volition that has come from West German research (Heckhausen, 1986; Kuhl, 1985, 1986, 1987). For example, Kuhl (1985, 1986, 1987) distinguishes motivation from volition and asserts that volition involves a degree of commitment to a planned course of action that is also protected and maintained once initiated by processes of action control. Kuhl relates volition and action control to an information-processing model that involves memory systems and motivational maintenance systems. This theoretical development enables one to deal with the effects of commitment and the more or less singleminded pursuit of a goal. For example, an unemployed person's commitment to seeking a particular kind of job (e.g. becoming an engineer or an actor) may lead to processes that maintain and protect the action tendency to pursue that goal even when there is a strong urge to take another sort of job. The particular form that job-seeking takes may therefore also relate to processes of volition and commitment.
This discussion of expectancy-value theory and its application to the analysis of unemployment effects indicates that it has been usefully applied to the study of job-seeking behaviour. In the case of the Feather & O'Brien (1987) study, this application was not completely successful because the job valence measure predicted job-seeking behaviour but the measure of control-optimism did not. Possible reasons for this negative result were discussed that involved consideration of both methodological and theoretical issues.
The possibility of applying expectancy-value theory to the analysis of psychological well-being was also considered. It must be acknowledged, however, that this approach was developed as a framework for analysing motivated behaviour rather than as a means of understanding psychological adjustment. Hence, it is more appropriately applied to the goal-directed behaviour of the unemployed than to the analysis of their subjective well-being.
The analysis of the psychological effects of unemployment and the variables that contribute to these effects will require theoretical inputs from different areas of psychology and from other social sciences as well. This article has considered one theoretical approach, expectancy-value theory, but this approach will have to be supplemented by other theoretical frameworks if we are to achieve a more general understanding of what it means to be unemployed.
Abramson, L. Y., Metalsky, G. I. & Alloy, L. B. (1989). Hopelessness depression: A theory-based subtype of depression. Psychological Review, 196, 358-372.
Abramson, L. Y., Seligman, M. E. P. & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49-74.
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckman (Eds), Action Control: From Cognitions to Behavior, pp. 11-39. New York: Springer Verlag.
Ajzen, I. & Fishbein, M. (1980). Understanding Attitudes and Predicting Social Behavior. Englewood Cliffs, NJ: Prentice-Hall.
Ajzen, I. & Madden, T. J. (1986). Prediction of goal-directed behavior: Attitudes, intentions, and perceived behavioral control. Journal of Experimental Social Psychology, 22, 426-435.
Atkinson, J. W. (1957). Motivational determinants of risk-taking behavior. Psychological Review, 643, 356-372.
Atkinson, J. W. (1982). Old and new conceptions of how expected consequences influence actions. In N. T. Feather (Ed.), Expectations and Actions: Expectancy-Value Models in Psychology, pp. 17-52. Hillsdale, NJ: Erlbaum.
Atkinson, J. W. & Birch, D. (1970). The Dynamics of Action. New York: Wiley.
Baltes, P. B. (1987). Theoretical propositions of life-span developmental psychology. Developmental Psychology, 23, 611-626.
Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147.
Bandura, A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall.
Bond, M. J. & Feather, N. T. (1988). Some correlates of structure and purpose in the use of time. Journal of Personality and Social Psychology, 55, 321-329.
Brown, D. R. & Veroff, J. (Eds) (1986). Frontiers of Motivational Psychology: Essays in Honor of John W. Atkinson. New York: Springer-Verlag.
Caplan, R. D., Vinokur, A. D., Price, R. H. & van Ryn, M. (1989). Job-seeking, reemployment, and mental health: A randomized field experiment in coping with job loss. Journal of Applied Psychology, 474, 759-769.
Edwards, W. (1954). The theory of decision making. Psychological Bulletin, 51, 380-417.
Feather, N. T. (1959a). Subjective probability and decision under uncertainty. Psychological Review, 66, 150-164.
Feather, N. T. (1959b). Success probability and choice behavior. Journal of Experimental Psychology, 58, 257-266.
Feather, N. T. (1982a). Actions in relation to expected consequences: An overview of a research program. In N. T. Feather (Ed.), Expectations and Actions: Expectancy-Value Models in Psychology, pp. 53-95. Hillsdale, NJ: Erlbaum.
Feather, N.T. (1982b). Expectancy-value approaches: Present status and future directions. In N. T. Feather (Ed.), Expectations and Actions: Expectancy-Value Models in Psychology, pp. 395-420. Hillsdale, NJ: Erlbaum.
Feather, N. T. (Ed.) (1982c). Expectations and Actions: Expectancy-Value Models in Psychology. Hillsdale, NJ: Erlbaum.
Feather, N. T. (1984). Protestant ethic, conservatism, and values. Journal of Personality and Social Psychology, 46, 1132-1141.
Feather, N. T. (1985). Attitudes, values, and attributions: Explanations of unemployment. Journal of Personality and Social Psychology, 48, 876-889.
Feather, N. T. (1986). Human values, valences, expectations, and affect: Theoretical issues emerging from recent applications of the expectancy-value model. In D. R. Brown & J. Veroff (Eds), Frontiers of Motivational Psychology: Essays in Honor of John W. Atkinson, pp. 146-172. New York: Springer-Verlag.
Feather, N. T. (1988). Values, valences, and course enrolment: Testing the role of values within an expectancy-valence framework. Journal of Educational Psychology, 80, 381-391.
Feather, N. T. (1989). Trying and giving up: Persistence and lack of persistence in failure situations. In R. C. Curtis (Ed.), Self-defeating Behaviors: Experimental Research. Clinical Impressions and Practical Implications, pp. 67-95. New York: Plenum Press.
Feather, N. T. (1990a). Bridging the gap between values and actions: Recent applications of the expectancy-value model. In E. T. Higgins & R. M. Sorrentino (Eds), Handbook of Motivation and Cognition: Foundations of Social Behavior, vol. 2, pp. 151-192. New York: Guilford Press.
Feather, N. T. (1990b). The Psychological Impact of Unemployment. New York: Springer-Verlag.
Feather, N. T. (1992a). An attributional and value analysis of deservingness in success and failure situations. British Journal of Social Psychology, 31, 125-145.
Feather, N. T. (1992b). Values, valences, expectations, and actions. Journal of Social Issues, 48, 109-124.
Feather, N. T. & Barber, J. G. (1983). Depressive reactions and unemployment. Journal of Abnormal Psychology, 92, 185-195.
Feather, N. T. & Bond, M. J. (1983). Time structure and purposeful activity among employed and unemployed university graduates. Journal of Occupational Psychology, 56, 241-264.
Feather, N. T. & Davenport, P. R. (1981). Unemployment and depressive affect: A motivational and attributional analysis. Journal of Personality and Social Psychology, 41, 422-436.
Feather, N. T. & Newton, J. W. (1982). Values, expectations, and the prediction of social action: An expectancy-valence analysis. Motivation and Emotion, 6, 217-244.
Feather, N. T. & O'Brien, G. W. (1986a). A longitudinal analysis of the effects of different patterns of employment and unemployment on school leavers. British Journal of Psychology, 77, 459-479.
Feather, N. T. & O'Brien, G. E. (1986b). A longitudinal study of the effects of employment and unemployment on school-leavers. Journal of Occupational Psychology, 59, 121-144.
Feather, N. T. & O'Brien, G. W. (1987). Looking for employment: An expectancy-valence analysis of job-seeking behaviour among young people. Journal of Occupational Psychology, 78, 251-272.
Fishbein, M. & Ajzen, I. (1975). Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley.
Forgas, J. (in press). Affect in social judgments and decisions: A multi-process model. In M. Zanna (Ed.), Advances in Experimental Social Psychology. New York: Academic Press.
Fryer, D. (1988). The experience of unemployment in social context. In S. Fisher & J. Reason (Eds.), Handbook of Life Stress, Cognition and Health, pp. 211-238. Chichester: Wiley.
Heckhausen, H. (1986). Why some time out might benefit achievement motivation research. In H. H. L. van den Berken, Th. C. M. Bergen & E. E. J. De Bruyn (Eds), Achievement and Task Motivation, pp. 7-39. Lisse: Swets & Zeitlinger.
Heckhausen, H., Schmalt, H. D. & Schneider, K. (1985). Achievement Motivation in Perspective. New York: Academic Press.
Heider, F. (1958). The Psychology of Interpersonal Relations. New York: Wiley.
Hewstone, M. (1989). Causal Attribution: From Cognitive Processes to Collective Beliefs. Oxford: Basil Blackwell.
Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological Review, 94, 319-340.
Jahoda, M. (1982). Employment and Unemployment: A Social-Psychological Analysis. Cambridge: Cambridge University Press.
Jahoda, M., Lazarsfeld, P. F. & Zeisel, H. (1933). Marienthal: The Sociography of an Unemployed Community (English translation, 1971). Chicago: Aldine.
Kuhl, J. (1985). Volitional mediators of cognition-behavior consistency: Self-regulatory processes and action versus state orientation. In J. Kuhl & J. Beckmann (Eds), Action Control: From Cognition to Behavior, pp. 101-128. New York: Springer-Verlag.
Kuhl, J. (1986). Motivation and information processing: A new look at decision making, dynamic change, and action. In R.M. Sorrentino & E. T. Higgins (Eds), Handbook of Motivation and Cognition: Foundation of Social Behavior, vol. 1, pp. 404-434. New York: Guilford Press.
Kuhl, J. (1987). Action control: The maintenance of motivational states. In F. Halisch & J. Kuhl (Eds), Motivation, Intention and Volition, pp. 279-291. New York: Springer-Verlag.
Lazarus, R. S. & Folkman, S. (1984). Stress, Appraisal, and Coping. New York: Springer-Verlag.
Lewin, K. (1936). Principles of Topological Psychology. New York: McGraw Hill.
Lewin, K. (1938). The conceptual representation and the measurement of psychological forces. Contributions to Psychological Theory, 1. New York: Johnson Reprint Corporation.
Lewin, K. (1951). Field Theory in Social Science. New York:
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Special Issue: Marienthal and Beyond: 20th Century Research on Unemployment and Mental Health|
|Author:||Feather, N. T.|
|Publication:||Journal of Occupational and Organizational Psychology|
|Date:||Dec 1, 1992|
|Previous Article:||Spare time use and psychological well-being in employed and unemployed young people.|
|Next Article:||The role of income, life-style deprivation and financial strain in mediating the impact of unemployment on psychological distress: evidence from the...|