THE INFLUENCE OF JOB INSECURITY ON JOB PERFORMANCE AND ABSENTEEISM: THE MODERATING EFFECT OF WORK ATTITUDES

Job insecurity was found to have relevant psychosocial consequences for both individuals and organisations. Recently, research is increasingly focusing on those variables that can moderate its negative influences. In this study, the impact of job insecurity on two indicators of organisational behaviour (i.e. job performance and absenteeism) was investigated. It was expected that job insecurity was negatively related to job performance and positively to absenteeism, and that this relationship was moderated by work related attitudes, such as job satisfaction and organisational commitment. Four-hundred and twenty five workers were interviewed with a structured questionnaire. Overall, the hypotheses were supported by the data: job insecurity was in fact negatively correlated with job performance and positively with absenteeism. However, work related attitudes moderated only the effect of job insecurity on job performance but not on absenteeism. werknemers is met ’n vraelys hipotese algemeen deur die data ondersteun: werksonsekerheid inderdaad negetiewe korrelasie tot werksprestasie en ’n positiewe korrelasie tot afwesigheid gehad. Werksverwante gesindhede het egter slegs die effek van werksonsekerheid werksprestasie gemodereer, maar nie op afwesigheid nie.


THE INFLUENCE OF JOB INSECURITY ON JOB PERFORMANCE AND ABSENTEEISM: THE MODERATING EFFECT OF WORK ATTITUDES
Requests for copies should be addressed to: N De Cuyper, nele.decuypre@psy.kuleuven.be (Klandermans & Van Vuuren, 1999;Hartley et al., 1991;Probst, 2003). However, feelings and perceptions of job insecurity may vary from one individual to another despite the objective situation they are currently living in (Greenhalgh & Rosenblatt, 1984;Klandermans & Van Vuuren, 1999;Hartley et al., 1991;Rosenblatt & Ruvio, 1996). After all, literature on stress has widely documented that reactions to stress depend on how individuals cope with the situation they are facing (Lazarus & Folkman, 1984), and has underlined the importance of internal resources (such as coping strategies, hardiness, and dispositional optimism) and external resources (such as material resources, social support) in coping with stress (see Parkes, 1994;Stroebe & Stroebe, 1995).
Similarly, research on job insecurity is recently focusing on those intervening variables that could mitigate and reduce its negative consequences . It was suggested that individual differences in personality traits such as negative affect and self-care (Roskies, Louis-Guerin & Fournier, 1993;Mak & Mueller, 2000), self-esteem and optimism (Makikangas & Kinnunen, 2003), emotional intelligence (Jordan, Ashkanasy & Hartel, 2002), locus of control and need for security (Ashford et al., 1989;Greenhalgh & Rosenblatt, 1984) may moderate the negative effects of job insecurity on well being outcomes. On the other hand, external resources such as social support (Lim, 1996), perceptions of process and work control (Barling & Kelloway, 1996;Tetrick & LaRocco, 1987), participations in decision processes (Parker, Chmiel & Wall, 1997), fairness of treatment and organisational justice were shown to affect work attitudes (like satisfaction and commitment) and well being, and thus may moderate the effects of job insecurity .
Following this line of research, the present study is aimed to investigate the impact of job insecurity on two indicators of organisational outcomes (i.e. job performance and absenteeism) and explore the potential moderation role of work attitudes that are strictly related to perceived organisational fairness and support, such as job satisfaction and organisational commitment (Rohades & Eisenberger, 2002). Sverke, Hellgren and Näswall (2002) made a theoretical distinction between short-term and long-term reactions to job insecurity. According to the authors, some kind of reactions, like work attitudes, would arise closer in time to the stress experience with respect to some others that are expected to emerge after a longer period of time, like behaviours and health complaints. There is empirical evidence proving that job insecurity is more strongly related to short-term outcomes (e.g. job satisfaction, organisational commitment) than to long-term reactions affecting the organization (e.g. performance, withdrawal behaviours like turnover intention, absenteeism, lateness and so on) . Therefore, firstly it is expected that job insecurity is negatively related to job satisfaction and organisational commitment (Hypothesis 1); secondly that job insecurity is negatively related to job performance and positively related to absenteeism (Hypothesis 2).
However, the most important predictions of this contribution concern the moderating role of work attitudes with respect to the impact of job insecurity on organisational outcomes. As previously stated, there is empirical evidence that different forms of perceived social and organisational support (e.g. family support, managerial support) may affect or reduce the negative consequences of job insecurity. Based on social exchange theory, perceived organisational support was conceptualised as employees' beliefs concerning the extent to which the organization values their role, treat them fairly and cares about their well-being in general (Eisenberger, Huntington, Hutchinson & Sowa, 1986). Thus, perceived organisational justice and support generate positive attitudes towards the organization and its activities, and are usually associated with more favourable evaluation of the organization, more satisfaction and more commitment (Rohades & Eisenberger, 2002). In turn, job satisfaction and organisational commitment predict job performance (Angle & Lawson, 1994;Judge, Thoresen, Bono & Patton, 2001;Leong, Randall & Cote, 1994;Meyer, Paunonen, Gellatly, Coffin & Jackson, 1989;Riketta, 2002). In fact, in two meta-analytic studies the mean true correlation between overall job satisfaction and job performance was estimated to be .30 (Judge, Thoresen, Bono & Patton, 2001) and .20 between organisational commitment and job performance (Riketta, 2002). It might be argued that if, to some extent, higher levels of job satisfaction and organisational commitment reflect higher perceived organisational support, then they may moderate the effects of job insecurity on the longterm organisational consequences (i.e. job performance and absenteeism). More precisely, the following hypotheses are stated: Hypothesis 3a: job satisfaction moderates the effect of job insecurity on job performance: namely, workers who perceive a high level of job insecurity in combination with a high level of job satisfaction will rate their job performance superior as compared to those perceiving high job insecurity together with low job satisfaction; Hypothesis 3b: job satisfaction moderates the effect of job insecurity on absenteeism: that is to say that workers who perceive a high level of job insecurity in combination with a high level of job satisfaction will report as having been less absent from work as compared to those perceiving high job insecurity together with low job satisfaction; Hypothesis 3c: organisational commitment moderates the effect of job insecurity on job performance: more precisely, workers who perceive a high level of job insecurity in combination with a high level of organisational commitment will assess their job performance as greater than those perceiving high job insecurity together with low organisational commitment; Hypothesis 3d: organisational commitment moderates the effect of job insecurity on absenteeism: i.e. workers who perceive a high level of job insecurity in combination with a high level of organisational commitment will report to have been less absent from work as compared to those perceiving high job insecurity together with low organisational commitment.

METHOD Participants
Four-hundred and twenty five workers participated in this study (219 males and 203 females; 3 neglected to state their gender). The average age of the participants was 37,5 years old (sd= 10,63), ranging from 18 up to 63 years old. Approximately 41.8% stated that they were married, while 44,7% declared that they were single. As regards education, 20,7% had a University degree, 70,1% had a secondary school degree, while the remaining 9.3% completed only primary school. The majority of the participants worked for private firms (64,9%), while 35,1% worked for a public company. As regards occupational positions, 72,5% had a full time job and 64,8% had a permanent job (compared to 35,2% who had a temporary job); on average, participants worked for 11.3 years (sd= 10,23), ranging from less than 1 year up to 42 years of work. About 21% were union members (mean= 13,1 years).

Measures and procedures
The study was conducted by autumn 2003 as part of a larger survey on related topics. Participants were administered a questionnaire by three interviewers, who were at the time undergraduate psychology students. Participants were contacted individually by the interviewers and then asked to volunteer for a survey on social and work attitudes. The questionnaire, on average, was completed in about 30 minutes, and contained the following measures: Job insecurity. Job insecurity was measured using five items focusing on workers' perception and worry of whether they would be able to keep their current job. This scale was fruitfully employed in previous cross-cultural studies, exhibiting excellent validity and reliability (i.a. Chirumbolo & Hellgren, 2003;Sverke, Hellgren, Näswall, Chirumbolo, De Witte & Goslinga, 2004). Examples of items were: "I fear I will lose my job" and "I am sure I can keep my job" (reverse scored item). Participants were asked to express their own agreement/disagreement with the statements on a Likert scale (from 1=strongly disagree; to 5=strongly agree). The scale reached a Cronbach alpha of 0.81. High scores indicate higher levels of job insecurity.
Job satisfaction. Job satisfaction was assessed with five items measuring the overall satisfaction with the present job (see De Witte, 2000;Hellgren, Sjöberg & Sverke, 1997). Participants were asked to express their agreement/disagreement with statements such as "I am very satisfied with my job" on a Likert-scale ranging from 1 (strongly disagree) to 5 (strongly agree). This scale was effectively employed in previous cross-cultural studies, exhibiting excellent validity and reliability (Sverke et al., 2004). In this study it reached a Cronbach alpha of 0,86 with high scores indicating increased job satisfaction.
Organisational commitment. Organisational commitment was measured using five items referring to the affective attachment towards the organisation (Allen & Meyer, 1990). A sample item is "I feel emotionally attached to my organisation". Participants expressed their agreement/ disagreement on a five-point scale (from 1 = strongly disagree; to 5 = strongly agree) such that higher scores indicate stronger organisational commitment. This scale was effectively employed in previous cross-cultural studies, showing excellent validity and reliability (Sverke et al., 2004). In the present study, the scale reached a Cronbach alpha of 0,89.
Job performance. Two items assessed overall job performance. The first item was "I achieved all my job goals in the last six months" to which participants had to state their agreement/disagreement (1 = strongly disagree; 5 = strongly agree). The second item was "In the last six months, your job performance was:" to which participants had to answer using a five point scale ranging from 1 = Low, to 5 = High. These two items were significantly correlated (r = 0,38, p< 0,001) and thus were aggregated to form a composite index of self rated job performance.
Absenteeism. Absenteeism was operationalised with a single item measure by asking how many days the interviewee was absent from work in the last six months.
Analysis of the data According to Baron and Kenny (1986), a variable (M) is said to moderate the relation of an independent variable (X) on a dependent variable (Y) when the strength of the association between X and Y depends on different levels of M. In general, moderator effects are indicated by the interaction of X and M in explaining Y, so that the following multiple regression equation must be calculated: The interaction between X and M, as shown by the product term X*M, measures the moderation effect. As suggested by Aiken and West (1991), before conducting the analyses, all the variables were standardized. Next, the two standardized predictors were multiplied to form the interaction term. The moderation was then tested via hierarchical regression analysis. The variables were regressed on the criteria (i.e. job performance and absenteeism in turn) with the predictors entered in the first step, followed by the interaction term in the second step (see Aiken & West, 1991, for more details).

RESULTS
Descriptive statistics of the scales and correlations among the variables are reported in Table 1. As expected, job insecurity was negatively correlated to job satisfaction (r= -0,19, p< 0,001) and organisational commitment (r= -0,24, p < 0,001). Job insecurity was also negatively correlated to job performance (r= -0,21, p < 0,001) and positively related to absenteeism (r= 0,23, p < 0,001). Additionally, job satisfaction and organisational commitment were found to be positively correlated to job performance and negatively to absenteeism (see Table 1). Overall, both hypothesis 1 and 2 were confirmed: job insecurity had a negative impact on work attitudes and organisational functioning.
In order to test the moderating hypothesis 3a, a hierarchic regression analysis was conducted using job insecurity, job satisfaction and the interaction term as predictors and job performance as criteria. The variables were entered in two steps. In the first step, job insecurity and job satisfaction were included in the equation as predictors, then the multiplicative term job insecurit y*job satisfaction was entered.
In the first step, both job insecurity (b = -0,11, t = -2,47) and job satisfaction (b = 0 .44, t= 9,91) showed a significant influence on job performance, explaining a significant proportion of variance of the criteria (R 2 = 0,32; F(2, 422)= 98,90, p< 0,001). However, most importantly, in the second step also the interaction term significantly predicted job performance ( b = 0,11, t= 2,71). [Refer to Table 2 for b, R 2 change and Fchange statistics]. In order interpret the direction of the interaction, a slope examination was conducted plotting the predicted values of job performance in function of job insecurity scores and three different levels of job satisfaction: low (one standard deviation below the mean), medium (the mean) and high (one standard deviation above the mean). As shown in Figure  1, job performance decreases when job insecurity is higher and job satisfaction is low or medium. However, there is no influence of job insecurity on job performance when job satisfaction is high. That is to say that job satisfaction moderates the impact of job insecurity on performance: hypothesis 3a was fully supported. In order to test hypothesis 3b, the same procedure was followed, employing absenteeism as criteria. In the first step, job insecurity and job satisfaction were included in the equation, then in step 2 the multiplicative term job insecurity*job satisfaction was entered.
In order to test hypothesis 3c, job insecurity and organisational commitment were included in the equation as predictors in the first step, then the multiplicative term job insecurity* organisational commitment was entered in step 2. In this case, job performance was the criteria.
In order interpret the direction of the interaction, a slope examination was conducted plotting the predicted values of job performance in function of job insecurity scores and three levels organisational commitment: low (one standard deviation below the mean), medium (the mean) and high (one standard deviation above the mean). As shown in Figure  1, job performance decreases when job insecurity is higher and organisational commitment is low or medium. However, when organisational commitment is high, there is a slight positive influence of job insecurity on job performance. In fact it seems that organisational commitment acts as a moderator of job insecurity on job performance: hypothesis 3c was fully supported.

Predictors\Criteria
Job performance Absenteeism

DISCUSSION
As some authors pointed out (Hartley et al, 1991;, the psycho-social factors that may intervene and moderate the negative outcomes of job insecurity represents a fruitful area of research from both the individual and the organisational perspective. Even if those factors can not modify the source of stress itself (i.e. they do not change the insecure employment situation into a more secure one), they may have beneficial effects for the individual and the organisation preventing the occurrence of the most negative outcomes, presumably through reducing the likelihood that job insecurity is perceived as stressful.
The main goal of the present study was to investigate the potential moderating role of work and organisational attitudes on the relationship between job insecurity and its organisational outcomes. Empirical evidence provided partial support to these expectations. Firstly, replicating previous results (e.g. Ashford et al., 1989;Davy et al, 1997;Hartley et al, 1991;Lim, 1996;Probst, 2002;Rosenblatt & Ruvio, 1996;Sverke et al., 2004), job insecurity was shown to have a negative impact to short-term organisational outcomes, namely job satisfaction and organisational commitment, as well as to long-term reactions, like job performance and absenteeism. Moreover, job satisfaction and organisational commitment were found to be positively correlated to job performance and negatively to absenteeism (Judge et al., 2001;Riketta, 2002). Nevertheless, this study has also shown that both job satisfaction and organisational commitment moderate the negative effects of job insecurity on job performance. More precisely, job insecurity exhibited its negative influence on performance only in conditions of lesser job satisfaction and lesser organisational commitment. On the contrary, when job satisfaction and organisational commitment were sufficient high, the influence of job insecurity on performance vanished. That is to say, regardless of job insecurity there is no difference in job performance rated by employees if job satisfaction and organisational commitment are high. No moderation effect was found as regards absenteeism: specifically, job insecurity, job satisfaction and organisational commitment do not interact each other in predicting employees' counter-productive behaviour like absenteeism. However, they have a main independent effect.
From a managerial perspective these results appear relevant for more effective human resource management. In fact, if on one hand this investigation has confirmed the negative influence of job insecurity on organisational outcomes, on the other hand it has underlined the importance of positive work attitudes in buffering some negative outcomes for the organization itself. Therefore, managers and employers should take into account that insecurity is detrimental for the organization, and try to cope with this phenomenon and prevent its negative reactions by enhancing employees' satisfaction and commitment, intervening, for instance, on their antecedents (e.g. Allen & Meyer, 1990;Cranny, Smith & Stone, 1992;Ostroff, 1992;Mathieu, 1991;Spector, 1997).
The present study has, still, some limitations. One shortcoming is represented by the measure of job performance used here. Several authors have pointed out the problematic issue of measuring performance through self-report questionnaires (Sverke, Hellgren & Naswall, 2002). There is empirical evidence that individuals tend to overrate their performance and that ratings of performance done by others (e.g. managers or colleagues) are usually more valid than selfrated performance measures (Ford & Noe, 1987). Moreover, job performance seem to be a multidimensional construct with a complex latent structure (Scullen, Mount & Goff, 2000). Therefore, it must be underlined that the promising results obtained in this study should be replicated in future research employing multiple measures of job performance, distinct from self report ratings.
Another problematic issue is the direction of the relationship between job insecurity and its supposed reactions. Although the present research was built upon a theoretical model derived from stress theory (Lazarus & Folkman, 1984;Jex, 1998;Stroebe & Stroebe, 1995), in which job insecurity is conceptualised as a stressor that would affect specific organisational behaviours and individuals' well-being (Ashford et al 1989;Dav y et al., 1991;Greenhalgh, & Rosenblatt, 1984;, correlational data do not allow to draw causal conclusions, resulting in a lack of internal validity of the research. In fact, it is also possible that employees dissatisfied with their job and not committed with the organization, achieve poor performance. Furthermore, they tend to carry out withdrawal behaviours and are also the ones more likely to experience job insecurity and to react more negatively to it (Hartley et al., 1991). The correlational data presented in this study cannot exclude this alternative explanation. Nevertheless, some empirical evidence provided by panel or experimental research seem to support the causal interpretation of the relationship and the validity of the theoretical model. A few longitudinal studies, in fact, illustrate that job insecurity, after controlling for prior level of well-being, actually predicts subsequent psychological well-being and physical complaints (e.g. Frese, 1985;Hellgren & Sverke, 2003). Moreover, in a field experiment, Probst (2003) found that organisational restructuring affected employees' perceptions of job security and had consequences on organisational commitment, turnover and psychological well-being. Likewise, in another laboratory experiment (Probst, 2002b) it was found that the threat of layoffs resulted in a lower quality of performance and in a greater violation of safety rules.
In conclusion, the cost of increasing flexibility and uncertainty of the job market appears to be considerable from social, individual as well as organisational perspectives. If some outcomes of job insecurity can have a direct effects for the individual and indirect impact on the organization (e.g., the case of mental and physical health), others instead tend to be directly damaging for the organization and only indirectly for the individual (e.g. the case of job satisfaction, organisational commitment and job performance) as was shown in this investigation.