European

Journal

of

Political

Research

46:

211-235,

2007

doi:

10.1111/j.1475-6765.2007.00692.x Workers, worries and welfare states: Social protection and job insecurity in 15 OECD countries CHRISTOPHER J. ANDERSON1 & JONAS PONTUSSON2 1Cornell University, USA; 2Princeton University, USA Abstract. This article examines a model of the domestic political economy of subjective employment insecurity in advanced industrial societies. Based on data on people's attitudes toward their job as well as levels of and kinds of social protection collected in 15 OECD countries, it shows that there are distinct manifestations of job insecurity that are affected differently by distinct aspects of social protection programs. While the analysis shows that social protection measures reduce employment insecurity, it also reveals that overall levels welfare state generosity do not have any systematic effect on whether workers feel secure. The article's findings suggest the need to decompose the different components of employment insecurity as well as disaggregate national systems of social protection when examining the impact of welfare states on job insecurity. Comparative political economists commonly assume that increased economic insecurity is one of the prime explanations for why welfare states have continued to grow, or why cutbacks have been relatively limited, despite tax fatigue and the economic pressures associated with globalization. Following Rodrik (1997) and Garrett (1998), much of the existing literature takes as axiomatic that insecurity, measured objectively, generates demand for social protection. Given the centrality that such reasoning has come to assume, the absence of any systematic analysis of the effects of social protection on individual perceptions of economic insecurity is striking. Drawing on 1997 survey data from 15 OECD countries, the following analysis focuses on whether and how cross-national differences in social protection affect the extent to which workers worry about losing their jobs. In estimating the effects of public policy and institutional arrangements, we draw on research in psychology to distinguish between different components of job insecurity, and explore how specific types of social protection (employment protection legislation, active labor market policies and unemployment insurance) affect these components. By disaggregating both job insecurity and systems of social protection, our analysis not only demonstrates that public provision of social protection does indeed reduce the job insecurity experienced by individuals, but also specifies the distinct causal pathways whereby this effect occurs.

We begin by reviewing general arguments about economic insecurity in the comparative political economy literature, motivating and situating our analysis in relation to this literature. We then present our model of the determinants of job insecurity and show how our survey data enable us to distinguish different components of job insecurity. We subsequently test this model in a two-step analysis, including country-level as well as individual-level variables in each step. In the first step, we develop and test hypotheses about the determinants of job insecurity and labor market insecurity and demonstrate that the determinants of these two manifestations of employment insecurity are systematically different. In the second step, we estimate the effects of unemployment compensation and overall social spending on worries about job loss while controlling for individuals' perceptions of their job insecurity (job-loss threat) and labor market insecurity. We conclude by discussing the implications of our empirical findings for the comparative political economy literature. Economic insecurity and social protection Economic insecurity matters greatly for people's private and public lives and, by implication, for the societies and polities in which they live. Political scientists have long been interested in economic insecurity because it affects people's views of the political system and their political behavior (Mughan & Lacy 2002). It is commonly accepted that high unemployment and economic insecurity have contributed to the downfall of democratic regimes (perhaps most famously the Weimar Republic), that economic grievances might lead to rebellion and revolution, and that governments are more likely to be voted out of office when the economy deteriorates. Finally, economic insecurity figures prominently in demand-driven explanations of the expansion of social insurance and other security-enhancing public policies, including restrictions on the ability of employers to fire workers at will. While economic insecurity, broadly defined, thus plays a major role in theories of public policy, political action, political behavior and democratic legitimacy, its actual definition and operationalization have frequently been ambiguous and variable. 'Economic insecurity' is often used as a kind of umbrella term for different manifestations of material well-being and, depending on the research question under investigation, its meaning ranges from a general sense of material well-being to job-related anxieties or individuals' assessments of recent changes in their personal financial situation. In part, this definitional variability has been a function of available data to measure people's sense of economic insecurity.

Our analysis of job insecurity is restricted to the determinants of job insecurity, but we conceive of 'job insecurity' as a multidimensional phenomenon that involves considerations beyond the probability of losing one's current job. As a number of studies have documented, employment insecurity appears to be on the rise across the industrialized democracies (OECD 1997). In recent literature on political economy, globalization figures as the most prominent culprit to explain these changes in employment insecurity and, again, increases in insecurity are commonly said to fuel demand for social protection. In Rodrik's (1997) much-cited formulation, globalization is a two-edged sword as it generates countervailing pressures on the welfare state. On the one hand, globalization puts downward pressure on the supply of social insurance by constraining the ability of governments to engage in deficit spending or to tax mobile factors of production. On the other hand, however, globalization increases the demand for social insurance by virtue of the economic insecurity that it breeds. In Garrett's words: '[T]he most important immediate effect of globalization is to increase social dislocations and economic insecurity, as the distribution of incomes and jobs across firms and industries becomes increasingly unstable' (Garrett 1998:7; see also Scheve & Slaughter 2004; Swank 2002; for earlier treatments of the openness/insecurity/ protection nexus, see Cameron 1978; Katzenstein 1985). In contrast, Iversen and Cusack (2000) argue that cross-national differences in economic insecurity in advanced capitalist societies are due to the transformation of labor markets produced by de-industrialization rather than globalization. More specifically, they argue that de-industrialization renders workers with asset-specific skills more insecure. Yet, Iversen and Cusack's explanation of patterns of welfare spending in OECD countries also hinges on the proposition that economic insecurity is the principal source of demand for publicly provided social welfare and protection. Consistent with this, Iversen and Soskice (2001) show that workers with asset-specific skills are particularly inclined to support the welfare state. Thus, regardless of the mechanism underlying the rise in job insecurity, such insecurity is assumed to fuel demand for social protection. Though theoretically central to the debates outlined above, the notion that social protection actually reduces insecurity has yet to be subjected to any systematic empirical examination. This is the lacuna that we seek to address in this article by focusing on the question of whether and exactly how public welfare provisions affect individual perceptions of job insecurity. Our model of job insecurity

Our conceptual framework draws on psychological research on job insecurity. Psychologists define 'job insecurity' in somewhat different ways, but their definitions invariably involve a perceived threat to continuity in one's job situation and typically also some sense of powerlessness in the face of this threat (e.g., see Greenhalgh & Rosenblatt 1984: 438; Sverke & Hellgren 2002: 25-26). In contrast to 'job loss', which is an objective state of affairs, 'job insecurity' is a product of people's interpretations of signals in the environment (Jacobson 1991; Sverke et al. 2002; Hartley et al. 1991). While the early literature on job insecurity as a psychological phenomenon conceptualized it in purely cognitive terms (Ashford et al. 1989; Greenhalgh & Rosenblatt 1984), more recent work argues that job insecurity involves affect as well as cognition (cf. Sverke & Hellgren 2002; see also Borg 1992; Borg & Elizur 1992). In what follows, 'cognitive job insecurity' is an individual's estimate of the probability that he or she will lose their job in the near future, while 'affective job insecurity' refers to worry or anxiety about losing one's job. Thus conceived, cognitive job insecurity is a determinant of affective job insecurity, but affective job insecurity involves more than a perceived threat to one's current job status. Our approach to explaining job insecurity posits that 'affective job insecurity' (i.e., the extent to which an individual worries about losing their job) is fundamentally a function of two variables: the individual's estimate of the probability that he or she will lose her job ('cognitive job insecurity', in the language of psychology) and the individual's perception of the consequences of losing their job.1 Our model in turn decomposes the expected consequences of losing one's job into two discrete variables: the prospects of finding another (more or less equivalent) job and access to sources of income (livelihood) that do not depend on finding another job.2 Prior research clearly demonstrates that cognitive job insecurity is a major determinant of affective job insecurity (Borg 1992; Borg & Elizur 1992). What, then, drives cognitive job insecurity? Moving down the causal chain, it seems reasonable to suppose that individuals take into account a number of objective factors in estimating the probability that they might lose their job. We group these factors under three headings: labor market conditions, individual employability attributes and institutions providing for employment protection. For one, workers look to the labor market for cues in seeking to assess how secure they are in their current employment (cf. Green et al. 2000). The second set of factors has to do with employability attributes or, in other words, the individual characteristics that make workers more or less valuable to their employers.3 The key issue here is 'human capital'. The standard Hecksher-Ohlin-Samuelson model of factor

endowments implies that unskilled labor is most likely to suffer economic insecurity in advanced industrialized economies. Our empirical analysis also considers the effects of age and gender on cognitive job insecurity. To the extent that employers formally or informally take seniority into account in workforce reductions, older workers can be expected to feel more secure in their current jobs than younger workers. We can distinguish two kinds of institutions - or institutional practices -that provide for employment protection. One has to do with labor relations in individual firms or workplaces, while the other has to do with government regulation of employment conditions. As indicated at the outset, one of our primary concerns here is to determine whether public policies affect job insecurity. When it comes to people's sense of whether their current job is secure, the relevant policy should relate to restrictions on employers to fire workers. Regarding institutional arrangements at the workplace, our empirical analysis tests two rather obvious hypotheses: first, workers who are employed in the public sector should feel more secure in their jobs than workers who are employed in the private sector; and second, workers who belong to unions or work in unionized firms should also feel more secure in their jobs (Bender & Sloane 1999). Our model posits further that worry about losing one's job entails more than some expectation that this might happen. The other dimension of affective job insecurity boils down to the question 'What happens to me (and my family) if I do lose my job?' Simplifying a great deal, the expected consequences of losing one's job can in turn be treated as a function of both replacement job prospects and access to sources of income (livelihood) that do not depend on finding another job. We use the term 'labor market insecurity' to designate a low probability of finding another job with more or less equivalent characteristics and the term 'income insecurity' to refer to the absence of other sources of income.4 Like cognitive job insecurity (the estimated probability of losing one's job), labor market insecurity can be expected to depend on labor market conditions and individual employability attributes. Higher unemployment reduces the prospects of finding another job at the same time as it increases the prospects of losing one's current job. Similarly, skilled workers are not only less likely to be laid off than unskilled workers; they are also more likely to find other jobs, especially equivalent jobs. With respect to age, however, our expectations for its effects on job insecurity and labor market insecurity diverge: older workers can be expected to feel more secure in their current jobs, but less secure in the labor market (cf. Naswall et al. 2002). There is no

obvious reason to expect that institutions that provide for employment protection should affect workers' assessments of their prospects in the labor market. Instead, the mechanism whereby public policy should affect such assessments consists of the effects of active labor market policies.5 For the most part, such policies seek to increase the employability of unemployed workers. We expect that greater commitment of societal resources to active labor market policies reduces labor market insecurity (i.e., beliefs about replacement job prospects). Finally, the issue of sources of income or consumption that are not dependent on employment brings us to the welfare- or security-enhancing functions of social programs emphasized by the welfare-state literature. In our model, the level of income replacement provided by unemployment insurance and the duration of unemployment benefits should affect workers' job insecurity to the extent that job insecurity is a function of worries about income loss while they are unemployed, or, in other words, what we refer to as 'income insecurity'. Because of 'decommodification' (the loosening of the link between income/ consumption and employment), we expect the size of the welfare state to be associated with a general sense of economic security and, as a result, less anxiety about losing one's job (i.e., reduced affective job insecurity). Income pooling within households (families) represents another source of economic support in case of job loss (Esping-Anderson 1999). Though increasing marital instability may have weakened the security-enhancing role of the family, someone who is part of a household that includes other income-earners should worry less about the possibility of losing his or her job than a person who is the sole 'breadwinner'. By the same logic, anxiety about losing one's job can be expected to decline as the income of other household members increases. Figure 1 graphically depicts the model of job insecurity sketched above. Measuring insecurity with survey data To test the hypotheses generated by our model of job insecurity requires information about individual-level attitudes and attributes as well as system-level information about labor market conditions and social-protection arrangements. The individual-level data that we use are drawn from surveys conducted in 1997 under the auspices of the International Social Survey Program (ISSP), as part of a study called Work Orientations II. Our analysis includes the following 15 OECD countries: Canada, Denmark, France, Germany, Great Britain, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland and the United States. As these countries differ widely with regard to public provisions of social protection as well as

economic and cultural characteristics, they would seem to constitute a most appropriate sample of countries with which to conduct our analyses. In particular, we note that each of Esping-Andersen's 'three worlds of welfare capitalism' (Esping-Andersen 1990) is represented by at least three or four countries (depending on the coding of particular countries). We make use of responses to three survey questions to measure different components of job insecurity. We measure cognitive job insecurity (i.e., a more or less dispassionate assessment of the probability that one might lose one's job in the near future) with the help of a question that asks respondents how much they agree or disagree with the statement that 'my job is secure' on a scale from 1 to 5. To gauge individuals' labor market insecurity, respondents were asked 'How easy or difficult do you think it would be for you to find an acceptable job?' Again, respondents were presented with five possible answers, ranging from 'very easy' to 'very difficult'. Finally, to measure affective job insecurity, respondents were asked 'Do you worry about the possibilities of losing your job?' Respondents were given four possible answers (coded 14): 'I don't worry at all', 'I worry a little', 'I worry to some extent' and 'I worry a great deal' (see Appendix A for variable coding). Our empirical analysis thus proceeds in two steps. First, we test models of cognitive job insecurity and labor market insecurity using answers to the first and second survey questions as the dependent variables. Second, we test models of affective job insecurity that include answers to the first and second survey questions as independent variables.6 Because our general model of employment pertains to job insecurity among employed workers, we restrict our analysis to individuals who declare

Figure 1. Our model of job insecurity. Note: Individual perceptions of job-loss threat and replacement prospects are measured directly. Perceived income insecurity is measured indirectly (dashed line). that they are gainfully employed at the time of the survey. Figure 2 shows the combined percentages of employed respondents who said that they worry 'a great deal' or 'to some extent' about losing their job (the third survey question) across the 15 countries. As the graph shows, there is a remarkable degree of cross-national variation in worries about potential job loss, ranging from over 50 per cent in Spain to only slightly more than 10 per cent in countries such as Denmark, the Netherlands and Norway Figure 3 plots the survey data in Figure 2 against total social spending as a percentage of GDP. Contrary to what some of the existing literature would seem to imply, there appears to be no association whatsoever between affective job insecurity and the size of the welfare state on a cross-national basis. This remains true if we control for rates of unemployment, which do correlate rather closely with levels of affective as well as cognitive job insecurity. In a simple cross-national OLS regression model with affective job insecurity as the dependent variable and total social spending and rates of unemployment as the independent variables, the coefficient for social spending is indeed negative, but the standard error is considerably larger than the coefficient. Cognitive job insecurity and labor market insecurity Because we combine information at the (micro) level of respondents and the (macro) level of countries, our data have a multilevel structure where one unit of analysis is nested within the other (Bryk & Raudenbush 1992). Tables 1 and 2 (below) therefore present the results of iterative generalized least squares

Figure 2. Worry about losing job. (IGLS) multilevel regression models designed to explain cognitive job insecurity (Table 1) and labor market insecurity (Table 2) using data both at the macro- (country) and micro- (individual) level. We estimate the same basic statistical model for both dependent variables in order to establish how their determinants differ:

We are primarily interested in the effects of political-economic arrangements associated with the welfare state or, in other words, public provision of social protection. In this first step, two such arrangements are particularly pertinent: employment protection laws and active labor market policies. To measure the stringency of legally mandated employment protection, we rely on a composite index developed by the OECD (1999). Active labor market policies essentially have to do with retraining unemployed workers and otherwise helping them find new jobs. We measure the scope of active labor market programs and the

Figure 3. Scatterplot of job worry and social spending. Table 1. Models of cognitive job insecurity in 15 OECD countries, 1997 Explanatory variables Model 1 Model 2 Model 3

Public policies Employment legislation (high = more protection) Active labor market (0.101) policy 0.006 (0.102) 0.006 (0.106) 0.006 protection -0.199* -0.220* -0.219*

(spending per unemployed) Workplace institutions Union member (1 - union -0.145** member; 0 = else) (0.025) (0.027) -0.314** (0.030) -0.321** -0.113** -0.114** (0.017) (0.017) (0.017)

Public sector employment (1 = public sector employee; 0 = else) Labor market conditions 1992-1997 unemployment rate 0.034* (%) (0.017) 1996-1997 employment rate (0.041) Individual attributes Education education) (0.003) Manual worker (1 = yes; 0 = else) Gender (1 = female; 0 = male) -0.036 (0.022) Age (actual age) -0.005** (high = high -0.017** change in 0.093* -

(0.031)

(0.032)

0.035*

0.036*

(0.016) 0.096**

(0.017) 0.095**

(0.040)

(0.042)

-0.009**

-0.005

(0.003) -

(0.004) 0.092**

0.002 (0.023) -0.005

(0.031) -0.015 (0.025) -0.005**

(0.001) Constant 3.057** (0.222) Variance components Country-level 0.046** (0.017) Individual-level 1.306** (0.017) N -2 log likelihood 11,461

(0.001) 3.075** (0.246)

(0.001) 2.975** (0.254)

0.042** (0.016) 1.305** (0.018) 10,273

0.044** (0.018) 1.328** (0.20) 8,846

35,635.07 31,936.44 27,657.44

Notes: Estimates are maximum likelihood estimates (IGLS) using MLwiN 1.10.0006 (2000); standard errors in parentheses. The dependent variable for job loss probability contains five response categories ranging from 'strongly agree' (1) to 'strongly disagree' (5) when asked about the statement 'my current job is secure'. Model 2 does not include the United States; Model 3 does not include the United States and the Netherlands. * p < 0.05; ** p < 0.01; one-tailed tests of significance. Table 2. Models of labor market insecurity in 15 OECD countries, 1997 Explanatory variables Public policies Employment legislation (high = more protection) (0.130) (0.102) -0.057** (0.102) -0.059** protection 0.134 0.018 0.029 Model 1 Model 2 Model 3

Active labor market policy (% of -0.042* GDP per capita unemployed) Workplace institutions Union member (1 - union 0.071** member; 0 = else) (0.024) (0.022)

(0.017)

(0.017)

0.062**

0.072**

(0.025) 0.053*

(0.028) 0.038

Public sector employment (1 = public

sector employee; 0 = else) Labor market conditions

(0.029)

(0.030)

1992-1997 unemployment rate 0.006 (%) (0.021) 1996-1997 employment rate (0.053) Individual attributes Education education) (0.003) Manual worker (1 = yes; 0 = else) Gender (1 = female; 0 = male) 0.182** (0.021) Age (actual age) 0.022** (0.001) Constant 2.836** (0.222) Variance components Country-level 0.077** (0.028) Individual-level 1.200** (0.016) N -2 log likelihood 11,513 (high = high -0.035** change in 0.129**

0.001

0.006

(0.016) 0.139**

(0.016) 0.136**

(0.040)

(0.040)

-0.035**

-0.027**

(0.003) -

(0.004) 0.179**

0.189** (0.022) 0.022** (0.001) 3.287** (0.243)

(0.029) 0.168** (0.024) 0.021** (0.001) 3.238** (0.244)

0.042** (0.016) 1.179** (0.016) 10,289

0.041** (0.016) 1.169** (0.18) 8,834

34,834.49 30,944.04 26,493.08

Notes: Estimates are maximum likelihood estimates (IGLS) using MLwiN 1.10.0006 (2000); standard errors in parentheses. The dependent variable for job loss probability contains five response categories ranging from 'very easy' (1) to 'very difficult' (5) when asked about 'How easy

or difficult do you think it would be for you to find an acceptable job?' Model 2 does not include the United States; Model 3 does not include the United States and the Netherlands. * p < 0.05; ** p < 0.01; one-tailed tests of significance. level of support that they provide by dividing the amount of money that governments spend on these programs, expressed in thousands of US dollars at purchasing power parities, by the number of unemployed.7 Consistent with the analytical model articulated above, we expect employment protection to be associated with cognitive job insecurity, but not with labor market insecurity, and we expect the association with cognitive job insecurity to be negative. Conversely, we expect government spending on active labor market programs to be negatively associated with labor market insecurity, but we do not expect active labor market policies to be associated, one way or the other, with cognitive job insecurity. For obvious reasons, it is important to control for other variables that might affect job insecurity and labor market insecurity in estimating these effects. At the individual level, we therefore control for public sector employment (1 = yes; 0 = no) and union membership (1 = union member, 0 = otherwise). While measured at the level of individuals, public-sector employment and union membership are first and foremost characteristics of workplaces that workers do not take with them when they leave an employer and re-enter the labor market. While there are strong reasons to expect that public-sector employment and union members are associated with lower levels of cognitive job insecurity, we cannot think of any obvious reason why these variables should affect labor market insecurity. Macro-economic conditions are likely to influence individuals' perceptions of the security of their current job, as well as their estimates of how difficult it would be to find another one. The most relevant macro-economic variable is surely the rate of unemployment. Persistently high levels of unemployment should be positively associated with both forms of employment security, but it also seems plausible to suppose that individuals are particularly affected by recent changes in the rate of unemployment as they assess their prospect of keeping their current job or finding another job. Thus the regression models reported in Tables 1 and 2 include the change in national unemployment rates over the year prior to the ISSP surveys (i.e., the change from 1996 to 1997) as well as the average rate of unemployment for the 1992-1997 period. We also control for characteristics that make individuals more or less attractive to employers. Again, the key issue here is human capital or skills. Drawing on the ISSP surveys, we construct

two individual-level variables that address this issue: education and manual labor (see Appendix A). We expect years of education to be negatively associated with insecurity in one's current job as well as insecurity about alternative employment opportunities, and we expect manual labor to be positively associated with both components of job insecurity. Our regression models also include the age of the respondent (a continuous variable) and a dummy for gender (1 = female, 0 = male). We expect age to be negatively associated with job insecurity on account of firms using more or less formalized seniority rules in decisions about lay-offs. At the same time, and partly for this very reason, we expect older workers to be more worried than younger workers about their ability to find another equivalent or acceptable job (Naswall & De Witte 2003). Our expectations concerning the effects of sex are equivocal. While women tend to occupy more marginal or precarious positions in the labor market than men, we also know that demand for female workers has increased in many OECD countries over the last several decades. Also, most of the countries covered by our analysis have laws against sex discrimination in employment practices, including lay-offs and other forms of termination. To the extent that informal sex discrimination persists, it should primarily affect women's estimates of their replacement job prospects rather than their estimates of the security of their current job. Three different models are reported for each dependent variable (cognitive job insecurity and labor market insecurity) in Tables 1 and 2 because of missing observations in the ISSP surveys on two important variables. The American survey does not distinguish between public and private-sector employees and the Dutch survey does not report occupational status. To check their robustness, we report results with and without these variables. For the most part, the results presented in Tables 1 and 2 confirm our expectations. We find that respondents in countries with higher levels of unemployment and those that experienced recent increases in unemployment rates are significantly more likely to state that their current job is insecure. When we switch from cognitive job insecurity to labor market insecurity as the dependent variable, the coefficient for unemployment levels becomes very small and no longer statistically significant, but the size of coefficient for recent changes in unemployment increases appreciably. When assessing their prospects of finding alternative employment, workers are apparently most sensitive to recent changes in unemployment and do not consider long-term unemployment rates to be a very meaningful cue.

In the first two models reported in Table 1, education has a significant negative association with cognitive job insecurity, but the size of the coefficient is nearly halved when we control for public-sector employment. The size of the coefficient diminishes further and loses statistical significance once we also introduce the dummy variable for manual labor. Education turns out to be more strongly and consistently associated with positive assessments of alternative job prospects. Relative to other workers, unskilled manual workers feel significantly more insecure in both their current jobs and in the labor market. The results concerning age are also statistically significant, and confirm our expectations: cognitive job insecurity declines with age, but labor market insecurity increases with age. With cognitive job insecurity as the dependent variable, the sign of the coefficient for sex changes depending on the model specification, but none of these coefficients are statistically significant. This indicates that women feel neither more nor less secure in their current jobs than men. However, we do find women have more negative assessments of their alternative employment prospects than men. Again consistent with our expectations, public-sector employees and union members feel significantly more secure in their current jobs than private-sector employees and unorganized workers. Though the effects are considerably smaller, these dummy variables also turn out to be significantly associated with labor market insecurity. In general, public-sector employees and union members have more negative assessments of their prospects in the labor market. Quite possibly, this result, which is not predicted by our analytical model, has to do with public-sector employment and union membership generating different (higher) standards for what constitutes an 'acceptable job'. Finally, and most importantly, the results of these regression models confirm our expectations concerning the effects of public policies. Stronger legal provisions for employment protection are indeed associated with less cognitive job insecurity, but have no effect on individuals' assessments of alternative employment prospects. Estimating the substantive impact of employment protection, our results indicate that increasing the OECD employment protection index by one standard deviation (while holding all other variables at their mean) decreases job insecurity by 0.21. Starting from mean levels of employment protection, job insecurity would drop from 2.44 to 2.23. Government spending on active labor programs thus generates more positive assessments of alternative employment prospects, but has no effect on cognitive job insecurity. With a one standard deviation increase of government spending, such programs labor market insecurity

decreases by 0.23 according to our results. Starting from mean levels of active labor market policy, labor market insecurity would drop from 3.43 to 3.20. More generally, the results reported in Tables 1 and 2 clearly indicate that cognitive job insecurity and labor market insecurity are distinct phenomena with different underlying determinants. Income insecurity and affective job insecurity We now consider the question of whether non-market sources of income support influence affective job insecurity by estimating the following model:

We use two measures of public income support: a broad one and a narrow one. The broad measure is also the most commonly used measure of the size of the welfare state and consists simply of total social spending in percentage of GDP (from the OECD's Social Expenditure Database). Our narrow measure of public income support pertains to the generosity of unemployment compensation. Recent literature uses the net income replacement provided by unemployment insurance as a measure of welfare-state generosity vis-a-vis the unemployed (Korpi & Palme 2003). For our purposes, this measure is problematic because it ignores cross-national variation in the coverage of unemployment insurance (the percentage of the labor force that is eligible for unemployment insurance) as well as the duration of unemployment benefits. Our preferred measure of the generosity of unemployment compensation is constructed in the same manner as our measure of societal commitment to active labor market policy programs - that is, as government spending on income support for the unemployed per unemployed person in thousands of US dollars (at purchasing power parities). As indicated earlier, income pooling within households can be seen as another (private) source of non-market support in case of employment loss. Worries about losing one's job can be expected to decline as the income of other household members increases. While the survey data do not include any information on total household income, they do indicate whether the respondent has an employed spouse/partner. We therefore include a dummy variable for 'employed spouse/partner', which we expect to be associated with less worry about job loss (cf. Lim 1996).8 As the results presented in Table 3 show, the individual-level variables included in our model are highly significant determinants of affective job insecurity. Not surprisingly, individuals who consider their jobs to be less insecure worry more about losing their job. Though the effects of

cognitive job insecurity are considerably greater, the results also show that individuals with more negative assessments of replacement job prospects worry more about losing their job. Controlling for these effects, having an employed spouse or partner does not have any statistically significant effect on job-loss worries, though the sign of coefficient is consistent with our expectations. The overall size of the welfare does not affect job insecurity, but the generosity of unemployment compensation does have a significant security-enhancing effect. Substantively speaking, our results suggest that increasing the generosity of unemployment compensation by one standard deviation decreases the level of worries about job loss by 0.15. Starting with unemployment compensation levels at their mean, job-loss worry drops from 1.56 to 1.41. The results presented in Table 3 should not be taken to mean that the unemployment compensation is the only feature of national systems of social protection that matter to worries about losing one's job (affective job insecurity). As we saw earlier, employment protection and government spending on active labor market programs are important determinants of job insecurity and labor market insecurity. Not surprisingly, cognitive job insecurity and labor market insecurity are in turn powerful predictors of affective job insecurity. For illustrative purposes, the estimates of the effects of employment protection and active labor market programs on cognitive job insecurity and labor market insecurity produced by our first-step regressions can be plugged into our second-step regressions to obtain estimates of the effects of these variables on affective job insecurity. Taking these 'pass-through effects' into account, Table 3. Models of affective job insecurity in 15 OECD countries, 1997 Explanatory variables Assessments employment Job insecurity (high = insecure) Labor market insecurity (high = difficult) Nonmarket support Total social -0.004 0.014 0.013 0.314** (0.007) 0.107** (0.007) 0.314*** (0.007) 0.107*** (0.007) 0.314*** (0.007) 0.107*** (0.007) 0.319** (0.008) 0.102*** (0.008) of current and future Model 1 Model 2 Model 3 Model 4

spending (% of GDP) Spending on unemployment compensation per unemployed Employed spouse (1 = yes; 0 = no) Constant 0.868** (0.299) Variance components Country-level 0.071** (0.026) Individual-level 0.670** (0.009) N -2 log likelihood 12,391 0.061** (0.022) 0.670** (0.009) 12,391 0.057** (0.020) 0.670** (0.009) 12,391 30,266.83 0.058** (0.021) 0.662** (0.010) 9,261 22,519.55 0.925** (0.111) 0.668* (0.288) -0.025 (0.019) 0.669* (0.292) (0.012) -0.017* (0.010) (0.014) -0.024* (0.012) (0.014) -0.023* (0.013)

30,270.22 30,267.73

Table 4 summarizes the substantive effects of different social protection measures suggested by our analysis. Rather than simulating the effect of one-standard-deviation changes, the counterfactual exercise presented in Table 4 asks 'What would happen to various components of job insecurity if the United States were to adopt Swedish levels of employment protection, active labor market programs and unemployment compensation?' While adopting the Swedish level of unemployment compensation alone would reduce the job-loss anxiety of the average American from 1.64 to 1.45, the combined effect of all three changes would be to reduce the job-loss anxiety of the average American from 1.64 to 1.26.9 The point of the counterfactual exercise presented in Table 4 is simply to illustrate the size of the substantive effects yielded by our analysis. As a benchmark in thinking about the substantive significance of the effects shown in Table 4, it might be useful to keep in mind that average job-loss anxiety for American survey respondents with less than twelve years of education ('high-school dropouts') was 1.74, while the corresponding figure for American respondents with

more than 15 years of education ('college graduates') was 1.62. From this perspective, the substantive effects of security-enhancing public policies uncovered by our analysis would appear to be quite significant indeed. Conclusion The analytical model developed in this article incorporates insights from the existing welfare-state literature as well as the literature on job insecurity in Table 4. Counterfactual estimates of the United States adopting Swedish levels of social protection Cognitive job Insecurity (actual level: 2.24) Employment -0.42 Active labor market programs Unemployment compensation Note: Entries denote changes in individual insecurity levels. Job-loss worry is measured on a 1-4 scale, while job insecurity and labor market insecurity are measured on 1-5 scales. EPL values are 0.7 for the United States and 2.6 for Sweden; ALMP value 1.98 for the United States and 12.20 for Sweden; and unemployment compensation value 3.41 for the United States and 11.72 for Sweden. psychology. Our model decomposes the determinants of affective job insecurity or worry about losing one's job into three separate components: estimates of the probability of losing one's current job (cognitive job insecurity), estimates of one's ability to find another job (labor market insecurity) and availability of income during an extended unemployment spell (income insecurity). Our empirical analysis does not exactly prove that this is the right way to think about job insecurity and its determinants, but the model yields a number of specific hypotheses that are indeed supported by the our empirical analysis. Our results confirm that more educated workers, and especially workers in non-manual occupations, tend to feel more secure in their current job while at the same time being more sanguine about their prospects of finding another job. Age, union membership and public-sector -0.19 -0.60 -0.06 protection Labor market insecurity (actual level: 2.53) Affective job insecurity (actual level: 1.64) -0.13

employment are associated with less cognitive job insecurity, but more insecurity about alternative employment prospects. Controlling for other individual attributes, women are significantly more worried about alternative job prospects than men. Turning to macro-level determinants, our analysis also confirms that labor market conditions powerfully shape employment-related worries of individuals. Not surprisingly, rising and persistent high national unemployment rates are associated with higher cognitive job insecurity, while rising unemployment is associated with more pessimistic assessments of alternative employment prospects. As for the security-enhancing effects of public policy provisions associated with the welfare state (the question of primary interest to political scientists), our empirical analysis yields strong support for the idea that there are three distinct pathways whereby public policy provisions influence workers' employment-related worries. First, government legislation restricting the ability of employers to fire workers and/or imposing costs on employers who do fire workers appears to have a quite significant impact on individuals' assessment of how secure their jobs are (cognitive job insecurity). Second, government spending on labor market programs designed to improve the employability of unemployed workers and to help them find new jobs reduces labor market insecurity. Third, generous unemployment compensation reduces worries about the income loss associated with unemployment. Again, the data at hand do not allow us to estimating the latter effect directly, but the circumstantial evidence obtained by estimating the effect of unemployment compensation on affective job insecurity while controlling for cognitive job insecurity and labor market insecurity is consistent with this inference. Somewhat surprisingly, our analysis fails to confirm the hypothesis that income pooling within households reduces affective job insecurity. One possible explanation for is that our measure of income pooling within households (a dummy for employed spouse) is too crude to capture the expected effect. Alternatively, our null finding on this score might be interpreted as support for the thesis that the salience of households for individual perceptions of economic prospects has declined as a result of rising marital instability in advanced industrial societies over the last 20-30 years (cf. Iversen & Rosenbluth 2003). Our conclusions about the effects of public policies must be tempered by recognizing that our analysis does not take into account institutional legacies, nor does it explore the effects of institutional change. In our analysis, incremental changes in spending on existing active labor

market programs or unemployment compensation are assumed to have the same 'per-unit effects' as the introduction of new programs, or the dismantling of existing programs. For the United States to adopt active labor market policy on a Swedish scale would obviously involve a major political struggle, and the effects of such a reform on workers' subjective sense of labor market insecurity could well be much greater than the simulation results reported in Table 4 suggest. Put differently, it may well be the case that public policy provisions come to be taken for granted and that their insecurity-reducing effects diminish over time. Much of the recent literature on welfare-state retrenchment emphasizes the incremental nature of the changes that have taken place in OECD countries, but public policy provision pertaining to employment protection and unemployment compensation certainly became politically contested in many of these countries in the 1990s. The story of this political contestation varies across countries and the implications of such variation for perceptions of job insecurity deserve further attention. Keeping the limitations of our analysis in mind, we do not wish to imply that governments should necessarily increase employment protection while boosting spending on active labor market programs and unemployment compensation if they seek to reduce affective job insecurity. For one, the rate of unemployment represents another important lever whereby government policies might workers' sense of insecurity. Moreover, conventional wisdom among economists suggests that employment protection and the payroll taxes associated with generous unemployment compensation might in fact reduce demand for labor and thereby expose workers to greater risk of losing their jobs. While the cross-national evidence on the growth-depressing effects of social protection is not as clear-cut as conventional wisdom among economists would lead us to expect, it is certainly plausible that employment protection and unemployment compensation might have negative consequences for cognitive and affective job insecurity via labor market effects that are not captured by our analytical model. To pursue this issue further would require an entirely different mode of empirical analysis. Similarly, we must leave aside the question of the extent to which governments have the ability to boost employment growth by means of demand stimulation or industrial policy. In a somewhat different vein, employment protection, active labor market policy and unemployment compensation might be seen as alternative means to achieve a certain, desired level of employment security. Arguably, parties across the political spectrum are distinguished by what they consider to be acceptable levels of job insecurity. Rather than pursuing further reductions in

insecurity ad infinitum, even the most security-oriented parties have a security 'target' that they seek to reach. As our analytical model and empirical analysis illustrate, any given target might be achieved by different combinations of security-enhancing measures. Again, the question of why governments might choose different combinations of employment protection, active labor market programs and unemployment compensation lies beyond the confines of this article. Suffice it to note that in the 1950s and 1960s Scandinavian Social Democrats eschewed employment protection, arguing that the proper role of government was to provide workers with greater security in the labor market through unemployment compensation and measures to enhance employabil-ity. In the 1990s, the politics of the Third Way again shifted the emphasis of labor-affiliated parties in Western Europe from employment protection to employability. Judging by our empirical results, relying exclusively on active labor market programs as the solution to job insecurity would not appear to be a politically viable formula for Social Democratic parties, but it may well be viable for these parties to trade employment protection for employment growth so long as they also maintain existing levels of unemployment compensation. As noted throughout our discussion, job insecurity is a subjective phenomenon. Our empirical results suggests that average individuals have a fairly clear idea of the objective variables that affect their job insecurity and that they are quite discriminating as they respond to survey questions about job insecurity. Consistent with ongoing related research in organizational psychology (cf. Sverke et al. 2002), job insecurity appears to have rather hard and precise connotations. Our findings that employment protection reduces cognitive job insecurity while government spending on active labor market programs reduces labor market insecurity suggest that individual respondents discern the distinctive purposes of these government policies. Also, it seems significant from this point of view that the public provision of unemployment compensation reduces affective job insecurity, but public provision of social welfare more generally does not have any impact on worries about losing one's job. These results would appear to be inconsistent with the idea that insecurity is a state of mind that can be affected simply through more generous social welfare provisions. The absence of any discernable effects of total social spending levels on individual perceptions of job insecurity casts some doubt on the importance assigned to insurance motives in much of the comparative welfare state literature. Our analysis suggests that rising job insecurity might well be invoked to explain public support for specific welfare-state programs, but the link between job

insecurity and the overall size of the welfare state appears is tenuous, at best. In this sense, our results might be interpreted as supporting the idea that redistributive motives (or interests) deserve a more prominent place in general theorizing about the politics of the welfare state. Acknowledgment We are grateful to David Rueda for his comments and help with data collection. Appendix A: Variables and question wording Worry About Job. 'Do you worry about the possibilities of losing your job?' I don't worry at all (0); I worry a little (1); I worry to some extent (2); I worry a great deal (3). Possible Job Loss. 'How much do you agree or disagree that these statements apply to your job?' My job is secure: Strongly disagree (0); Disagree (1); Neither agree nor disagree (2); Agree (3); Strongly agree (4). Job Replacement. 'How easy or difficult to you think it would be for you to find an acceptable job?' Very easy (0); Fairly easy (1); Neither easy or difficult (2); Fairly difficult (3); Very difficult (4). Employed. Are you currently working for pay?' yes = 1; no = 0. Spousal Employment. What is the employment status of your spouse/ partner?' Employed full- or part-time = 1; otherwise = 0. Sex. Female = 1; male = 0. Age. Actual age. Manual Laborer. Recoded from variable categorizing respondent's occupation. Unskilled professions = 1; otherwise = 0. Education. Age when respondent left school. Union Membership. Are you a member in a trade union at the present time?' yes = 1; no = 0. Public Sector Employment. Do you work for the public sector?' yes = 1; no = 0. Spending. Social expenditures (including mandatory private benefits) as percentage of GDP, 1995 (source: OECD 2001). EPL. Composite employment protection index (source: OECD 1999). Active labor market policies. Government expenditures on active labor market programs per unemployed in thousands of US dollars at purchasing power parities, 1995 (source: OECD 2001). Unemployment compensation. Government expenditures on income support for the unemployed per unemployed in thousands of US dollars at purchasing power parities, 1995 (source: OECD 2001). Unemployment rate. Average unemployment rate, 1992-1997.

Change in unemployment rate. Percentage change in unemployment rate, 1996-1997. Appendix B: Descriptive statistics Variables Mean Standard deviation Worry about job 1.824 0.964 Minimu m 1 1 Maximu m 4 5

Security of current 2.372 1.171 job Difficulty new job Education Unskilled labor Union member Public employee Sex Age Spouse work status Employment protection legislation Active labor market 5.620 3.846 policies Unemployment compensation Welfare spending Change unemployment rate in -0.9 2.033 state 23.769 5.881 8.923 6.715 0.476 0.499 40.478 12.519 0.529 0.499 2.127 0.967 0.381 0.486 sector 0.214 0.410 12.304 3.642 manual 0.236 0.423 of find 3.330 1.231

1 0

35 1

0 0

1 1

0 16 0 0.7

1 99 1 3.7

1.338

12.734

3.164

26.925

13.47

32.95

-4.6

2.0

Average unemployment rate Notes

8.533 4.605

3.1

22.5

1. These uncertainty perceptions are likely to vary across individuals for a number of reasons, including differences in personality. Our model sidesteps the influence of personality traits and assumes that individuals are reasonably well informed about the objective conditions in which they find themselves. 2. This model builds on OECD (1997). Note that we have data on individual perceptions of job-loss threat and replacement prospects, but exclusively objective measures of income insecurity. 3. While psychologists focus on individual perceptions of employability (Sverke & Hellgren 2002), we use the term 'employability' to denote objective characteristics such as age and education. 4. The concept of 'equivalent job' is by no means straightforward. Workers who lose their job commonly have to settle for replacement jobs that pay less, and the value of one's current job relative to alternative jobs find depends on considerations other than pay (benefits, job security, etc.). Given the data available, we must abstract from these ambiguities and sidestep the concerns over maintaining desirable job features versus having a job in the first place (see Greenhalgh & Rosenblatt 1984). 5. The effects of macro-economic demand management can be expected to operate through the rate of unemployment, which is included in our model. 6. Our two-step approach to explaining job insecurity does not involve a hierarchical set-up in the statistical sense of the word since our model of general job insecurity is independent of our models of job insecurity and labor market insecurity. This helps to avoid difficulties in estimating standard errors. As importantly, the second step of our analysis is designed to ascertain the effects of non-market sources of income support on affective job insecurity while controlling for cognitive job insecurity and labor market insecurity. Thus, it is not directly relevant how well our first-step models predict job and labor market insecurity. 7. Our data on government spending on active labor market programs in percentage of GDP come from the OECD Social Expenditure Database (OECD 2001). The OECD defines active labor

market policies as including spending on public employment services and administration, labor market training, programs for youth when in transition from school to work, programs to provide or promote employment for unemployed and other persons (excluding young and disabled persons), and programs for the disabled. 8. While the number of dependents should also be associated with more worry about job loss, the survey data do not allow us to test this proposition. Including the 'employed spouse/partner' variable means that we lose the Netherlands because of missing data. 9. Americans were less worried about losing their jobs than Swedes in 1997 (average Swedish job-loss worry was 1.81). Labor market conditions provide the most obvious explanation for this. The counterfactual estimates presented in Table 4 were estimated while holding all other variables constant. These estimates are based on Model 3 for job and labor market insecurity (Tables 1 and 2) and Model 4 for overall employment insecurity (see Table 3). References Ashford, S.J., Lee, C. & Bobko, P. (1989). Content, causes and the consequences of job insecurity: A theory-based measure and substantive test. Academy of Management Journal 32(4): 803-829. Bender, K.A. & Sloane, P.J. (1999). Trade union membership, tenure and the level of job insecurity. Applied Economics 31(1): 123-135. Borg, I. (1992). Uberlegungen und Untersuchungen zur Messung der subjektiven Unsicher-heit der Arbeitsstelle. Zeitschrift fur Arbeits- und Organisationspsychologie 36(3): 107116. Borg, I. & Elizur, D. (1992). Job insecurity: Correlates, moderators and measurement. International Journal of Manpower 13(2): 13-26. Bryk, A.S. & Raudenbush, S.W. (1992). Hierarchical linear models. Newbury Park, CA: Sage. Cameron, D. (1978). The expansion of the public economy: A comparative analysis. American Political Science Review 72(4): 1243-1261. Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Princeton, NJ: Princeton University Press. Esping-Andersen, G. (1999). Social foundations of post-industrial economies. London: Oxford University Press. Garrett, G. (1998). Partisan politics in the global economy. New York: Cambridge University Press.

Green, F., Felstead, A. & Burchell, B. (2000). Job insecurity and the difficulty of regaining employment: An empirical study of unemployment expectations. Oxford Bulletin of Economics and Statistics 62(4): 855-884. Greenhalgh, L. & Rosenblatt, Z. (1984). Job insecurity:Toward conceptual clarity. Academy ofManagement Review 9(3): 438-448. Hartley, J. et al. (1991). Job insecurity: Coping with jobs at risk. London: Sage. Iversen, T. & Cusack, T. (2000). The causes of welfare state expansion. World Politics 52(3): 313-349. Iversen, T. & Rosenbluth, F. (2003). The Political Economy of Gender. Paper presented at APSA, Philadelphia, PA. Iversen, T. & Soskice, D. (2001). An asset theory of social policy preferences. American Political Science Review 95(4): 875-893. Jacobson, D. (1991). Toward a theoretical distinction between the stress components of the job insecurity and job loss experiences. Research in the Sociology of Organizations 9: 1-19 Katzenstein, P. (1985). Small states in world markets. Ithaca, NY: Cornell University Press. Korpi, W. & Palme, J. (2003). New politics and class politics in the context of austerity and globalization: Welfare state regress in 18 countries. American Political Science Review 97(3): 425-446. Lim, V.K.G. (1996). Job insecurity and its outcomes: Moderating effects of work-based and nonwork-based social support. Human Relations 49(2): 171-194. Mughan, A. & Lacy, D. (2002). Economic performance, job insecurity and electoral choice. British Journal ofPolitical Science 32(4): 513-533. Naswall, K. & De Witte, H. (2003). Who feels insecure in Europe? Predicting job insecurity from background variables. Economic and Industrial Democracy 24(2): 189-216. Naswall, K. et al. (2002). Who feels job insecurity? What characterizes insecure workers in Europe? In M. Sverke et al. (eds), European unions in the wake of flexible production. Stockholm: National Institute for Working Life. Organisation for Economic Cooperation and Development (OECD). (1997). Employment outlook. Paris: OECD.

Organisation for Economic Cooperation and Development (OECD). (1999). Employment outlook. Paris: OECD. Organisation for Economic Cooperation and Development (OECD). (2001). OECD social expenditure database, 1980-1998. Paris: OECD. Rodrik, D. (1997). Has globalization gone too far? Washington, DC: Institute for International Economics. Scheve, K. & Slaughter, M. (2004). Economic insecurity and the globalization of production. American Journal of Political Science 48(4): 662-674. Sverke, M. & Hellgren, J. (2002). The nature of job insecurity: Understanding employment insecurity on the brink of a new millennium. Applied Psychology: An International Review 51(1): 23-42. Sverke, M., Hellgren, J. & Naswall, K. (2002). No security: A meta-analysis and review of job insecurity and its consequences. Journal ofOccupational Health Psychology 7(3): 242264. Swank, D. (2002). Global capital, political institutions and policy change in developed welfare states. New York: Cambridge University Press. Address for correspondence: Christopher J. Anderson, Department of Government, Cornell University, Ithaca, NY 14853, USA. E-mail: cja22@cornell.edu Research Paper Study on Staff Development and Employee Welfare Practice and their effect on productivity in five College Libraries in Charusat University, Gujarat Staff development and employee welfare are valuable assets in an organization since an organization's primary aims are productivity and profitability The study examines whether staff development policies exits in five libraries in Charusat University whether training programmes are being offered to increase staff competence, efficiencies and performance In conducting survey, Two sets of questionnaires were drawn up .one set was administered to management and the other set went to library staff . The survey revealed that all the organizations under study have staff development policies and training programmes for staff to enhance their capabilities and efficiency Key Words : Staff development, Employee welfare, productivity, College libraries, Charusat University

Staff Development is a process of raising productive potentialities of manpower resources in terms of knowledge, skills and capabilities through appropriate mechanisms such as education & training, counseling, career planning, performance or self-appraisals, awards or rewards etc. among all these, education and training play a vital role in strengthening both academic and professional capabilities of human resource to meet the goals of an organization. Every organization needs, committed and dedicated staff that will help it to meet it's tactical and strategic objectives, these objectives can only be achieved if the knowledge and skills of staff are upgraded through training and development .this will serve as a source of motivation to employee and drive them to contribute their quota towards the achievement of the organization's goal. All good things management require such efficiency, effectiveness, productivity, creativity can best be achieved when everybody works effectively as a team prayer and not as individuals. Gohil Devika G Research Scholar Singhania University, Rajasthan (India) Abstract Introduction This study focuses on staff development and employee welfare practices and their effect on productivity in five college libraries in Charusat University, these are the libraries of the Charotar Institute of Technology, Charotar Institute of Computer technology, Ramanbhai Patel College of Pharmacy, Charotar institute of applied science, Indubhai Ipkowala Institute of Technology. Literature Review Giroux (1989) studied on "A Blueprint for Human Resource Development in the Third Decade of the Ontario Colleges of Applied Arts and Technology: Executive Summary". It was for the assessment of its human resource efforts; that the system was initiated in order to identify improvement possibilities and areas in the light of the futuristic directions. In this regard a population was consulted which comprised of staff, faculty, management and members of the board of governors of educational institutes and tried to find the traits of their perception regarding HRD programs. Khosravi (2000) surveyed the opinions of librarians working in central University libraries in Iran about the problem of library stress. Fred Luthan's categorizations stressor is applied. A questionnaire with 36 questions using Likert Scale is used. 150 filled questionnaires were received. Results are analyzed using percentage and means. Luthan's five factors, causing the most

amount of stress in libraries are identified as a lack of library director's support, lack of job security, lack of library director's cooperation, lack of correct measures for encouragement, and lack of scientific methods of performance evaluation. Clough (1991) conducted a study on "Improving Staff Development in South Country Community College District, California". A study was conducted to investigate staff development needs at Chabot College and Las Positas College as perceived by members of the faculty, administration, and classified staff; to develop strategies to address the perceived needs; and to recommend short and long-term implementation plans. A Staff Development Needs Assessment Questionnaire, an eight-page instrument designed specifically for the study, was administered to the seven hundred and forty five full and part-time administrators, faculty and classified staff at colleges. The questionnaire asked about demographic information; opinions about the colleges organizational operating environment; willingness to participate in focused on meeting changing institutional needs, developing academic-and technical knowledge and skills, and personal development Methodology The questionnaire method was used to gather data, two sets of questionnaire were drawn up one set administered to management and the other sat went to library staff of the organization under study certain questions were peculiar to management staff and library staff only, In all 25 questionnaire were distributed to the management staff and out of the total number 23 were retrieved, this represented 80 percentage .he responses from management staff to analyze the responses. The writer again distributed 7 of the second questionnaire to the library staff. Out of the total of 7, 6 were retrieved representing 93% of the total. Responses from the library staff were analyzed alongside with that of the management staff. Questions were asked to elicit responses from management and library staff that there light on their opinion and attitude on staff development policy, training programmes, impact of training avenues to motivate staff and incentive packages. Findings Staff Development Policy When probed further to find out whether a budget was allocated for such programme, all the twenty five (25) respondent from management staff responded, yes .this was an indication of a healthy development in these organizations and it went to buttress several scholar's belief that the prime asset any manager of any organization for that matter, is that which relate to people. The

manager has many responsibilities, but two are especially high, those of ensuring that people are able to develop and undertake whatever is necessary to meet their objectives, and ensuring that people produce the desire performance. Staff Training When asked further about the frequency of the training, all the 25 respondents said as and when the need arises. This was also confirmed by all the six library staff Impact of training on staff All 25 management staff responded positively to the question .their responses included the following: 1. Workers are able to achieve better result 2. Training improves staff performance and increases output The same question was asked the library staff. All the six (6) respondents responded favorably .the following were some of their responses: 1. New ideas are discovered, wrong approach to work is corrected, and one's confidence 2. New knowledge is acquired, improved performance and higher productivity is achieved 3. One is exposed to new developments, and new skills for performance enhancement Motivational avenues When management was asked to indicate avenues available to motivate staff, All 25 respondents said salary is one of the means to motivate workers. This was also confirmed by the 6 library staff Conclusion Staff development and employee welfare practice have become major issues in management in this global economy where human resource management has received much organization, many employers are only interested in seeing their employees work without caring for their welfare and development. The study was carried out under the following assumption: The much attention is paid to employee's welfare The staff development and employee welfare have corresponding effect on productivity The organizations under study have staff development policies The survey further revealed that productivity cannot increase without training and staff motivation. They are inseparable for, one cannot do without the other Reference

1) Treasury of Polish Literature (1999), URL: http://www.literatura.zapis.net.pl/. 2) Virtual Library of Polish Literature (2000), URL: http://julia.univ.gda.pl/~literat/. 3) Virtual Scientific Library (2003), URL: http://vls.icm.edu.pl/. Please Cite this Article as : Gohil, Devika. "Study on Staff Development and Employee Welfare Practice and their effect on productivity in five College Libraries in Charusat University, Gujarat" Research Expo International 193-195. <www.researchjournals.in.>. Impact of Stress on Employees Job Performance A Study on Banking Sector of Pakistan Usman Bashir Faculty of Management Sciences, International Islamic University Islamabad, Pakistan E-mail: bbashir.usman@gmail.com Muhammad Ismail Ramay Associate Professor, Chairman FMS Faculty of Management Sciences, International Islamic University Islamabad, Pakistan E-mail: proframay@iiu.edu.pk Abstract Bankers are under a great deal of stress and due to many antecedents of stress such as Overload, Role ambiguity, Role conflict, Responsibility for people, Participation, Lack of feedback, Keeping up with rapid technological change. Being in an innovative role, Career development, Organizational structure and climate, and Recent episodic events. One of the affected outcomes of stress is on job performance. This study examines the relationship between job stress and job performance on bank employees of banking sector in Pakistan. The study tests the purpose model in relation of job stress and its impact on job performance by using (n=144) data of graduate, senior employees including managers and customers services officers of well reputed growing bank in Pakistan. The data obtained through questioners was analyzed by statistical test correlation and regression and reliabilities were also confirmed. The results are significant with negative correlation between job stress and job performances and shows that job stress signifincently reduce the performance of an individual. The results suggest that organization should facilitate supportive culture within the working atmosphere of the organization. Keywords: Job performance, Stress, Banks Multidisciplinary Research Journal 2.2 (2012) :

1. Introduction Over the past few decades stress is emerging as an increasing problem in organizations. Stress is vigorous state in which a person is confronted with an opportunity, demand, or resource related to what the individual wishes and for which the outcome is perceived to be both vague and vital. (Selye, 1936) first introduced the idea of stress in to the life science. He defined stress as the force, pressure, or tension subjected upon an individual who resists these forces and attempt to uphold its true state. Basically what is stress? The HSE (Health Safety Executive Uk) defines stress is an undesirable response people have to tremendous pressures or other types of demands placed upon them. It arises when they worry they cannot deal with. Some stress can be good, and some can be bad. HSE distinguishes between stress and pressure. Pressure is seen as positive and something that actually helps improve our performance. We all need a certain amount of pressure to perform well - ask any athlete, actor or actress. However, the problems arise when the sources of pressure become too frequent without time to recover, or when just one source of pressure is too great for us to cope with. Stress can be understood more comprehensively as, it is a condition which happens when one realizes the pressures on them, or the requirements of a situation, are wider than their recognition that they can handle. If these requirements are huge and continue for a longer period of time without any interval, mental, physical or behavioral problems may occur, (Health &Safety Executive UK). Stress has a positive effect on employees of any organization but up to a certain extent up to which an employee can cope with it, mostly it exceeds the bearable limits and have a negative result on employees. This is the base of the research study which has not yet been conducted in Pakistan. A lot of work has been conducted outside Pakistan. Number of studies has been conducted in different area of world but a huge gap exists in third world countries like Pakistan. The purpose of this study is to examine stress in employees of banking sector of Pakistan 2. Review of Literature Stress is an unwanted reaction people have to severe pressures or other types of demands placed upon them. A huge and multi fields literature points a lot of key factors such as work environment, management support, work load etc in determining the stressful the work can be and its effect on employee physical and mental health, (Ganster & Loghan, 2005). According to (Anderson, 2002) work to family conflicts is also a predecessor which creates stress in employees of an organization. Job stress has been also viewed as dysfunctional for organizations and their members (Kahn,

Wolfe, Quinn, Snoek, & Rosenthal, 1964). Although stress has been variously viewed as an environmental stimulus to an individual (Kahn et al., 1964) Selye, 1956 defined stress as an individual's reaction to an environmental force that effect an individual performance. Job related stress can be mostly immobilizing because of its possible threats to family functioning and individual performance. Job related stress can create an difference between demands on families and the ability of families to provide material security for them (McCubbin & Figley, 1983). While there is a significant body of research which deals with work and family there is relatively little research (e.g., Jacobson, 1987) which deals specifically with perceived job insecurity (i.e., concerns or fears about job loss) and marriage and family life (e.g., Buss & Redburn, 1983). Stress condition which happens when one realizes the pressures on them, or the requirements of a situation, are wider than their recognition that they can handle, if these requirements are huge and continue for a longer period of time with out any interval, mental, physical or behavior problems may occur, (Health &Safety Executive UK ). Stress exists in every organization either big or small the work places and organizations have become so much complex due to which it exists, work place stress has significant effects over the employees job performance, and the organizations in Uk are trying to cope with this scenario, (R. Anderson, 2003). Eleven forces are used as an antecedents of stress by researches (Overload, Role vagueness, Role conflict, Responsibility for people, Participation, Lack of feedback, Keeping up with quick technological change, Being in an innovative role, Career growth, Organizational structure and environment, and Recent episodic events.,) Overload :excessive work or work that is outside one's capability(Franch and Caplan ,1972; Margolis et al, 1974 ; Russek and Zohman, 1958) Role Ambiguity : Role insufficient information concerning powers, authority and duties to perform one's role (French and Caplan, 1972; Kahn, et al, 1964 ), Role Conflict: Supervisors or subordinates place contradictory demands on the individual(Beehr et al, 1976; Caplan and Jones, 1975; Caplan, et al, 1975; Hall and Gordon, 1973; Kahn et al, 1964) Responsibility for people: Responsibility for people, well-being works, job security, and professional development (French and Caplan, 1972; Pincherle, 1972) Participation: Extent to which one has influence over decisions relevant to one's job (Kasl, 1973) Margolis et al, 1974). Lack of Feedback: Lack of information about job performance (Adams, 1980 Cassel, 1974) Keeping up with rapid technological change: Keeping up with rapid changes in the information processing field (Ginzburg, 1967) Being in an innovative role: Having to bring about change in the organization (Kahn, et al.1964) Lawrence and

Lorsch 1970. Career development: Impact of status dissimilarity, lack of job security, let down ambition (Brook 1973) Erikson and Gunderson 1972; Kahn, et al. 1964) Recent episodic events: Certain life events, such as divorce and bereavement, that are highly stressful (Adams 1980 ; Cobb, 1977 Holmes and Rahe 1975). (Rose, 2003) In every organization and at every level of management and workers an elevated average level of stress is to be found which mostly has an effect on employee's job satisfaction. According to (Rose ,2003) employees have tendency towards high level of stress regarding time, working for longer hours which reduces employees urge for performing better. Management support helps in reducing or increases stress in employees, (Stamper & Johlke, 2003) apparent organizational assistance, management support work as a cushion which acts positively in decreasing work related stress in employees. There are a lot of reasons causing stress work family conflicts work over load one of reason indentified by (Stamper & Johlke ,2003) that if the organization or management does not appreciates its employees for their hard work or contribution toward the organization creates stress and mostly creates intention to leave. (Ivancevich & Donnelly, 1975) studied the link between anxiety stress with satisfaction and performance of employees, that lower anxiety stress improves performance of employee's which he studied in different managerial level of an organization. (Beehr, Jex, Stacy & Murray, 2000) found the relationship between occupational stressors an the performance of employees of an organization as well as it can affect the employees psychologically. (Jamal, 1984) studied a association between job stress and job performance between managers and blue-collar employees. Stress on job can be stated as the outcome of an individual due to the working environment from which he feels unsecured. Different relationships are projected between job stress and performance: U-shaped and curvilinear, positive linear, negative linear and no relationship between the stress and performance. A random sample of 305 blue-collar and 325 managerial workers in Canadian firm are surveyed through structured questionnaire. Variables used for this study were job stress, job performance, and organizational commitment. A negative linear relationship between job stress and job performance was found. Very limited evidence is seen for curvilinear or no association. 3. Theoretical Framework & Methodology The present study was conducted among employees of a well reputed growing bank in Pakistan who has three categories under which employees fall those are FTE (Fixed Tenure Employees) bank contract and third party contract employees. The data was collected by means of a structured

questionnaire with the help of supervisors and managers, copies of the questionnaire were given to respondents by hand. As the study is pointed towards employee stress and job performance, target sample source is well reputed growing bank in Pakistan, due to time and budget limitations convenient sampling technique is used to explore the relationship between stress and job performance of employees , my sample size is 144.The majority of respondents were male (87%) and married (79%). The average respondent was 39 years of age, having graduate and postgraduate qualifications. HYPOTHESIS: Job stress is negatively associated to job performance of employees 3.1 Job Related Stress 3.1.1 Description Four stressors work conflicts, resource inadequacy, work overload and work ambiguity this measure was designed by Khan, Wolfe, Quinn, and Snoek (1964). It described employee perception of job stress using 15 items asking about the incidence of stressful occurrences and the degree of role burden. This scale evaluate psychological indications of stress, such as feeling of over burdened with work, not having the adequate resources and tools to complete given tasks or projects, and usually being incapable to handle all of the work. In many studies, the items have been used to form subscales for role uncertainty, role overload, and resources insufficiency (Jamal, 1990; Shirom & Mayer, 1993). The scale is rated on five likert scale form 1=strongly disagree to 5= strongly agree. 3.1.2 Validity Job-related tension linked positively with troubles at work, work participation, work expectations, lack of mental support from a mentor, and family exception (Abraham & Hansson, 1996; Duxbury & Higgins, 1991: Seibert, 1999). Job tension linked negatively with quality of work life, job satisfaction, goal accomplishment, organizational commitment, self-esteem at work, quality of family life, and life satisfaction (Abraham & Hansson, 1996; Duxbury & Higgins, 1991; Seibert, 1991). In Jamal (1990), the job tension subscales for role ambiguity, role overload, and resource inadequacy all correlated positively with psychosomatic problems. All the subscales except role overload correlated negatively with job satisfaction. In Shirom and Mayer (1993), role overload correlated positively with parent teacher conflict, teacher conflict, teacher-principal conflict, role overload correlated positively with parent teacher conflict, teacher-principal conflict, and work-home conflict.

3.2 Job Performance (In Role Behavior) This measure was developed by Williams and Anderson (1991). It uses 21 items to describe three types of organizational citizenship behaviors. These are behaviors directed at specific individuals (OCBI), Behaviors directed at an organization (OCBO), an employees in-role behaviors (IRB). In this paper only the in role behavior scale having five items was used as mentioned in appendix A the scale is rated on five likert scale form 1=strongly disagree to 5= strongly agree The questionnaire was exclusively filled by the line managers and supervisors of the employees and was kept personal form the employees and no information was leaked 4. Results & Discussion The reliabilities of both the dependent variable and independent variable are shown in table 1(a) and table (b) which are 0.694 and .637 respectively, in table the relationship between job stress and job performance is negatively correlated at significance level of P .01 that is correlation is significant. In table 2 (a) and 2 (b) as the value of beta is -.527 that shows 52.7 means that for every one percent increase in job stress will have an effect of 52.7 effect on job performance which is negatively correlated to job stress. In other words 52.7 of job performance is being effected by job stress and the remaining by other factors. Regression analysis shows that relationship between job stress and job performance is proved the value of R square .278 it shows that the impact of job stress on job performance is 27.8% is explained. The purpose of this study was to find out the relationship between the job stress and job performance of employees of banking sector in Pakistan. As per hypothesis job stress had a negative relation with job performance that when stress occurs it effects the performance of employees negatively, (Ivancevich & Donnelly, 1975) that lower the stress it increases the performance so both these are inversely proportional each other as to the results the correlation in table 2 shows both job stress and job performance is negatively correlated. Management support helps in reducing or increases stress in employees, (Stamper & Johlke ,2003) apparent organizational assistance, management support work as a cushion which acts positively in decreasing work related stress in employees. (Rose, 2003) studied the results that the stress in work environment reduces the intention of employees to perform better in jobs with the increasing level of stress the employees thinking demoralize and his tendency to work well also decreases. No doubt stress is necessary for increasing performance of employees but up to a certain level. In this study the employees do their job regularly but due to workloads and time constraints their performance reduces.

Banks timing is mostly from 9am to 5pm in Pakistan but originally there is no time limit so employees have to work for longer hours as compared to other jobs which is also a reason of concern. The study was conducted only in industry that was banking sector and the impact job stress on job performance was measured only in one sector, if we want to generalize the results of this study then this study should be replicated in other sectors of Pakistan as well. References Adams, J.D. (1980). Improving Stress Management: An Action-Research-Based OD Intervention, I Under-standing and Managing Stress: A Book of Readings. University Associates, San Diego, California, pp. 179-198. Anderson E.S., Coffey S.B., & Byerly T.R. (2002). Formal Organizational Initiatives and Informal Workplace Practices: Links to Work-Family Conflict and Job-Related Outcomes. Journal of Management 28,787. Anderson R. (2003). Stress at work: the current perspective. The Journal of The Royal Society for the Promotion of Health, 123; 81 Beehr A. T, Jex M.S., Stacy A. B., & Murray A.M. (2000). Work Stressors and Coworker Support as Predictors of Individual Strain and Job Performance. Journal of Organizational Behavior, Vol. 21, No. 4 , pp. 391-405. Beehr, T.A. (1976). Perceived Situational Moderators of the Relationship Between Subjective Role Ambiguity and Role Strain. Journal of Applied Psychology, Vol 61, pp. 35-40. Brook, A. (1973). Mental Stress at Work. The Practitioner, Vol 210, pp. 500-506. Caplan, R.D., and Jones, K.W. (1975). Effects of Work Load, Role Ambiguity, and Type A Personality on Anxiety, Depression, and Heart Rate. Journal of Applied Psychology, Volume pp. 713-719. Cobb, S., and Kasl, S.V. Termination. (1977). The Consequences of Job Loss, U.S. Government Printing Office. Washington, D.C. Erickson, J., Pugh, W.M., and Gunderson, E.K.E. (1972). Status Congruency as a Predictor of Job Satisfaction and Life Stress. Journal of Applied Psychology, Vol 56, pp. 523-525. French, J.R.P., Jr., and Caplan, R.D. (1972). Organizational Stress and Individual Strain. in A.J. Marrow, ed., The Failure of Success, AMACOM, New York, New York, .

Hans S. (1936). A syndrome produced by diverse noxious agents. Nature, 138:32. Health and Safety Executive. (2001). Help on Work-related Stress: A Short Guide. INDG281 Rev1 2001.Sudbury: HSE Books. Health and Safety. (2001). Executive. Tackling Work-related Stress: A Guide for Employees. INDG341.Sudbury: HSE Books, Ivancevich M.J., & Donnelly H. J. (1975). Relation of Organizational Structure to Job Satisfaction, Anxiety-Stress, and Performance. Administrative Science Quarterly, Vol. 20, No. 2 , pp. 272-280. Jamal M. (1984). Job Stress and job Performance controversy: an empirical assessment in two countries. Organizational Behavior and Human Performance, 33:1-21. Kahn, R. L., Wolfe, D. M., Quinn, R. P., Snoek, J. D., & Rosenthal, R. A. (1964). Organizational stress: Studies in role conflict and ambiguity. New York: Wiley Kasl, S.V. (1973). Mental Health and the Work Environment. Journal of Occupational Medicine, Vol 15, pp. 509-518. Lawrence, P., & Lorsch, J. (1967). Organization and Environment: Managing Differentiation and Integration. Harvard University Press, Cambridge, Massachusetts . Margolis, B.L., Kroes, W.H., & Quinn, R.P. (1974). Job Stress: An Unlisted Occupational Hazard. Journal of Occupational Medicine, Vol, pp. 659-661. McCubbin, H. I., & Figley, C. R. (Eds.). (1983). Coping with normative transitions (Vol. 1). New York: Brunner/ Mazel. Pincherle, G. (1972). Assessment of the Relation-ship Between Stress and Work Performance. Proceeding of the Royal Society of Medicine, Vol, pp. 321-324. Rose M. (2003). Good Deal, Bad Deal? Job Satisfaction in Occupations. Work Employment Society, 17; 503. Stamper L.C., & Johlke C.M. (2003). The Impact of Perceived Organizational Support on the Relationship Between Boundary Spanner Role Stress and Work Outcomes. Journal of Management, 29; 569.

The effects of job characteristics and working conditions on job performance Emin Kahya*

Department of Industrial Engineering, Eskisehir Osmangazi University, Bademlik Campus, Eskisehir, Turkey Received 26 September 2006; received in revised form 3 January 2007; accepted 8 February 2007 Available online 26 March 2007 Abstract In performance evaluation literature, although the combination of some variables such as age, gender, experience, observation time, and interpersonal affect has been widely considered in determining employee performance, no investigation has indicated the influence of workplace conditions on job performance. This study reports the effects of job characteristics (physical efforts and job grade), and working conditions (environmental conditions and hazards) in addition to experience and education level on task performance and contextual performance. A total of 154 employees in 18 teams at a medium-sized metal company participated in this study. Seven criteria for task performance and 16 for contextual performance were used for measuring employee performance. The results showed that there were substantial relationships between employee performance both job grade and environmental conditions. Poor workplace conditions (physical efforts, environmental conditions, and hazards) result in decreasing employee performance consisted of following organization rules, quality, cooperating with coworkers to solve task problems, concentrating the tasks, creativity, and absenteeism. Relevance to industry Unpleasant working conditions in workshops have different effects on each of the job performance indicators. This study highlighted that training program designed to enhance job performance of the employees working under poor workplace conditions should focus on organizational rules in terms of occupational health and safety. 2007 Elsevier B.V. All rights reserved. Keywords: Task performance; Contextual performance; Experience; Education level; Job grade; Working conditions; Environmental conditions; Hazards 1. Introduction Perhaps, the most important dependent variable in industrial and organizational psychology is job performance. For all of the main applications of this branch of psychology, such as employee training and job redesigning, the focus is almost always on improving job performance (Borman, 2004). Borman and Motowidlo (1993) identified two broad classes of employee behavior: task performance and contextual performance. Both types of behavior are presumed to contribute to

organizational effectiveness, but in different ways (Kiker and Motowidlo, 1999). Task performance involves patterns of behaviors that are directly involved in producing goods or service or activities that provide indirect support for the organization's core technical processes. Such criteria including quantity, and quality of output were widely used task performance criteria to measure employee job performance in the ergonomic studies. Contextual performance is defined as individual efforts that are not directly related to their main task function but are important because they shape the organizational, social, and psychological context that serves as the critical catalyst for task activities and processes (Werner, 2000). When employees help others complete a task, cooperative with their supervisors, or suggest ways to improve organizational processes, they are engaging in contextual performance (Van Scotter et al., 2000). As interest grows in the type of helpful, cooperative, and innovative job performance behavior, it becomes more important to understand its influence on organizational and individual outcomes. In performance evaluation literature, many systematic studies have been devoted to the potential effects of some variables such as age, gender (Lee and Alvares, 1977), experience (Schmidt et al., 1986), observation time (Moser et al., 1999), interpersonal affect (Antonioni and Park, 2001), rating format (Yun et al., 2005), workplace deviant behaviors (Dunlop and Lee, 2004), and organizational politics (Witt et al., 2002; Miron et al., 2004) on job performance. The major result from these studies was that job experience and education level had direct or indirect effects on job performance. Several authors (e.g., Schmidt et al., 1986; Moser et al., 1999; Posthuma, 2000) have reported a mean correlation of 0.09-0.18 between experience and job performance. Schmidt et al. (1986) concluded that job experience leads to the acquisition of skills, techniques, method, psychomotor habits, etc., that directly produce improvements in performance capabilities. In Borman, Motowidlo and their colleagues' studies (e.g., Borman et al., 1995; Borman and Motowidlo, 1993; Motowidlo et al., 1997; Motowidlo and Van Scotter, 1994; Van Scotter and Motowidlo, 1996), experience was significantly correlated with task performance (r 0.30-0.40) and contextual performance (r~0.15). In contrast, Ferris et al.'s (2001) findings, r 0.02 for task performance and r 0.23-0.25 for contextual performance, did not support the expected result. These variations indicate that the relative importance of experience may depend on complexity of job family. It may be that, for jobs of greater complexity, an increase in job experience results in higher job knowledge and task performance. The opposite may be true for jobs of lower complexity.

There is a substantial body of research from both sociology and economics that has revealed the relationship between education, productivity, job satisfaction, and salary (e.g., Groeneveld and Hartog, 2004; Voon and Miller, 2005). Education level required for a job in the workforce is a persistent problem in all industrialized countries. It is expected that the higher the level of education the more productive the employee, wherever he or she is employed. Posthuma's (2000) finding, that education level is positively associated with supervisor evaluations of job performance (r 0.24), supported this hypothesis. A major effect of education is that overall productivity falls short. Education does not guarantee increased productivity. Indeed, it seems likely that over-educated employees would be more prone to morale problem, not to be rewarded with higher salary. Higher dissatisfaction generates lower work effort, which reduces employee productivity. Several authors such as Ferris et al. (2001) have found a negative correlation for task performance (e.g., r 0.04) and overall performance (e.g., r 0.13). Despite the paucity of research literature, personnel psychologists appear to have been aware that education could impact job performance. From this expectation, employers tend to hire overeducated employees. One of the major concerns of manufacturing companies has focused on improving worker productivity, which is one of the job performance measures. Some of the common features of these companies are heavy loadings, adverse environment, poor human-machine system design, unpleasant working conditions, etc. Heavy loadings, workplace conditions such as inclement weather, extreme heat/ cold, chemical smell, noise, poor lighting, vibration, and dust have direct or indirect effects on employee job performance. These conditions decrease employee concentration towards tasks which lead to low employee performance such as low productivity, poor quality, physical and emotional stress, which cause high cost. Effective applications of ergonomics in working conditions enhance employee job performance, provide worker safety, physical well-being, and job satisfaction. Many studies in ergonomic area (e.g., Das and Shikdar, 1999; Resnik and Zanotti, 1997; Shikdar and Sawaqed, 2003; Yeow and Sen, 2006) have focused on the positive effects on quality, productivity, hazards, occupational health, and their cost effectiveness of ergonomic improvements in a workstation or workshop at a manufacturing company. There has been a lack of systematic empirical research designed to investigate the relationship between job performance, or several aspects of job performance and working conditions. To our knowledge, as one of the studies on this topic, Shikdar and Sawaqed (2003) investigated ergonomic factors which lead to low worker productivity in six different industries by applying questionnaires to 50

production managers. In this study, measures related to performance were productivity, quality, and absenteeism, and environmental factors including heat, humidity, noise, light, dust, and pollution. From the statistical analysis, they concluded that there was a significant correlation (r 0.234) between performance indicators and environmental factors. It indicates that companies with higher environmental problems had more performance related problems such as low productivity, high absenteeism. In the ergonomic literature, no research has examined the relationship between job characteristics, working conditions and employee job performance. The current research is the first attempt to shed some lights on the issue of which job characteristics (job grade and physical effort) and working conditions (environmental conditions and hazards) play a more important role in influencing any job performance measure. 2. Relationship between job characteristics, working conditions and job performance In job performance literature, although age, gender, experience, observation time, and interpersonal affect have been considered in many studies, no research has been devoted to the potential effects of job characteristics and working conditions on task and contextual performance. Some jobs in the workshops of mechanical processing, machining, and maintenance are high level jobs of complexity. Job type, job level, and job context create different influences on job performance. Some jobs are required high level skill and responsibility to perform tasks successfully. It is plausible that those employees who carry out these jobs should undertake higher level of job performance to fulfill the job responsibilities satisfactorily. Blue-collar employees working in manufacturing companies exert different levels, frequencies, and durations of the physical effort during performing the tasks of their jobs. They use strength of kneeling, crouching/crawling, walking, standing, balancing, lifting, and pulling/pushing objects. The physical effort requirement for a job changes from a combination of sitting, standing, and working with little requirement to an intense requirement for lifting objects weighing over 30 Kg with no mechanical aid. An increase in the level of physical effort was accompanied by an increase in energy expenditure. The majority of the energy consumption generally is converted to the waste activities such as static effort (e.g., lifting, pulling/pushing objects), to other awkward postures, or to inefficient equipment or method. These waste activities cause decreasing productivity. Some jobs in the workshops such as mechanical processing, painting, maintenance are performed in unpleasant working conditions. In job evaluation literature, working conditions imply two dimensions: environmental conditions and hazards. Environmental conditions range from ordinary

to extreme conditions in terms of the factors such as heat, humidity, noise, smell, light, and dust. Unpleasant environmental conditions have both direct and indirect effects on employee job performance. The concentration to tasks of an employee who exposes to these impacts decreases, which leads to low employee performance including productivity, quality, emotional stress, and in turn this causes high cost. Hazards are unavoidable direct or indirect exposure to light wound/scald, flammable danger, electrical hazards, occupational disease, and mortal hazards. It is believed that ergonomic deficiencies are the root causes of workplace health hazards, low level of safety (Shikdar and Sawaqed, 2003). The application of the relevant human factor principles can reduce the likelihood of accidents and injuries. These impacts reduce worker productivity, and cause high absenteeism. 3. Method 3.1. Subjects Our research was conducted in a medium-sized manufacturing company, which produces various types of tractor cabin. The company was founded approximately 10 years ago, and has grown to its current size of over 200 employees for all the jobs. This is an important context for examining that not enough research in field settings has focused on blue-collar workers. Fourteen immediate supervisors participated in this study as raters; each was a first-line manager of one or two work teams. The number of workers in a team is ranging from 4 to 29, with an average of 10.21 workers per team. All supervisors participated in several one-day workplace training courses. These courses covered principles of assessment such as reliability, fairness, halo errors in ratings, definitions of the items used to measure employee performance, and the assessment procedure. A total of 154 blue-collar employees working in the departments including stock control (one team, four workers), mechanical processing (7 teams, 98 workers), assembly (4 teams, 23 workers), maintenance (3 teams, 9 workers), R&D (one team, 7 workers), and quality (2 teams, 13 workers) participated in this study as part of a larger HRM project. Several authors (e.g., Rothstein, 1990) have argued that new employees should be evaluated only after a certain time has passed because supervisors need adequate opportunities to observe subordinates' behaviors. In this study, the employees were those who are working in the company much longer than four months so that they can be assessed fairly by their own supervisors. Job experience (number of years that employee is working in the company) and education level of each ratee were obtained from the

company's archival sources. The average length was 8.61 (s.d. 5.52) years for experience. Among them, the majority (91) had graduated from high schools or below. Education level was a grouped measure that was coded from 1 (high school or below) to 4 (junior college two years university). 3.2. Job analyses A job analysis questionnaire was designed to obtain current job information under 14 job evaluation factors adopted from Metal Industry Job Grading System (MIJGS) (1996) for the purpose of determining job characteristics and working conditions. Forty-nine blue-collar jobs whose base duties, responsibilities, skills, personal attributes necessary for successful execution, and working conditions were different from each other, were analyzed in 18 teams. The questionnaires were distributed to less than five employees for each identified job. Employees were asked to fill them out by checking an appropriate alternative or answering the questions as far as possible, seeking assistance from their supervisors. A total of 129 questionnaires were returned, with a response rate of 83.77%. Job grade: The questionnaires for each job were evaluated to score, and the results were checked with the same or similar jobs in MIJGS (Metal Industry Job Grading System) (1996) for consistency. It was found that job scores varied from 204.50 to 645.50 points. Job score was graded as 1 for 250 points and below, 2 for 251-300 points, 9 for 601 points and above. Greater job score and grade reflects higher level of complexity in terms of job knowledge, responsibility, ability, and effort requirements. Physical effort: Eight factors were identified for physical effort. Two (lifting and pulling/pushing) had strength levels ranging from (1) 5 Kg and below to (5) 30 Kg and above. The others were frequencies of kneeling, crouching/ crawling, walking, standing, sitting, balancing items ranging from (1) never to (5) continuous. Physical effort was grouped into five levels; each level consisted of a combination of these factors. Levels ranged from 1 ''A combination of sitting (low), standing and walking with an ongoing requirement for physical effort. There may be a requirement to exert light physical effort (lifting, pushing, pulling objects below 5 Kg) for short periods.'' to 5 There is an occasional requirement to exert intense physical effort (lifting, pushing, pulling objects weighing over 30 Kg with no mechanical aids) for long periods during a shift.'' Environmental conditions;: Seven factors were identified for environmental conditions. Five were conditions including fumes/bad odors, dust, illumination, dirt, and humidity scored from (1) never

to (5) continuous. Noise consisted of three levels: (1) low, (2) 75 dBA and below, (3) higher than 75 dBA. Temperature had three levels: (1) about 20 C, (2) 22-30 C (or cold), (3) 30 C and above. In job evaluation, environmental conditions are sub-divided into levels ranging from 5 to 6. In this study, they were grouped from 1 Exposure to environmental conditions is rare.'' to 5 Continuous exposure to highly unpleasant environmental conditions'', where each level consisted of a combination of these factors. Hazards: Twelve factors such as light wound/scald, flammable danger, electrical hazards, occupational disease, and mortal hazards were identified for hazards consisted of four levels ranging from (1) sometimes to (4) continuous. Hazards were grouped into five levels ranging from 1 Exposure to hazards is rare'' to 5 Exposure to mortal hazards''. 3.3. Job performance measures Task performance: Task performance was measured using seven criteria. Criterion development was based on the empirical work of Motowidlo and Van Scotter (1994) and the results of our job analysis applied for identifying important tasks and behaviors. In a meeting, the evaluation committee consisted of department directors reviewed the criteria for appropriateness and then came to an agreement for seven criteria (see Table 3). The last four criteria in the table came from Motowidlo and Van Scotter's (1994) empirical work, and the othersjob knowledge (measure of the knowledge required to get the job done), problem solving, and concentrating to duties were the results of the job analysis. Each employee was rated by his supervisor on a five-point Likert scale ranging from (1) inadequate to (5) excellent. Contextual performance: Twenty-five contextual performance criteria were generated from previous studies reporting contextual performance criteria (e.g., Motowidlo and Van Scotter, 1994; Goodman and Svyantek, 1999; Coleman and Borman, 2000) and tools implemented by several companies in the city. The appropriateness of each criterion was discussed in a meeting. The evaluation committee accepted 16 criteria which were important for the company's vision and values. As a result, the criteria were adopted from Coleman and Borman (2000) (six items), Motowidlo and Van Scotter (1994) (three items), and the tools in use (seven items). Coleman and Borman's (2000) categorization provided a useful model (a three group model) for classifying the criteria. Supervisors used a five-point scale ranging from 1 ''fails to meet expectations'' to 5 clearly and consistently exceeds expectations'' to rate their employees' performance.

A performance evaluation questionnaire consisting of the criteria to assess employee performance, employee name, and job information (job identification, job grade, team, and department) was designed for each of two separate tools. Ratings were made at the end of a six-month period of observation. One questionnaire per employee was distributed to the supervisors at work. They were asked to fill out the questionnaires and give them back to HRM department in a week. It was guaranteed that their responses would remain completely confidential. All of the supervisors completed the performance rating questionnaires for all their employees. 4. Results 4.1. Correlations! among the variables! Table 1 reveals descriptive statistics (means and standard deviations) and Pearson correlation coefficients of the variables. There are several interesting findings that should be noted. Job grade was strongly correlated with task performance (r 0.456, /?<0.01) and contextual performance (r 0.411, jp <0.01). It is worthwhile that one of the most positive correlations among the variables was obtained with job grade, which exposes the contribution of this study. Supervisors indicated that employees who work at the qualified jobs had higher performance than the others. There were significant but small positive correlations (not high enough to indicate empirical redundancy) between job grade and the other study predictors (r 0.191, 0.203, and 0.237). Second, environmental conditions were significantly correlated with task performance (r 0.332, p< 0.01), not related to contextual performance (r 0.058). In other words, unpleasant environmental conditions cause decreasing the task performance scores of the workers. Surprisingly, it was found that there was a second significant (negative) interaction between environmental conditions and employee's education level (r 0.461, jp<0.01). In general, those employees who were less educated perform the jobs in poor working conditions. Thirdly, there was the highest correlation coefficient (r 0.622, /p<0.001) between environmental conditions Table 1 Descriptive statistics and Pearson correlation coefficients Mean Std. dev. 1. Task performance 3.20 2. Contextual 3.08 0.36 0.46 0.535*** 1 2 3 4 5 6 7

performance 3. Experience 4. Education level 5. Job grade 6. Physical effort 7. 8.61 1.53 5.08 2.95 5.52 0.78 1.81 0.98 1.31 0.166* -0.254** 0.456** 0.043 0.332** 0.057 0.063 -0.428** 0.073 0.622** -

0.411** 0.038 -0.160* 0.058 0.175* 0.139

-0.369** 0.191* -0.461** 0.203*

Environmental 2.51

conditions 8. Hazards 2.53 0.83 -0.119 -0.029 -0.036 0.205* 0.237** 0.324** -0.07 2 Notes:*p<0.05 (2-tailed), **p<0.01 (2-tailed). Table 2 Hierarchical multiple regression analyses Independent variables Step 1 (a) Task performance Experience Education Job grade Physical effort Environmental conditions Hazards R2 (adjusted) DR2 (b) Contextual 0.06 0.28 + 0.22 0.30 +0.02 0.36 +0.06 -0.058 0.36 +0.00 0.070 -0.224** 0.030 -0.276*** 0.475*** 0.030 0.349*** 0.517*** -0.189** 0.062 -0.236** 0.470*** -0.357*** 0.341*** 0.063 -0.222** 0.480*** -0.321*** 0.320*** Step 2 Step 3 Step 4 Step 5

performance Experience Education Job grade Physical effort Environmental conditions 0.102 0.107 0.068 0.063 0.404*** 0.069 -0.043 0.465*** -0.277*** 0.088 0.024 0.437*** -0.376*** 0.203* 0.088 0.023 0.436*** -0.378*** 0.204*

Hazards R2 (adjusted) DR2

0.00

0.16 +0.16

0.22 +0.04

0.23 +0.01

0.003 0.26 +0.03

Notes: Beta values for the models. *p<0.05, **p<0.01, ***p<0.001. and physical effort. The general conclusion from this result is that the worse the working conditions, the higher the physical effort. Employees who expose to highly unpleasant environmental conditions also perform their jobs with intense requirement for physical effort. 4.2. Multiple regression analyses Multiple regression analyses were conducted to reveal the relative contribution of each variable to the prediction of job performance. To control the incremental effects of the proposed variables (job grade, physical effort, working conditions, and hazards) in the regression analysis, five hierarchical regression models were established. In the base model (Step 1), job (task or contextual) performance was dependent variable, and two variables (experience year and education level) were a fixed block as independent variables. For the next models, each variable was extended to the previous model; job grade in Step 2, physical effort in Step 3, environmental conditions in Step 4, and hazards in Step 5. Such an approach shows the effect of the extended variable on job performance. Table 2 reports the standardized coefficient (b), the amount of variation in the dependent (predicted) variable by the model (R2), and the change in R2 (DR2) for task performance (Table 2a), and contextual performance (Table 2b). The first column in Table 2a (Step 1) shows that only education level (b = 0.24, jp < 0.01), significantly but negatively, explained the variation in the ratings of task performance. As some authors have depicted, increase in the education level results in lower score in task performance. Experience was not a significant predictor. Nevertheless, all the variables accounted for 6% of the total variation of ratings (R2 = 0.06) which is quite small. In all the analyses (Steps 1-5), the variables had a similar pattern. When job grade was entered into regression equation (Step 2), the change in R2 was 22%. As predicted, job grade accounted for a significant incremental portion of R2 in predicting task performance after controlling the effects of the variables in the base model. In support of our

Fig. 1. Interactions between the study variables. (a) Physical effort and job grade. Note: The levels for job grade imply that low: 1-3 grades; average; 4-6 grades; high: 7-9 grades. The levels for physical effort imply that low: levels 1 and 2; average: level 3; high: level 4. (b). Hazards and environmental conditions. Note: The levels for environmental conditions and hazards imply that low: levels 1 and 2; average: level 3; high: levels 4 and 5. High hazards levels were found for both average and high environmental conditions. predictions, job grade was a statistically significant predictor of task performance (b 0.475, /?<0.001). In Model 3 (including physical effort), AR2 was 2%. Neither physical effort nor hazards had an important effect on task performance. The results of the Model 5 indicate that the most significant predictors were job grade (b 0.480, /p<0.001) and environmental conditions (b 0.320, p< 0.001). The results in Table 2b show that the variables except for job grade had no significant effects on contextual performance. According to Model 2, the amount of the variation explained by job grade was quite important (AR2 0.16). As a result, job grade and environmental conditions (less) were better predictors on task and contextual performance ratings. To identify the form of the interactions between job performance and the study variables, we plotted the prediction of job performance scores at the means as well as at high and low levels of

the variables. As shown in Fig. 1, which illustrates the interaction on task performance, we found that there were significant but small interaction effects of physical effort and hazards on task performance. Those employees who perform qualified jobs under light physical conditions reached to higher level task performance than the others. This implies that physical conditions have a substantial effect on qualified jobs; getting the worst physical effort produces a larger decrease in task performance. In a poor physical effort, employees with any level of job grade reached to the same level of task performance (see Fig. 1a). We also considered the interaction from the standpoint of the relationship between task performance and both environmental conditions and hazards. Hazards demonstrated at least a low decreasing effect on task performance among workers under high and low environmental conditions 4.3. The effects of variables on job performance criteria An additional correlation analysis was performed to reveal the contribution of each variable to various measures of job performance. The results of this analysis (Table 3) provided a general support that each variable had a different contribution to each performance measure. Surprisingly, it was found that the variables had significantly greater effects on contextual performance criteria. The substantial performance criteria were ''Engaging in self-development to improve own effectiveness'', and ''Creativity to solve a work problem'' for job grade, ''Quality'', and ''Cooperating with others to solve problems'' for physical effort, ''Following organization rules and proper procedures'', and ''Cooperating with others to solve problems'' for environmental conditions, and ''Following organization rules and proper procedures'', and ''Concentrating to the duties'' for hazards. 5. Discussion In performance evaluation literature, although many systematic studies have been devoted to the potential effects of some variables such as age, gender, experience, observation time, interpersonal affect, rating format, workplace deviant behaviors, and organizational politics, to our knowledge, this is the first investigation to indicate the influence of job characteristics and workplace conditions on job performance. Some jobs in the manufacturing and maintenance departments differ from the traditional jobs. They are performed under the working conditions ranging from rarely to extremely unpleasant in terms of heat, cold, smells, noise, humidity, etc. These conditions influence employee to demonstrate low level job performance than the others who work under

better conditions. In this study, we intended to explain the effects of job characteristics and working conditions on task and contextual performance. The argument for distinguishing between task and contextual performance gains force if they are correlated with different demographic characteristics. Borman and Motowidlo (1993) suggested that the major source of variation in task performance is the proficiency with which a person can carry out task activities. This means that individual differences in knowledge, skills, and abilities Table 3 Pearson correlation coefficients between the variables and criteria Dimension Criteria Physical effort Task Job knowledge 0.088 0.003 Environmental conditions 0.299 Hazard s 0.12 9 performanc Overcoming obstacles to complete a task e Problem solving (ability to solve problem quickly 0.072 and correctly) Operating equipment, using tools, or both Working safely Concentrating to the duties 0.115 0.066 0.052 0.089 0.176 0.358 0.123 0.168 0.303 0.182 0.256 0.113 0.072 0.034 0.084 0.15 4 Protecting the resources 0.052 0.152 0.360 0.052 0.09 8 Interperson Assisting co-workers with personal matters al Citizenship Cooperating with others to solve problems 0.057 0.073 0.00 2 0.090 0.258 0.15 3 Engaging responsibly in meetings and group 0.202 activities Organizatio Treatment the supervisor with respect nal Citizenship Absenteeism 0.171 0.063 0.092

Working systematically Following procedures Participating responsibility in the organization organization rules and

0.104 proper 0.141 0.089 0.087 0.164 0.363 0.007 0.277

0.105 0.463

0.113 0.20 5 0.04 9 0.086 0.00 8 0.082 0.044

0.084

Completing a task on time

0.078

Job dedication

Attention to important details

0.032 0.166

Quality

Productivity

0.141 0.184 0.054

Creativity to solve a work problem

0.126 0.00 7 0.03 5

Engaging in self-development to improve own 0.095 effectiveness Generating new ideas to make things (tasks) 0.202 better (innovation) Planning and organizing work 0.070

0.098 0.067

0.039

should covary more with task performance than with contextual performance. Experience may have a direct or indirect impact on job performance. Increase in experience may produce increase in job knowledge, which leads to improved job performance. In Van Scotter and colleagues' studies, experience has significantly correlated with task performance (r 0.30-0.40) and contextual performance (r~0.15). In contrast, Ferris et al.'s (2001) findings, r 0.02 for task performance and r 0.23-0.25 for contextual performance, did not support the expected result. Our results (r 0.17 for task performance, r 0.06 for contextual performance) differed from their findings; they did not support the above sight but overlapped too much with Moser et al.'s (1999) results. However, experienced employees may, generally, get difficulty in adjusting to social or new situations or engaging in self-development to improve own performance and also do not assist and cooperative with others. The correlation between job performance and education level did not yield the predicted pattern of result. Education was negatively correlated with task

performance (r 0.25, jp<0.01), but positively and weakly correlated with contextual performance (r 0.06); thus, task performance may diminish somewhat with increasing education level, which has been concluded by some authors as well. Higher education level may not guarantee higher job performance. The finding showed that job grade was strongly correlated with task and contextual performance. As expected, job grade was more strongly associated with task performance (r 0.456) than with contextual performance (0.411). Job grade accounted for a significant incremental portion of the variation in predicting task performance (DR2 0.22) and contextual performance (DR2 0.16). It indicates that, generally, those employees who perform the qualified jobs had higher levels of job performance. It is a foresight that while employees who had higher performance work at high graded jobs, the others in the same team perform low graded jobs. For example, although the employees carrying out the ''Heavy welding'' scored as 524 points (grade 7) were rated as 3.57, the others working at the ''Light welding'' job scored as 458 points (grade 6) were rated as 3.14. When job gets difficult, employees who are perceived as being good at task performance will also achieve the highest rating on contextual performance. This pattern shows that job performance differs depending on job difficulty level. As shown in Fig. 1a, employees who perform high graded jobs under a requirement to exert light physical effort reached to higher level task performance than the others; physical conditions had a substantial effect on qualified jobs. Getting the worst physical effort produced a larger decrease (from 3.65 to 3.29 of 5.00) in task performance. This implies that ergonomic improvements focused on physical effort will contribute to task performance, when they are conducted on the jobs with high level grade. In an intense physical effort, employees with any level of job grade reached to the same level of task performance. In a word, physical effort contributes less to task performance while a job has a low or average level grade. One of the major contributions of the present study is to report the effect of working conditions on job performance. Some jobs in the workshops such as welding and painting are performed in such environmental conditions as extreme cold or heat, excessive noise (above 90 dBA), chemical smell. It is expected that these environmental conditions can show a negative effect on job performance. The findings showed that environmental conditions had a second important effect on job performance. It accounted for approximately 6% of the total variation.

The results of this research have practical implications on employees and organizations. It may be one of the first empirical studies in the related literature to investigate the point effects of workplace conditions on job performance measures. An implication for ergonomic practice may be to implement workplace improvements designed to enhance employee job performance on areas where a factor is poor. The correlation coefficients provided a general support that each variables had a different contribution on each job performance indicator as depicted in Table 3. For three relevant variables, the most important job performance criteria will be discussed below. Physical effort: Blue-collar employees working in workshops exert different levels, frequencies, and durations of the physical effort while performing the tasks of their jobs. The requirement for physical effort of a job changes from a combination of sitting, standing, and working with little requirement to an intense requirement for lifting objects weighing over 30 kg with no mechanical aid. In a general sense, it is expected that intense physical effort leads to decreasing productivity of an employee while performing the job. Although physical conditions impress a variety of performance criteria, our results showed that physical conditions had a greater effect on ''quality'', ''cooperating'', and ''creativity'', respectively. Emphasis on the quality of products and services has increased with the establishment of the ISO 9000 quality standard. The major requirements of ISO 9000 are that organizations develop and implement a set of routines and procedures for product design, manufacturing, delivery, service, and support. Standardization assures that all customers get the same product and service as promised. Reliable and standardized production with minimal variation can be achieved only when employees confirm and adhere to existing rules. Nevertheless, our results indicated that ergonomic improvements based on physical effort will contribute employee to reach to high quality. Environmental conditions: There were significant relationships between environmental conditions and many task and contextual performance criteria. Surprisingly, environmental conditions were strongly correlated with ''following organizational rules'' (r 0.463), ''cooperating'', (r 0.360) and ''concentrating'' (r 0.303) criteria, respectively. The first and third results highlight that those employees who work under unpleasant conditions obey the organizational rules and also concentrate the tasks much more than the others. However, environmental conditions decrease cooperation among co-workers to solve task problems.

Hazards: These findings are evidence that, according to our results, as predicted, employees demonstrate a high level of absenteeism and concentration, whereas the likelihood of accidents and injuries increases. The current study has several limitations that need to be considered when interpreting the results, and that should be addressed in future research. The major limitation of this study is the generalizability to other task environments. These findings were obtained from a medium-sized metal manufacturing company where some jobs are at intermediate levels of complexity. It may be that, for jobs of greater complexity and/or greater autonomy and discretion, improvements in the conditions that reduce or enhance the impact on task performance and contextual performance across a range of different occupations do continue to contribute to job performance. The opposite may be true for jobs of lower complexity. Different findings could reflect real differences in the way that workplace conditions affect task performance and contextual performance in these jobs. Future research, to be conducted in more than one organization, would strengthen the generalization of the present findings. Another limitation is the source of the ratings. The validity of supervisors' ratings as performance indicators has been widely criticized. As Podsakoff et al. (1997) suggested, subjective job performance ratings are less reliable because they are affected by rater's instinct factors (e.g., personality, cognitive errors). If alternative objective indicators for some performance measures such as productivity ratio, percentage of products that was rejected (quality), and the number of suggestions acquired from company record are used for assessments, much reliable results can be produced. References Antonioni, D., Park, H., 2001. The relationship between rater affect and three sources of 360-degree feedback ratings. Journal of Management 27, 479-495. Borman, W.C., 2004. The concept of organizational citizenship. Current Directions in Psychological Science 13 (6), 238-241. Borman, W.C., Motowidlo, S.J., 1993. Expanding the criterion domain to include elements of contextual performance. In: Schmitt, N., Borman, W.C. (Eds.), Personnel Selection in Organizations. Jossey-Bass, New York, pp. 71-98.

Borman, W.C., White, L.A., Dorsey, D.W., 1995. Effects of rate task performance and interpersonal factors on supervisor and peer performance rating. Journal of Applied Psychology 80 (1), 168-177. Coleman, V.I., Borman, W.C., 2000. Investigating the underlying structure of the citizenship performance domain. Human Resource Management Review 10 (1), 25-44. Das, B., Shikdar, A.A., 1999. Participative versus assigned production standard setting in repetitive industrial task: a strategy for improving worker productivity. International Journal ofOccupational Safety and Ergonomics 5 (3), 417-430. Dunlop, P.D., Lee, K., 2004. Workplace deviance, organizational citizenship behavior, and business unit performance: the bad apples do spoil the whole barrel. Journal of Organizational Behavior 25, 67-80. Ferris, G.R., Witt, L.A., Hochwarter, W.A., 2001. Interaction of social skill and general mental ability on job performance and salary. Journal of Applied Psychology 86 (6), 1075-1082. Goodman, S.A., Svyantek, D.J., 1999. Person-organization fit and contextual performance: do shared values matter. Journal of Vocational Behavior 55, 254-275. Groeneveld, S., Hartog, J., 2004. Overeducation, wages and promotions within the firm. Labour Economics 11, 701-714. Kiker, D.S., Motowidlo, S.J., 1999. Main and interaction effects of task and contextual performance on supervisory reward decisions. Journal of Applied Psychology 84 (4), 602-609. Lee, D.M., Alvares, K.M., 1977. Effects of sex on descriptions and evaluations of supervisory behavior in a simulated industrial setting. Journal of Applied Psychology 62, 405-410. Metal Industry Job Grading System, 1996. MESS, Ankara. Miron, E., Erez, M., Naveh, E., 2004. Do personal characteristics and cultural values that promote innovation, quality, and efficiency complete or complement each other? Journal of Organizational Behavior 25, 175-199.

Moser, K., Schuler, H., Funke, U., 1999. The moderating effect of raters' opportunities to observe ratees' job performance on the validity of an assessment centre. International Journal of Selection and Assessment 7 (3), 133-141. Motowidlo, S.J., Van Scotter, J.R., 1994. Evidence that task performance should be distinguished from contextual performance. Journal of Applied Psychology 79 (4), 475-480. Motowidlo, S.J., Borman, W.C., Schmitt, M.J., 1997. A theory of individual differences in task and contextual performance. Human Performance 10 (2), 71-83. Podsakoff, P.M., Ahearne, M., MacKenzie, S.B., 1997. Organizational citizenship behavior and the quantity and quality of work group performance. Journal of Applied Psychology 82 (2), 262-270. Posthuma, R.A., 2000. The dimensionality of supervisor evaluations of job performance. Journal of Business and Psychology 14 (3), 481-487. Resnik, M.L., Zanotti, A., 1997. Using ergonomics to target productivity improvements. Computers & Industrial Engineering 33 (1-2), 185-188. Rothstein, H.R., 1990. Interrater reliability of job performance ratings; growth to asymptote level with increasing opportunity to observe. Journal of Applied Psychology 75 (3), 322-327. Schmidt, F.L., Hunter, J., Outerbridge, A.N., 1986. Impact of job performance and ability on job knowledge, work sample performance, and supervisory ratings of job performance. Journal of Applied Psychology 71 (3), 432-439. Shikdar, A.A., Sawaqed, N.M., 2003. Worker productivity, and occupational health and safety issues in selected industries. Computers & Industrial Engineering 45 (4), 563-572. Van Scotter, J.R., Motowidlo, S.J., 1996. Interpersonal facilitation and job dedication as separate facets of contextual performance. Journal of Applied Psychology 81 (5), 525-531. Van Scotter, J.R., Motowidlo, S.J., Cross, T.C., 2000. Effects of task and contextual performance on systematic rewards. Journal of Applied Psychology 85 (4), 526-535.

Voon, D., Miller, P.W., 2005. Undereducation and overeducation in the Australian labour market. Economic Record 81 (255), s22-s33. Werner, J.M., 2000. Implications ofOCB and contextual performance for human resource management. Human Resource Management Review 10 (1), 3-24. Witt, L.A., Kacmar, K.M., Carlson, D.S., Zivnuska, S., 2002. Interactive effects of personality and organizational politics on contextual performance. Journal of Organizational Behavior 23, 911-926. Yeow, P.H.P., Sen, R.N., 2006. Productivity and quality improvements, revenue increment, and rejection cost reduction in the manual component insertion lines through the application of ergonomics. International Journal of Industrial Ergonomics 36 (4), 367-377. Yun, G.J., Donahue, L.M., Dudley, N.M., McFarland, L.A., 2005. Rater personality, rating format, and social context: implications for performance appraisal ratings. International Journal of Selection and Assessment 13 (2), 97-107.