Portugal apresenta, um pouco contraditoriamente, uma grande concentração de trabalho relativamente pouco qualificado e uma acentuada dispersão salarial entre os seus assalariados. É nosso entender que estas características fundamentais do nosso mercado de trabalho decorrem de uma clara opção pelo emprego, assente, por seu turno, numa matriz de relações industriais que combina de forma singular a flexibilidade e a segurança nas relações de trabalho. Decorrente desta aposta, que não se prevê sofrer alterações sensíveis no futuro próximo, é de admitir que a relativa escassez da procura de trabalho qualificado está para ficar.
AbstractThe backdrop to this inquiry into the relationship between worker job satisfaction and workplace representation in European nations is twofold. The first is that the bulk of research has focused on union membership and job satisfaction in Anglophone nations with their very different industrial relations systems and bargaining arrangements. The second and more immediate context is the dramatic shift from negative to positive in the association between union membership and job satisfaction (inter al.) observed in the most recent literature. Using data on 28 European nations from the last two waves of the European Working Conditions Survey, however, we report that workers in establishments with formal workplace representation record lower job satisfaction than their counterparts in plants without such representation. These findings of conditional correlation are then upgraded by constructing a pseudo‐panel with cohort fixed effects to take account of unobserved worker heterogeneity. First‐difference estimates suggest that the negative relationship between worker representation and job satisfaction found in cross section continues to hold. Next, an endogenous treatment effects model is deployed to address the possible endogeneity of worker representation. The results are supportive of a causal negative relationship between job satisfaction and worker representation. One interpretation of our findings is that in the matter of the association between unions and job satisfaction the jury is still out.
AbstractUsing a cross section of matched data from the employee and management questionnaires of the European Company Survey for 28 nations, this paper investigates the determinants of worker commitment and the potential contribution of commitment to establishment performance. An index of worker commitment is constructed from employer perceptions of the motivation of workers and their retention and absenteeism propensities, while the determinants of commitment are fashioned from observations taken from the worker representation side, ordered along dimensions such as perceived organizational trust and involvement. The commitment index is then linked to establishment performance outcomes. Key findings from the commitment equation are the positive role of trust in management, the quality of information exchanged, and the degree of worker representation influence in respect of major decisions taken by management. In turn, commitment emerges as a key correlate of establishment financial performance and labor productivity growth. Our supplemental sensitivity analysis is supportive of the interpretation of commitment as a driver of performance.
Using data from the 2013 European Company Survey, this chapter operationalizes the representation gap as the desire for greater employee involvement in decision-making expressed by the representative of the leading employee representative body at the workplace. According to this measure, there is evidence of a substantial shortfall in employee involvement in the European Union, not dissimilar to that reported for the United States. The chapter proceeds to investigate how the size of this representation gap varies by type of representative structure, information provided by management, the resource base available to the representatives, and the status of trust between the parties. Perceived deficits are found to be smaller where workplace representation is via works councils rather than union bodies. Furthermore, the desire for greater involvement is reduced where information provided the employee representative on a range of establishment issues is judged satisfactory. A higher frequency of meetings with management also appears to mitigate the expressed desire for greater involvement. Each of these results is robust to estimation over different country clusters. However, unlike the other arguments, the conclusion that shortfalls in employee involvement representation are smaller under works councils than union bodies is nullified where trust in management is lacking.
"In a characteristically combative treatment, Jirjahn (2008a) argues that Addison and Teixeira's (2006) finding of a negative relationship between works council presence and employment growth is a chimera produced by the way in which establishment size is measured. We reject his assertion of misspecification for two reasons; the second of which undoubtedly contributed to leading Jirjahn astray. And while Jirjahn's treatment is of interest in its own right, he does a poor job of portraying our overall analysis. Thus, he neglects our treatment of survival bias while ignoring our presentation of a dynamic labor demand model. Elsewhere he seems to grudgingly support the former (Jirjahn 2008b), and implicitly to accept our findings pertaining to employment adjustment (where we report that works councils do not slow the tortuous pace of employment adjustment in Germany). At root, the thrust of his treatment is adversarial and his position on the economic effects of works councils over-optimistic. But the main lesson of Jirjahn's critique is that more work is required of all of us in this area. Issues raised by the present exchange, apart from the need for a wider set of covariates and longer time frame, include the selection of firms into collective bargaining and works councils and out of the system, and the consequences for the raw point estimates. Pending this work, it would be idle to overstate the robustness of the extant results. We hinted at this in our own treatment in comparing cross-sectional results with dynamic panel estimates." (author's abstract)
