Labour protection and productivity in EU economies: 1995-2005.

• Author: Damiani, Mirella
• Position: Report

1. Introduction (3)

   Over the past decade, disappointing productivity growth has been recorded in the European economy and the catching-up of Europe on the U.S. has slowed significantly. Productivity differentials have mainly involved market services. Indeed, new research (van Ark et al., 2008) sheds some light on differential patterns of growth in labour-intensive sectors, such as services, explaining the different performances of multifactor productivity recorded in the European economy and the US. These findings encourage further inquiry into the role of management of labour resources and their regulation with regard to successes or failures in Europe.

   As known, various hypotheses on the role of labour regulation have been advocated, and their relevance has been tested in a growing number of empirical studies. Many investigations analyse the impact of these policies on employment and unemployment rates, or on unemployment inflows and outflows, but reserve less space for productivity growth. Conversely, the present paper examines the more controversial issue of the impact of labour institutions on productivity outcomes, only recently addressed by some country studies (Dew-Becker and Gordon, 2008), or industry-level cross-country research (Micco and Pages, 2006; Bassanini and Venn, 2007; Bassanini, Nunziata and Venn, 2009) (4).

   The expected impact of the role of labour market institutions on productivity is ambiguous and mainly related to restrictions on firing. Scarce attention is devoted to fixed-term contracts.

   The deeper motives for promoting labour market flexibility are found in the theoretical literature on the potential costs of labour protection. Such protection, as argued by Hopenhayn and Rogerson (1993), perturbs the reallocation of resources from declining firms to more dynamic ones with above average productivity growth. In addition, these protective devices tend to alter the allocation of resources among sectors.

   Economies with rigid labour markets manifest a distortion in their innovation activities, since they adopt mainly secondary innovations which determine a cost reduction in existing goods, but they do not experiment with primary innovations, such as those related to new products, characterised by higher returns but also higher variance (Saint Paul, 2002) (5). Such economies, which are prevalent in Europe, show an international specialization in secure goods, at an advanced stage of their product lifecycle, and this contributes toward explaining why "Europe appears as less high tech than the United States" (Saint Paul, 2002, p. 376) and why it falls behind in terms of long-run productivity growth.

   Other key channels able to explain unfavourable consequences are related to capital returns and worker effort. Returns of investment are lowered by job protection provisions, as shown by Bertola (1994). Analogous negative impacts are brought about by worker incentives since labour protection lowers the probabilities of layoffs for disciplinary reasons; under less threat of dismissal, opportunistic behaviour is encouraged (Boeri and Jimeno, 2005).

   By contrast, theoretical arguments in favour of employee protection underline the positive effects on productivity of long-term relationships and incentives for workers to upgrade their skills. Two main channels might be advocated. First, labour policies may increase the stock of human capital through the promoting of training. Second, for a given stock of human capital, "policies that reduce social conflict might condition workers' effort and their willingness." (OECD, 2007, p.65)

   In this vein, Nickell and Layard (1999) signal that employment protection has a positive effect, because it increases job tenure and stimulates on-the-job training. Belot, Boone and van Ours (2007) also highlight that, when effort and investments in human capital are non-contractible, employment protection solves hold-up problems; protection of this kind encourages employees to invest in match-specific human capital by increasing the probability of the survival of the match, and this beneficial effect is stronger in those sectors where firm specialization in competences is more important. However, there is a trade-off between these positive effects and the negative consequences of firing costs. In any case, there is a positive optimal level of employment protection. In addition, Wasmer (2002) shows that different levels of employment protection affect skills in different ways: low levels favour general portable skills, higher levels determine firm-specific skills. (6)

   Institutional interactions are of strategical importance, some of which are related to the wage setting system. A limited discretionality of employers in wage setting lead them to adjust employment levels, and only protective provisions make union wage bargaining power effective; this explains why employment protection and unions are quite often complementary institutions, since binding rules on employment contracts may be useless if firms are free to adjust wages downwards (Boeri and Van Ours, 2008, chapter 13).

   On the other hand, collective bargaining rules and sectoral agreements which generate low spreads between firms may prevent opportunistic behaviour by employers when on-the-job training is observable, but not verifiable. In such circumstances, workers may accept a wage cut with a promise of obtaining on-the-job training, but the firm has an incentive to renege on its promise and reaps some gains from cheap labour (Malcomson, 1999). Institutional constraints on wage setting, more enforceable if collective contracts are extended by law to third parties at national or sectoral level, thus impede hold-ups and generate the upgrading of skills (7).

   Other links with incentive schemes are provided by efficiency wage models. As shown by Guell (2000), the choice of fixed-term contracts is plausible in a context of efficiency wages, but only if there is a sufficiently high transformation rate of fixed-term contracts into permanent ones; otherwise, shirking is always strictly preferred. "The idea behind this is very simple: if a worker always becomes unemployed independently of the effort expended, there is no way to give incentives to the worker by paying him a higher wage" (Guell, 2000, p.10).

   In a context of asymmetric information, other institutional complementarities occur and are related to the role of works councils and the presence of employees on supervisory boards. As shown by Freeman and Lazear (1994), management tends to use information strategically, declaring bad states of the firm to extract more effort from employees; in their turn, employees, knowing this, disregard management and rationally choose low-effort strategies (8). Legal norms and elected works councils overcome this communication problem, as they provide more information, inducing more effort; Freeman and Lazear also argue that, by enhancing job security, employees attribute more importance to expected future profits, show greater loyalty, and invest more in firm-specific skills.

   To sum up, this brief excursus demonstrates that multiple dimensions of labour regulation must be taken into account and that their impact helps to explain national and sectoral disparities in productivity growth.

   In the present paper, we analyse these disparities in Europe and then focus on some driving forces such as flexible employment contracts and collective labour regulation to explain various patterns of multifactor productivity. As said above, other recent studies examine the influence of labour protection on productivity and focus most of their analysis on dismissal rules, with some extensions to other forms of regulation (see, for instance, Bassanini, Nunziata and Venn, 2009). The present paper does not consider rules on firing restrictions and mainly examines only those norms that increase flexibility on the margin; it focuses on those institutional devices which involve co-decision making and verifies the role of co-management when works councils and codetermination rules apply.

   In order to obtain a comprehensive database, we considered two distinct 'areas' of labour regulation, employment laws and collective relations laws. As regards the first, we selected some of the indexes which summarise the regulation of allocation of labour inputs in the productive process: part-time, fixed-term contracts and hours worked. (9) For the second area, collective relations, we included not only payoff rights, but also workers' decision rights, which may have some impact on productivity (10). For instance, computing the presence of employees in worker councils or on boards of directors yields a more precise evaluation of arrangements that give a 'voice' to employees and which may represent additional channels influencing productivity (11).

   The second set of information is on growth and is taken from the EU KLEMS Growth and Productivity Accounts. This database is the outcome of a research project, financed by the European Commission and aimed at analysing productivity in the European Union (Timmer et al., 2007). It allows a detailed analysis of European economies and explicitly considers important issues such as average skills of the labour force, capital services in information and communication technology, and their diversities across sectors and countries.

   By merging statistical information made available by these two sets of databases, with additional data from EUROSTAT, as described in Section 2, we explore the potential impact of labour regulation on productivity performances for a sample of European Economy countries over the period 1995-2005.

   The paper is organised as follows: Section 2 provides data description; Section 3 presents our main findings on country-sectoral growth differentials and labour arrangements; Section 4 offers econometric estimates; Section 5 concludes.

2. Data description

   As mentioned above, our empirical investigation...