Firm-sponsored training in regulated labour markets: evidence from Spain

Abstract

Using data from the 1994 European Community Household Panel Survey, the author examines who receives formal firm-sponsored training in Spain. The author finds that the distribution of firm-sponsored training in the work force is uneven and concentrated among more skilled workers in the upper deciles of the wage distribution. The data show that the likelihood of receiving firm-sponsored training for a low education employee is much lower. Also, the better-educated employees in high wage occupations of the largest establishments have higher probabilities of receiving specific training. Spain has a highly regulated labour market, and the labour market frictions and institutions compress and distort the structure of wages. However, the results suggest that the highly compressed wage structure do not provide firms with the incentive to invest in general training.

Acknowledgements

A preliminary version of this paper was published by the Instituto Valenciano de Investigaciones Económicas, Working Paper No. EC 2001-15. The author is grateful for financial support by the Ministerio de Ciencia y Tecnología under Research Project No. SEC 2001-2763.

Notes

 However, the authors find that the correlation between employer and employee measures are less than 0.5, which are much lower than correlation for other variables that have been used in wage equations.

 Therefore, I can not control for the type of contract.

 Our final subset contains only workers with more than 21 months in their current jobs. Unsuccessfully, we undersample recently hired workers and turnover jobs. The problem of tenure variable is because the 1994 ECHP Survey was conducted in the fourth quarter of 1994 and the Survey asks retrospective questions about the training that workers have received in 1993. This suggests that the subset may miss spells of firm-sponsored training because it does not contain newly hired workers (see Barron et al., 1997).

 It must be noted that the comparisons of training incidence differ across countries, and Spain has a different industry mix from that of the USA.

 The impact of public sector versus private sector on probabilities of receiving training was also tested in the model. The results present limited evidence that workers in public companies are more likely to participate in firm-sponsored training than private company employees. Overall, this coefficient was not statistically significant.

 The ECHP data contain two experience variables; years of experience in current job and worker's age. The years of tenure is a measure of the relevant experience that workers have (as Barron et al., 1999, refer to it): experience previously acquired on-the-job training. The statistical significance of the tenure effect is easily rejected for all specifications considered in both models. The measure ‘worker's age’ is a proxy for general experience, because they accumulate general skills in jobs that are not relevant to their current employment.

 The critical chi-squared value at the 0.005 significance level is 29.8 for13 degrees of freedom.

 The sample was stratified by establishment size in the following manner: 36.0% of all workers in establishments with 1–19 regular paid employees, 12.5% of workers in establishments with 20–49 employees, 25.5% of workers in establishments with 50–499 employees, and 25.7% of workers in establishments with 500 or more employees.

 Black et al. (1999) argue that differences in formal training by firm size and establishment size arise from cost advantages for larger firms. Additionally, because unions are more prevalent in firms with a large number of employees, and since training is positively correlated with unionization, these coefficients could be picking up ‘collective voice’ effects.

 These findings point in the same direction as the Krueger and Rouse (1998) study of impact of a workplace education programme at two companies, one in the manufacturing sector, the other in the service sector. They estimate a small positive impact of the training programme on earnings at the manufacturing company but an insignificant impact at the service company.

 See Acemoglu and Pischke (1999b), Tables 2 and 3 in comparison with our Table 5.

 Booth and Zoega (2000) suggest that better-than-average firms (with monopsony power due to a high quality workforce) can offers higher wages to its well-trained workforce while enjoying monopsony profits due to the complexity of tasks performed within its ranks. Thus, in the presence of monopsony power, firms are willing to pay for training, which is specific to the task performed but general to the industry.

 Fig. 1 may be compared with the corresponding Fig. 1 in Acemoglu and Pischke (1999c).