Temporary work tends to be associated with low participation in vocational training and thereby a reduced accumulation of human capital. Using Spanish 2001 data, Albert et al. (2005) found that workers with only temporary contracts were less likely to receive firm-provided and/or financed training than those with permanent ones. Using data from the European Community Household Panel (ECHP), the OECD (2002) also found that having a temporary job had a negative impact on the probability of participating in training.
The greater turnover of temporary workers, coupled with low conversion rates into permanent contracts, reduces incentives to invest in (jobspecific) human capital (Dolado et al., 2002; Bentolila et al., 2008). Guell and Petrongolo (2003) argue that the negative impact of temporary work on vocational training also depends on whether temporary contracts are mainly used to lower wage costs or as a screening device for entry-level jobs, with the effect being greater in the former case.
Dolado et al. (2002) and Dolado and Stucchi (2008) report evidence of a negative impact of a large use of temporary work on labour productivity growth, mainly via low participation rates in vocational training. Since the mid-1990s, Spain has been registering, together with a very high share of temporary work, a significant slowdown in labour productivity growth(32) which cannot simply be explained by the rapid growth in unskilled/low productivity labour. In fact, in the same period a sharp decline in the growth rate of Total Factor Productivity (TFP) has also occurred(33). Using Spanish data for the period 1991 to 2005, Dolado and Stucchi (2008) found that firms with a higher share of temporary workers are less productive, while those with high conversion rates of temporary contracts into permanent ones are more productive (for a given share of temporary work). The former effect can be explained by the lower investment in training of temporary workers, whereas the latter should be attributed to the lower in-work effort of workers, reflecting their perceived low probability of becoming permanent workers. The OECD (2010) report also found no evidence of, or a negative, cross-country correlation between the share of temporary work and total factor productivity (TFP).
Using a sample of Italian firms, Boeri and Garibaldi (2007) found that the share of temporary workers has a large negative impact on firm-level productivity growth, after controlling for a number of firm and worker characteristics. These authors argue that rising employment levels, in the aftermath of two-tier EPL reforms that facilitate the use of temporary work, led to falling labour productivity via decreasing marginal returns for labour. They also found that countries introducing flexibility at the margin have subsequently experienced a rise in the employment intensity of growth – the counter-part of declining productivity.
Chart 22 shows the cross-country correlation between the use of temporary contracts and the growth rate of TFP in the EU.
The information, based on ECHP data, suggests that firms provide considerably less education or training opportunities to temporary workers than they do to permanent workers (Chart 23).
Data from the European Company Survey (2008) also suggest that firms regularly pay less attention to the training needs of temporary workers compared to those of their permanent counterparts – a gap of about 30% in the EU - which is likely to be reflected in fewer training opportunities financed by enterprises (Chart 24).
Likewise, the Adult Education Survey(34) (AES) carried out in 2007 also suggests that participation rates in Continuous Vocational Training (CVT) - using non-formal education as a proxy indicator - are higher for permanent workers than for temporary workers in a majority of EU Member States (Chart 25).
On the other hand, figures for participation rates in formal (as opposed to non-formal) education suggest the presence of a ‘catching-up’ response, because take-up rates are higher for temporary than permanent workers (Chart 26). Mroz and Savage (2006) argue that after a spell in unemployment, young people will seek to increase their investment in human capital as a ‘catch-up’ response to the unemployment spell.(35) This result is corroborated by the analysis of the determinants of formal education in box 3 using LFS micro data, which suggests that temporary workers, being in a disadvantaged position in terms of pay, career prospects and job stability, do try to improve their lot by taking part in further formal education.
The remaining part of this section uses LFS anonymised microdata for the years 2006 and 2007 to evaluate the determinants of participation in ‘formal’(36) education. It should be highlighted that ‘formal’ education is not necessarily provided (i.e. free or subsidised) by firms, therefore there is no ex-ante presumption that participation rates for temporary workers will be lower than for permanent ones. Furthermore, the analysis of its determinants is carried out from the perspective of employees.
An econometric model is used to identify individual and firm characteristics that influence the likelihood/odds of employed people participating in further ‘formal’ education (see box 3). Overall, results suggest the existence of a ‘catching-up’ effect: a temporary worker with a lower or median level of education is more likely to take part in further education than a permanent worker and/or one with tertiary education. Work-life reconciliation aspects also seem to play a role, as single workers and part-time workers are more likely to participate in further education than full-time or married ones.
According to the estimated oddsratio(37), male workers have a 23% greater likelihood than female workers of participating in further ‘formal’ education. Taking part in further ‘formal’ education shows a very marked age profile (Chart 27), reflecting the well-established fact that returns to education decrease strongly with age (Cahuc and Zylberberg, 2005). The evidence shows that young adults (aged 20-24) are ten times more likely to participate in further ‘formal’ education than workers aged 60-64(38), but that participation in further education declines rapidly with age as workers aged 35-44 and 50-54 are only respectively three and two times more likely to do so than those aged 60-64. Firm size also plays a significant role in that workers in small or medium sized firms are less likely to take part in further education than workers in larger firms.(39)
Overall, the main conclusion of this section is that, on the one hand, temporary workers tend to have reduced access to on-the-job training than permanent ones while, on the other hand, they are more likely to participate in further ‘formal’ education (i.e. distinct from firm-provided training) than permanent workers, suggesting the existence of a catching-up effect, i.e. temporary workers attempting to overcome their disadvantaged economic situation.
A logistic regression is estimated to identify the determinants of participation in ‘formal’ education from the perspective of employed people. Explanatory variables include both individual and firm characteristics. The number of observations totals close to 2.3 million, covering the years 2006 and 2007. Twenty five EU Member States are covered (EU27 minus MT and IE). Regressions are estimated weighting observations by the weighting variable provided in the dataset (COEFF). The odds ratio of participating in ‘formal’ education is modelled as:
where
i: individual
j: country
t: year
P: probability
yijt: {0: non-participation; 1: participation}
X1: Year
X2: Country
X3: Experience (0-1, 2-3, 4-5, 6-10, 11-15, 16-20, 21-25, 26+) proxied by the number of years since highest education level was completed
X4: Gender
X5: Age groups (20-24, 25-29, 30-34, 35-39, 40-44, 45-49, 50-54, 55-59, 60-64)
X6: Marital status (1: widowed, divorced or separated; 2: single; 3: married)
X7: Work status (1: self-employment; 2: permanent work; 3: temporary work)
X8: Full/part-time
X9: Education level (1: primary; 2: secondary; 3: tertiary)
X10: Size of the firm (number of people working in the local unit): (1-10; 11-19; 20-49; 50+)
X11: Supervisory role (1: Yes; 2: No)
X12: Economic activities (NACE at 1 digit)
X13: Occupations (ISCO at 1 digit)
(32) | The average annual growth rate of GDP per hour worked fell from 2.9% in the 1970-1994 period to 0.3% in 1995-2005 period. |
(33) | The average annual growth rate of TFP has decreased from 0.6% in the 1980-1994 period to -0.8% in 1995-2005 period. |
(34) | The Adult Education Surveys (AES) are part of the EU Statistics on lifelong learning. The surveys cover participation in education and lifelong learning activities (formal, non-formal and informal learning). All definitions apply to all persons aged 25-64. The AES are planned to be conducted every five years. The total sample size for the first 24 countries participating in the 2007 wave is about 170.000. |
(35) | Mroz and Savage (2006) find that recent unemployment has a significant positive effect on whether young people take part in training today. However, despite this catch-up response and an absence of persistent unemployment effects, they find evidence of long-lived ‘blemishes’ from unemployment. |
(36) | Student or apprentice in regular education during the last 4 weeks (LFS’s EDUCSTAT variable). The same exercise was carried out for ‘informal’ education (COURATT), but results were not found to be sufficiently conclusive to be worth reporting. |
(37) | The odds ratio is the ratio of odds of an event occurring in one group (e.g. men) to the odds of it occurring in another group (e.g. women). An odds ratio greater than 1 indicates that the event is more likely to occur in the first group. An odds ratio less than 1 indicates that the event is less likely to occur in the first group. |
(38) | The reference category. |
(39) | With more than 50 workers. |