Labor Market Training for the Unemployed in Denmark

Lars Skipper
Fieldwork by:
1994 - 1994
Target group:
  • Job seekers
  • Adults
Outcome of interest:
  • Earnings and income
  • Employment
Intervention type:
  • Job counseling
  • Training
  • Vocational training (TVET)

Policy issue

Vocational training is often advocated as a means of maintaining and improving the qualifications of the labor force. In developed countries, vocational training programs are often initiated to facilitate adjustments to the structural changes in the labor market produced by the forces of globalization. However, if training does not provide relevant skills, it is also possible that time spent in vocational training simply displaces time spent working or looking for employment. The actual effect of vocational training programs is not known because very few empirical evaluations of such programs have been conducted, particularly in Europe.

Context of the evaluation

In the 1960s, in order to support the transition from an agricultural to an industrial society, the Danish government initiated a series of adult vocational training programs, called "AMU programs". These programs were intended to promote a flexible labor force, by giving individuals opportunities to acquire the skills currently demanded by the labor market. Over time, the programs were adapted in accordance with the changing needs of the Danish labor market; in the 1990s, the training programs were focused on knowledge, service, and information.1

Teacher scrawling notes on his flipchart
Teacher scrawling notes on his flipchart
Aude Guerrucci

Details of the intervention

This evaluation examines the impact of AMU programs on participants' subsequent employment and wage rate. In April 1994, the AMU program application process was randomized so that 423 individuals were accepted into the program (treatment group) and 387 individuals were rejected (the comparison group). Over the next two months, individuals in the treatment group were enrolled in AMU courses, which lasted from one to seven weeks, with an average duration of two weeks.

A survey, which included detailed questions on the respondent' s labor market history, marital status, and reasons for applying to the program, was conducted in December 1994. Administrative records, which contained weekly unemployment and employment status, and annual earning measures, were collected beginning in April 1993 and then matched with the individual surveys.

The random assignment of applicants into treatment and comparison groups was complicated by the presence of no-shows (members of the treatment group who did not show up at the course they signed up for) and crossovers (members of the comparison group who ended up enrolling in training). Twelve percent of the total sample13 percent of the treatment group and 14 percent of the comparison groupdid not show up for the survey after applying to the program, and 22 percent of the comparison group ended up receiving training. There is no clear explanation for the high degree of crossover, though it is not uncommon. To account for the presence of no-shows and crossovers, researchers created an additional experimental estimator to confirm the results of the program on participants.

Results and policy lessons

Impact on unemployment: Results showed that unemployment actually increased by around ten percentage points among program participants who had been initially assigned to treatment. However, this effect dissipated over time, and nine months after training there was no significant difference in employment rates between training participants and non-participants.

Vocational training may affect employment rates through two mechanisms: the locking-in effect, which reduced the intensity of an individual' s job search while in training, and the post-program effect. If the training provided participants with formal qualifications, the post-program effect may have increased unemployment in the short-run because it narrowed the pool of appropriate jobs and increased expected wages. For example, if an individual acquired a bus driver' s license through the training program, he may look only for jobs as a bus driver, which limits the number of available vacancies to choose from.

Impact on wage rates: The training program did not appear to have any significant impact on the subsequent hourly wages of participants. However, as the average length of the program was quite short (2 weeks), this was not unexpected.

Rosholm, Michael, and Lars Skipper. 2009. "Is labour market training a curse for the unemployed? Evidence from a social experiment." Journal of Applied Econometrics. 24 (2009): 338-365.
Danish Ministry of Education. "AMU: The Danish Adult Vocational Training Programmes."