Do Public or Private Providers of Employment Services Matter for Employment? Evidence from a Randomized Evaluation in Denmark

Researchers:
Michael Svarer
Kai Reywald
Fieldwork by:
Location:
Copehagen, Aarhus
Sample:
3,107 individuals
Timeline:
2011 - 2012
Target group:
  • Job seekers
Outcome of interest:
  • Employment
Intervention type:
  • Job counseling

Though sometimes referred to as “a feature of the welfare state,” some employment and training programs have recently been contracted out to private enterprises. Researchers conducted a randomized evaluation to test the impact of public and private employment service providers on labor market outcomes of unemployed university graduates. Overall, researchers found that public and private employment service provision did not differ in their impacts on labor market outcomes. Private providers were less cost-effective and, despite providing more intensive and employment-oriented services, had lower client satisfaction compared to public providers.

Policy issue

Though sometimes referred to as “a feature of the welfare state,” the public provision of some employment and training programs have recently been contracted out to private enterprises. Previous research has focused on economically disadvantaged individuals who were hard-to-place or at-risk for long-term unemployment. However, less is known about how different employment service providers impact highly educated, unemployed individuals who may face different labor markets and specialized recruitment processes. Are private providers able to provide better employment services more efficiently than public providers? How does the public or private provision of employment services affect the employment rate of highly educated individuals?

Context of the evaluation

In Denmark, about 100 government municipal centers are responsible for helping the unemployed back to work by providing placement services and labor market programs.

From 2007 to 2011, municipalities were legally required to contract out employment services for individuals with higher education (master’s degree or equivalent). After the financial crisis hit Denmark in 2008, unemployment rates for highly educated individuals rose sharply and enrollment in these employment services increased.

In November 2011, a law repealed the private provider contracting requirement. As of July 2012, all highly educated, unemployed individuals were referred to public providers for employment services.

An individual getting assistance at an employment office.

Details of the intervention

Researchers conducted a randomized evaluation to compare the impact of public and private employment service providers on labor market outcomes of university graduates. The sample consisted of 3,107 unemployed individuals, all with masters’ or equivalent degrees, who had been referred to a private provider. 1,541 individuals were randomly assigned to receive employment services directly from municipal job centers. The remaining 1,566 individuals received these services from private providers.

Private providers were required to contact their clients at least once per quarter and ensure their clients’ CVs were updated regularly. In turn, unemployed individuals had to participate in programs such as classroom training, job counseling, subsidized or temporary employment, and education or vocational training programs. Private providers also received performance-based financial incentives for helping their clients back to work as soon as possible. The National Labor Market Authority, part of the Ministry of Employment, selected these private providers through a competitive bidding process.

Researchers used administrative and survey data to measure the impact of the intervention on a number of labor market outcomes including regular employment, subsidized employment, and unemployment rates

Results and policy lessons

Overall, researchers found that labor market outcomes for unemployed individuals did not differ whether they received employment services from public or private providers. Thirteen weeks after the program started, unemployment rates fell by 42 –43 percentage points, (from 80 percent at baseline to 36.4 percent in the private provider group and 37.6 percent in the public provider group).

Additionally, private providers were less cost-effective and, despite providing more intensive and employment-oriented services, had lower client satisfaction compared to public providers. Excluding the expenses of social benefits that individuals collected, the cost per individual assigned to private providers was 10,392 DKK (US$2,007 in 2011) compared to 8,248 DKK (US$1,593 in 2011) per individual assigned to the public provider. These results suggest that private providers may not be a cost-effective method for increasing employment rates.

Reywald, Kai, Michael Rosholm, and Michael Svarer. "Do Public or Private Providers of Employment Services Matter for Employment? Evidence from a Randomized Experiment." Labor Economics Volume 45, April 2017, Pages 169–187.