Active Labour Market Policy: How to Measure What Works

To measure an active labour market policy programme, track people at three points: when they leave, six months later and two to three years later, and compare them with similar people who did not take part. ESF+ already requires the first two. The third matters because a review of over 200 evaluations found average impacts "close to zero in the short run", turning positive two to three years after the programme. A programme judged only at exit can look like a failure when it works. I deliver coaching inside Germany's publicly funded activation programmes. This guide is for the ministries, municipalities and ESF+ teams who fund and report on them.

What ESF+ asks you to report

Annex I of the ESF+ Regulation (opens in a new tab) sets common indicators for programmes under the employment objective, which covers "activation measures for all jobseekers", with long-term unemployed and inactive people named.

The longer-term indicators must rest at least on "a representative sample of participants". Where data comes from registers, member states may use national definitions.

These indicators tell a funder who took part and where they stood afterwards. They do not tell you whether the programme caused the change. For that you need two more things: time and a comparison group.

Active labour market policy effects take years

David Card, Jochen Kluve and Andrea Weber reviewed over 200 recent evaluations of active labour market programmes (Journal of the European Economic Association, 2018 (opens in a new tab)). Four findings shape any measurement plan:

1. Average impacts sit close to zero in the short run and turn more positive two to three years after the programme ends.
2. The time profile depends on the type of programme. Programmes that emphasise building skills and knowledge show larger average gains.
3. Impacts are larger for women and for participants who enter from long-term unemployment.
4. Programmes are more likely to show positive impacts in a recession.

The practical point: a skills-heavy programme for long-term unemployed adults can post weak six-month numbers and still pay off later. If your reporting stops at six months, add a later follow-up from register data.

Compare against what would have happened

A placement rate counts who found work. It cannot separate the people the programme helped from the people who would have found work anyway.

Germany's Institute for Employment Research (IAB) solves this with matching. For its evaluation of activation measures for people on basic income support, it picked, for each participant, a statistical twin who did not take part but looked alike on the factors that drive participation, such as education and length of unemployment. Then it compared both groups three years after the start (opens in a new tab). Measures run by providers raised the chance of regular, unsubsidised employment by between 6 and 16 percent. Measures run at an employer raised it by between about 44 and 62 percent.

You do not need a national research institute to borrow the idea. An employment service with register data can draw a comparison group from people with similar profiles who registered in the same months and did not join. Agree the method before the programme starts, so nobody picks it after seeing the results.

Measure what changes first

Employment is the end of a chain. Songqi Liu, Jason Huang and Mo Wang found in a 2014 meta-analysis (opens in a new tab) that job search interventions work partly through three channels: job search skills, job search self-efficacy (a person's belief that they can do what the search requires) and job search behaviour.

That gives you indicators you can collect during the programme:

• Self-efficacy: a short self-assessment at the first and last session, with the same questions both times.
• Skills: a finished CV for a named target job, a completed mock interview.
• Behaviour: applications sent, interviews attended, contacts made, per week.
• The plan: a written next step with a date and a named person who follows up.

These numbers move within weeks. They show a funder whether the programme is doing its job long before the employment data arrives, and they show the trainer which part of the group is stuck.

Track the person, not only the programme

The EU Council Recommendation on long-term unemployment (opens in a new tab) sets three monitoring questions: whether registered long-term unemployed people "have regained employment, whether their integration into the labour market is sustainable and the use of job-integration agreements".

The middle question, sustainability, needs data from well after the exit date. Germany builds it into payment: a success-paid placement provider under § 45 (6) SGB III (opens in a new tab) receives €1,250 after six weeks in the job and the rest only after six months. If you fund by outcome, tie part of the payment to the job lasting.

In Lithuania, Užimtumo tarnyba asks municipalities to approve their employment promotion programmes (opens in a new tab) for no longer than one year, with clear participant numbers, services, measures and results for that period. It publishes integration indicators for programme participants. A coaching module inside such a programme should report into the same frame.

A measurement plan on one page

Where I fit in

Since September 2022 I have worked as a career coach for International Motivation Training Deutschland (IMT), an approved provider, in programmes funded by the Jobcenter and the Arbeitsagentur (AVGS, § 45 SGB III). Across Germany's career and integration programmes I have coached more than 2,000 adults, in groups and one to one, in German and Russian. When groups finish, the feedback I hear most is about individual attention, a clearer plan for the next step, and not feeling alone in it. More in what 2,000 adults starting over in Germany taught me.

When I build a programme for a funder, I start from the indicators you report on and fit the sessions, hours and documentation to them. See how I design and deliver employability programmes, read what a good employability training curriculum covers, or book a call.