What prevents job-seekers from training? Reframing training access to reduce non-take up

From July 2022 to July 2023, we launch a field experiment designed in cooperation and implemented by the Public Employment Service (PES) of Lower Austria (Arbeitsmarktservice Niederösterreich, AMS NÖ). The aim is to increase training and employment among the unemployed by increasing completion of training programs. The treatment consists of an email newsletter sent to job-seekers with information on training programs offered by the public employment service. The newsletter includes a voucher framing that informs about the range of available training programs and signals their monetary value to job-seekers. In a second treatment arm, job-seekers receive on top additional information on the type of jobs with the highest number of open vacancies.

Two different treatment arms and one control arm vary the type of information that the job-seekers receive on training programs. The job-seekers in the sample will be randomly allocated to each of the three groups on an individual level. This randomization is conducted separately for each sending date. The different treatment and control arms are as follows:

1. Group: control
2. Group: treatment with a newsletter that includes a voucher and information on training provided by the PES targeted to the educational attainment of job-seekers.
3. Group: identical newsletter as the 2nd group complemented with job information also targeted to the educational attainment of job-seekers.

The treatment for group 2 is intended to spur self-driven initiative by raising perceived autonomy over training choices, and increasing reciprocity and reducing social stigma by signalling the monetary value of training programs. On top of it, the treatment for group 3 should also reduce asymmetric information on the demand for labor. Except for receiving the email newsletter, participants in the treatment groups continue to be subject to the same rules and obligations, including regular appointments with their job counselors, as participants in the control group.

2022-07-20

2023-07-31

Our two primary outcomes are training completion and days in employment.

Training completion refers to whether or not a training program was completed after two years
since the intervention.

Employment measures days in employment after two years since the
intervention.

Training take-up, Type of training,
Activation measures, Counseling, Subsidized employment
Unemployment duration, Earnings
Social and health outcomes

The secondary outcomes are not per se targeted outcomes, but can be
seen as mechanisms leading to the primary outcomes described above or providing additional
information on economic and social consequences. To get a better understanding of the immediate
effects of our intervention, we differentiate by the type of training that was completed. This
can be shorter training programs focused on employment orientation or longer training programs
focused on enhancing qualifications, which aim at increasing human capital. Further, we look at
potential spill-over effects to other active labor market programs, such as activation measures, such
as application courses, job search assistance and counselling as well as employment promotion such
as subsidized employment and direct job creation. To get a better understanding of employment
effects, we also look at effects on unemployment duration as well as earnings and employment
stability in post-intervention employment. Unemployment duration allows us to quantify lock-in
effects. Earnings and employment stability provide us with measures for the quality of employment.
Finally, using the additional post-intervention survey, we will collect outcome data on job search
intensity, and social outcomes including well-being, social inclusion and health. Other studies have
found positive effects of active labor market policies on mental health and social inclusion (Wang
et al., 2021).

The randomization is conducted for each sending date separately. In addition, we use the other
stratification variables as specified in table 1 to construct strata, i.e. blocks. The treatment assignment
is in the next step conducted randomly within these strata. Ideally, we would like to stratify
by more variables than just the three used, but the sample size does not allow more stratification
variables, because then the strata would become too small. In the first wave, we additionally
stratify on unemployment duration, as this population also includes unemployed with varying unemployment
durations, as already explained above.

Stratified randomization

Individual level

No

-

-

Our sample comprises of all eligible unemployed with an unemployment duration of at least six
months to maximum 2 years in June 2022. In the subsequent bi-monthly treatments, new unemployed
who exceed a duration of over 6 months will be treated. Every individual will only be
treated once. This means that job-seekers who were part of the pre-cursor study are excluded to
allow for clean inference of treatment effects. The sample size is thus unclear and will depend on
the labor market situation. It will, however, most likely be larger than the sample we had in the
precursor intervention of 2021.

Lehner_Schwarz_2022_PAP_What prevents job-seekers from training-rand1.pdf