The Tilburg Trust Experiment: A Dutch RCT Experiment with Social Assistance

The Tilburg RCT experiment contains four treatment conditions, one control condition and a reference group.
- The three treatments are: (1) claimants are exempted from the standard obligations, such as applying for jobs, accepting job offers indicated by the municipality or joining reintegration programmes combined with an additional work bonus in the form of an extra earnings release option; (2) the coaching treatment combined with earnings release: participants receive intensive mediation support (5 contacts a year) through tailor-made supervision focused on reintegration into paid work or social participation activities when paid work is not yet feasible; (3) intensive tailor-made mediation (6 contacts a year) but without additional earnings release options.

Work bonus
According to article 83 the existing release conditions are relaxed: whereas the current conditions are that people may keep 25% of additionally earned income up to a maximum of 199 EUR/month for a maximum of six months, the experimental treatment conditions are that people may keep 50% of additionally earned income up to max. 202 EUR/month for the duration of the experiment, that is 24 months. Tilburg has put on top of this a work bonus of an extra 200 EURO/month when mediation in group 1 or 2 will lead to fulltime reintegration into paid work. Such a bonus that is paid out once a year may be rendered 2 times at a maximum.

- In the control condition, claimants are subject to the standard treatment of the municipality of Tilburg which is largely set - with some discretionary power at the side of the municipality - by the rules and regulation of the Dutch Participation Law as of January, 1, 2015. Social welfare in the Netherlands is a welfare benefit scheme of last resort (hence, with a means and work test) for people having no other means of existences. The social assistance scheme provides monthly income support to single persons and households at a minimum income level, that is 70% of the minimum wage for a single person. For couples and families with children there are different amounts. The monthly amount for a single person is 992 EUR. On top of that welfare claimants may be eligible for child, housing and healthcare allowances. Recipients have to comply with certain rules, such as writing application letters, applying to job vacancies indicated by the municipality or following skills or personal development programmes. Recipients that fail to comply with the rules may be sanctioned by freezing or cutting their monthly benefit payment. Beneficiaries may be exempted from these obligations in the case people are incapacitated to work or face severe health problems. Local governments (municipalities and regional councils) are charged with the task of executing the scheme.

- Additionally, for methodological reasons a reference group of randomly selected claimants is designed consisting of social assistance beneficiaries who have any involvement in the experiment. The outcomes for the three treatment groups will be compared with the outcomes for the standard group and this reference group.

2017-12-01

2019-10-01

Our primary outcome indicator concerns reintegration into fulltime and parttime paid work.

The reintegration into fulltime or parttime paid work will be measured using local municipality data and administrative micro-data from Statistics Netherlands.

Our secondary outcome indicators are (1) social participation and activation, (2) participants’ health and well-being (3) participants' self-reliance and self-management capacities (5) satisfaction with support (6) participants' capabilities and (7) participants' financial situation (earnings income, poverty, financial stress).

Social participation and activation is measured through self-reported survey data (e.g. frequency of contacts with social network; hours of volunteer work). Participants’ self-reliance and self-management capacities and capability scores, satisfaction with support and health and well-being is measured using survey data on self-efficacy, capability list of questions, support evaluation, subjective health, mental health and subjective wellbeing but possibly also health data from Statistics Netherlands. Participants’ financial situation is measured using administrative data and self-reported survey data.

The experiment is a cooperation between the University of Tilburg and the Municipality of Tilburg. Participating welfare claimants are randomly assigned to the four different conditions described above (three treatments groups and one control group).The target population for the experiment consists of all welfare claimants in Tilburg, except for some exclusion groups (not speaking Dutch, fully disabled, 65 years and older, homeless people, living in institutions). The target population is randomized over the four groups (three treatments groups and one control group). After this randomized assignment procedure claimants were invited to participate for that specific treatment during an information and enrolment campaign. Later on each participant was invited for a talk with the case-worker in which the client became informed about his/her treatment group and what it means for the support provided during the two years of experimentation. In the meeting also final confirmation of participation is asked for. Participation had to be confirmed through informed consent. Participation is voluntary, which means that participants can decide to drop out or withdraw at any time.

In total, 783 claimants have signed up to take part in the study after which 147 participants in the end withdrew. The net total number is 636 participants which makes up for around 10% of the targeted population. Data is collected using administrative data as well as three longitudinal surveys. Data collection takes place at three points of time: Oct – Nov 2017 (baseline), Oct - Nov 2018 (midline) and Aug – Sep 2019 (endline).

After selection of the target group the randomization is conducted in office using a computer programme operating on State 14.1.

individuals (social assistance recipients)

No

800 persons without clustering

800 observations in 4 groups

Sample size: planned number of clusters
0
Sample size: planned number of observations
planned 800; realised 783; 147 attrition/refusal/drop-outs; net number: 636
Treatment 1 (exemption group): 165 claimants; Treatment 2 (coaching group): 161 claimants; Treatment 3 (intensive mediation): 138 claimants; Control: 172 claimants.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

200 persons in each group; power 95%; alpha 5%; effect size 10%; p=0.20 (control); 0.30 (treatment); sd 0.25 (control); 0.35 (treatment)