FORWORK - Fostering Opportunities of Refugee Workers

Last registered on December 27, 2021


Trial Information

General Information

FORWORK - Fostering Opportunities of Refugee Workers
Initial registration date
December 23, 2021

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
December 27, 2021, 10:52 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.



Primary Investigator

Bocconi University

Other Primary Investigator(s)

PI Affiliation
Fondazione Ing. Rodolfo Debenedetti
PI Affiliation
Bocconi University
PI Affiliation
University of Maryland
PI Affiliation
Università of Brescia

Additional Trial Information

On going
Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Notwithstanding the recommendations of the Italian Ministry of Labour and Social Policies, the development of fully fledged programmes for integration of asylum seekers in Italy has not mirrored the overwhelming number of arrivals in reception centres. Pilot studies are needed to inform policy at the national level and to improve the opportunities of social and labour market integration for asylum seekers. The present project deploys a new model of integration which we pilot in Piedmont, a large region of Northern Italy characterized by an integration model for asylum seekers based on widespread territorial reception in small centres.

The intervention offers to asylum seekers the assistance of a mentor for the creation of an individualized integration plan consisting in an initial skills assessment, in the offer of Italian short language and vocational training courses and, finally, in professional assistance for job search.

The effectiveness of this intervention package on labour market and social integration outcomes will be evaluated via a clustered randomized controlled trial with blocking. All eligible managing host structures will be included in the experimental sample. Given the peculiarity of the Piedmont system outlined above, randomization will be performed at the asylum seekers’ managing entity level. Expected outcomes will be measured by integrating quantitative data from administrative sources and ad-hoc surveys.

Registration Citation

Abbiati, Giovanni et al. 2021. "FORWORK - Fostering Opportunities of Refugee Workers." AEA RCT Registry. December 27.
Experimental Details


The intervention was originally planned to run over 11 months and to offer language and skills assessments combined with short language and vocational courses to develop an individualized integration plan. It consists of four phases:
1) Mentoring to facilitate beneficiaries’ access to different services and the creation of an individual work integration plan, with mentors and cross-cultural facilitators working as a team to accompany participants for about one year
2) Language and civic education workshops
3) Targeted vocational training
4) Apprenticeship experiences and job placement services.

The Skills Profile Tool for Third Country Nationals, being developed by the DG Employment, Social Affairs and Inclusion of the European Commission, will be used by mentors and cross-cultural facilitators in the initial phase of the individual plan to learn about the individual, her skills and experiences.

* The COVID-19 crisis had a severe impact on how the project phases unfolded: first, the national lockdown imposed by national authorities (March to May 2020); second, the social distancing measures affected a wide-range of in-person activities all throughout 2020, 2021 and the beginning of 2022. For this reason, the original planning of the activities saw a significant delay.
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Labour market outcomes:
- Employment rates (overall and by type of contract) from administrative and survey data
- Survey measures of job search effort, wages, and illegal employment
Primary Outcomes (explanation)
We intend to measure employment outcomes combining administrative data and qualitative indicators from surveys designed for the purpose of this study. The partnership with the Agenzia Piemonte Lavoro will ensure access to administrative data from SILP (Sistema Informativo Lavoro Piemonte), the regional labour market information system of Public Employment Services (PES). This SILP database tracks the population of all unemployed individuals who registered with PES in the region as well information that private employers are legally bound to file in order to start, modify, or terminate a job (including apprenticeships). In addition, it provides personal information on age, educational attainment, and previous work experience in Italy. This will provide us with linked employer/employee data updated in real time for all work episodes in Piedmont, ensuring month-to-month monitoring of labour market outcomes for refugees. This data will be complemented by a post-treatment survey of beneficiaries, where additional information about the type of job held (length, intensity, salary) will be retrieved. The survey will also investigate the extent of informal or irregular job contracts, a widespread phenomenon in Italy that cannot be captured using SILP data.

Secondary Outcomes

Secondary Outcomes (end points)
- Survey measures of social integration (knowledge of Italian, level of trust (general and towards natives), frequency of contacts (general and towards natives)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The evaluation will be conducted via a clustered randomized control trial with blocking. The statistical units of analysis are individuals. Clusters are defined by individuals belonging to the same reception centre managing entity. Managing entities will be stratified in mutually exclusive “blocks” defined using the administrative geography of the Piedmont region.

1. Reasons for clustering
Despite FORWORK being tailored around the specific needs of each individual, there are good reasons to believe that the unit of intervention (that is, the level at which “treatment” is randomized) should not be the individual. Interactions among asylum seekers may generate externalities in the same centre, and particularly so if its size is small. Moreover, in Piedmont a single entity (charity, NGO, private organization) typically manages more than one reception centre. Reception centres under the same managing entity are normally very close to one another and served by the same staff.
For these reasons, randomly denying access to the treatment within the same centre, or between different centres managed by the same entity may be politically sensitive and ethically not acceptable, and fraught with operational problems exactly because of social interactions. Because of this, we will consider the reception centres’ managing entity as the statistical unit of intervention. All (eligible) individuals in the centres selected will be included in the study sample.

2. Eligibility
2a Area selection (LMAS - Labour Market Areas)
In order to foster the comparability of the job opportunities faced by the experimental and the control group, we will limit the scope of our evaluation within the province capital LMAs and the Chieri LMA. LMAs are aggregations of adjacent municipalities defined by the Italian National Statistical Institute based on workers’ commuting patterns. They identify a finely defined geography with similar labour market opportunities and economic conditions. We adopt here the 2011 version provided by ISTAT (see, accessed February 20, 2019).
2b Structure eligibility criteria
All managing entities operating in selected LMAs are included in the sample, conditioning on the characteristics of their reception centres. Eligible reception centres need to comply with the following criteria: a) to be located in a selected LMA; b) to be located not farther than a 30 minutes driving distance from the closest Public Employment Services PES, in municipalities served by public transportation; c) to host no more than 30 guests.
2c Individual guest eligibility criteria
In eligible reception centres target asylum seekers need to be a) 18 years old or older and b) should be waiting for their first meeting with the territorial commission, which means being in the early stages of applying for refugee status. This last restriction has been included, in agreement with local Prefectures, to avoid as much as possible sample shrinking during the year of the intervention.
Experimental Design Details
Randomization Method
Randomization was performed before the start of the intervention on a dataset containing, for each asylum seeker hosted in Piedmont socio-demographic characteristics (age, sex, and nationality), asylum status, address of the reception centre, name of the managing entity. First, we built 7 blocks designed to take into account geographical location of the managing entities. Most managing entities are comparable in terms of size (number of reception centres and number of guests) with the exception of two entities located in the province of Turin. The blocks are the following: 1) Piedmont, South-East (province of Cuneo); 2) Piedmont, South-West (provinces of Alessandria and Asti); 3) Piedmont, North-East (provinces of Vercelli, Verbano-Cusio-Ossola and Biella); 4) Turin, city centre and close neighbourhoods; 5) Turin, countryside; 6) Turin, mixed (entities with centres in both countryside and in the city); 8) Turin, large managing entities. The last block have been built to assure the randomization against an unlucky draw of the two outlier entities in terms of size. Within each block we drew treated and control units in proportion 1:1.
Randomization Unit
Reception centres’ managing entities
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
58 Eligible entities, managing a total of 262 reception centres
Sample size: planned number of observations
1262 eligible individuals within eligible centres
Sample size (or number of clusters) by treatment arms
29 treated and 29 control managing entities
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Power calculations suggest that involving approximately 1,200 (600+600) individuals in 58 managing entities equally split into “treatment” and “control” arms, allows to detect a treatment effect of about 10 percentage points.

Institutional Review Boards (IRBs)

IRB Name
Boccony University
IRB Approval Date
IRB Approval Number


Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information


Is the intervention completed?
Data Collection Complete
Data Publication

Data Publication

Is public data available?

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials