Methodology

Methodologically, the TAKE project makes use of a range of data sources and innovative research methods.

First, we will collect new survey data which allow for a proper investigation into NTU in Belgium. Existing administrative data have important shortcomings for studying NTU: usually they do not contain sufficient information for assessing eligibility for those that do not receive the policy measure concerned; and they contain limited information for assessing the causes of NTU. In contrast, general-purpose surveys such as SILC do contain more elaborate income information for all respondents, but the data are not sufficiently detailed for assessing eligibility very precisely, and the share of vulnerable groups (e.g. recipients of social assistance) in the total sample tends to be rather low limiting the possibilities for detailed analysis of their characteristics. Finally they usually do not contain information on the reasons for NTU. Therefore, in this project we will carry out a new survey, which is expressly tailored to the needs of making an integrated in-depth study of the size, characteristics, causes and consequences of NTU of policy measures targeted at vulnerable groups. A sample will be drawn from administrative data available to the Crossroads Bank on Social Security (CBSS).

Second, for identifying NTU, one also needs a model which replicates the eligibility tests on the basis of the variables in the survey. The existing static tax-benefit microsimulation model MEFISTO will be extended and refined for doing so. MEFISTO has been developed in the framework of the IWT-SBO project ‘FLEMOSI : a tool for ex ante evaluation of socio-economic policies in Flanders’ (2010-2013). Given that the MEFISTO-model runs on EU-SILC eligibility tests are currently designed to the data availability in EU-SILC. Within the framework of TAKE, MEFISTO will be adapted to the TAKE survey and extended with policies that are currently not covered in MEFISTO (e.g. reduced tariffs for mobility and utilities). On the basis of the refined and extended MEFISTO model we will be able to identify in a first step NTU in the TAKE survey, such that we can study its characteristics, size, determinants and direct budgetary and social impact. In a second step, we will make use of MEFISTO to evaluate the impact of changing the eligibility tests in such a way that they would facilitate lower levels of NTU.

Third, TAKE will make use of a field experiment to test the effect of various triggers and encouragements on the uptake of the “Beneficiary of Increased Reimbursement” (BIR) statute within health insurance. Within the Belgian Health Insurance, people living in low-income families with an income below a certain threshold can be assigned the BIR statute, after which they pay lower co-payments. For persons for whom sufficient income information is available from administrative data, the BIR statute is assigned automatically. For persons for whom this is not the case, a supplementary income test by the sickness funds is required. The uptake of the BIR statute by this latter group is very low. Between 2015 and 2017 the Belgian sickness funds will organise a large scale campaign to stimulate the uptake of the BIR statute by persons for whom automatic enrolment is not possible. To this end, the Belgian tax administration have provided them with information on who of their members had an income below the BIR eligibility threshold. The sickness funds will invite these potential BIR beneficiaries for an income test between August 2015 and December 2017. As they are free to choose the way in which members are contacted (letter, e-mail, telephone call, home visit) the members will be exposed to various triggers aimed at fostering take-up of the BIR statute. This provides us with a unique opportunity to investigate the effect of various triggers on take-up of the BIR statute.

Fourth, in order to study the institutional context, we will develop TAKE_ISSOC (working title), i.e. a structured and searchable database which contains the details of eligibility tests of the social benefits covered by the project and how they are implemented in practice. Among others, the database will allow for easily comparing how definitions of ‘household composition’, ‘household income’, etc. vary across welfare benefit systems in Belgium. In addition, TAKE_ISSOC will cover measures that public administrations currently take to reduce NTU, and current monitoring practices and their outcomes with regard to NTU. The TAKE project will gather similar information on Sweden and the United Kingdom in order to identify and assess good practices in two countries that have been generally recognised as frontrunners in the use of administrative data and actively monitoring and reducing NTU of social benefits. This should lead to the identification of possible best practices in monitoring and reducing NTU.

Fifth, TAKE will exploit existing administrative data collected by public agencies to investigate NTU of employment subsidies targeted at the low-skilled and other vulnerable groups in the period 2004-2013. Longitudinal administrative data from the National Social Security Office (NSSO) and from the National Employment Office (NEO) will be used for doing so. Whereas the NSSO dataset will be used to study NTU from the perspective of employers, NEO data will be used to estimate NTU from the perspective of the unemployed. The estimation of NTU rates will then allow us to investigate the causes of NTU of wage subsidies with a focus on employer and employee characteristics. In particular, TAKE aims to examine if certain types of employers exhibit higher NTU rates than others and to measure the effectiveness of targeted measures in terms of their main objective, increasing employment of their respective target group. In particular, we will examine the length of the employment relationship once a subsidy has been granted and compare it to a “control” group which does not benefit from the (same) subsidy. Measures which exhibit relatively low take-up rates due to the difficulty of finding an eligible candidate might still generate stable employment relationships while those with higher take-up rates might not. The combination of take up rates and effectiveness will allow us to contribute significantly to the debate on the effectiveness of targeted wage subsidies in generating employment for the low-skilled.

Six, to better understand the reasons for NTU of employment subsidies, the existing survey on administrative charges will be used. This survey measures the costs of administrative charges related to employment legislation incurred by employers in Belgium. TAKE will be able to gather valuable information on non-take up of wage subsidies by adding specific questions to this survey.

In summary, the TAKE project aims at making a significant contribution to (1) the theoretical underpinnings of research into NTU; (2) the available research infrastructure for studying NTU in Belgium; (3) insights into the characteristics, size, consequences and remedies for NTU; (4) an improved monitoring of NTU and knowledge exchange between the relevant actors in the field. TAKE attempts to make significant contributions not only to the international scientific literature and debate on NTU, but also to a better informed public debate and policy making process on NTU and how to improve participation in public transfers.

Advertisements