The Use of Discretion by Client Managers in Social Assistance Practice

Objectives: To study the variation in case managers’ treatment of claimants concerning eligibility to benefits and activation practices. Design: A factorial survey in which case managers are presented with 12 stories... [ view full abstract ]

Objectives: To study the variation in case managers’ treatment of claimants concerning eligibility to benefits and activation practices.

Design: A factorial survey in which case managers are presented with 12 stories (vignettes) of claimants. The vignettes consist of 12 experimentally varied categories (e.g. gender, mental health, attitude…). Vignettes are treated as the first level in the multi-level analysis, respondents as the second level.

Setting: Seven social assistance agencies in Belgium.

Subjects: 170 case managers (90% response rate of all case managers in the municipalities).

Main outcome measures: The likelihood of clients: (1) receiving a financial benefit, (2) being withdrawn if declining a job offer and (3) not receiving activation offers.

Results: Large variation exists between the case managers participating in the survey, both in a general way (intercepts) and in responses on specific client characteristics (slopes). The VPC’s of the multi-level null models range from 0,35 to 0,5, meaning that more than one third of the variation is explained by respondent and not by client characteristics. Respondent’s characteristics that may explain this variation are the municipality where one is working, the gender of the case manager and resemblance between the respondent and the client. Furthermore, specific client characteristics influence the variation between respondents. Respondents react differently to client characteristics such as parenthood, addiction, housing situation, ambition and attitude.

Conclusions: This study shows, for the first time quantitatively, that similar clients do not receive a similar treatment from different case managers in different municipalities. This might raise concerns about the fairness of the social assistance offered.