Disentangling the dynamics of social assistance: A linked survey-Register data cohort study of long-term social assistance recipients in Norway

• Published in: PloS one Vol. 15; no. 3; p. e0230891
• Main Authors: Heggebø, Kristian, Dahl, Espen, van der Wel, Kjetil A
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 27-03-2020; Public Library of Science (PLoS)
• Subjects: Assistance; Child & adolescent psychiatry; Child welfare; Childhood; Children; Cohort analysis; Data collection; Disabilities; Drug abuse; Dynamics; Economic conditions; Education; Employment; Families & family life; Family relations; Finance; Financial support; Health; Health behavior; Health status; Immigrants; Labor; Labor market; Medicine and Health Sciences; People and Places; Polls & surveys; Probability; Public assistance; Randomness; Self sufficiency; Social dynamics; Social networks; Social policy; Social Sciences; Social systems; Sociodemographics; Statistical analysis; Studies; Surveys; Taxation; Variables; Withdrawal; Young adults; Norway; Sweden; Finland
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0230891

Social assistance is a means-tested benefit that is supposed to be a short-term, temporary economic support. Understanding why some individuals are in repeated or continuous need of social assistance is thus of obvious policy relevance, but the dynamics of social assistance receipt remain poorly understood. In 2005, a survey among long-term recipients of social assistance in Norway collected data on (a) childhood disadvantages, (b) health status, (c) health behaviors, (d) psychological resources, and (e) social ties, in addition to basic sociodemographic information. This rich survey data has been linked with tax register data from 2005-2013, enabling us to explore the detailed characteristics of long-term social assistance recipients who are unable to reach financial self-sufficiency. Results from linear probability models show that surprisingly few of the 28 explanatory variables are statistically associated with social assistance dynamics, with two important exceptions: People with drug problems and immigrants both have a much higher probability of social assistance receipt. Yet overall, it is challenging to 'predict' social assistance dynamics, indicating that randomness most likely plays a non-negligible role. The 28 explanatory variables do a far better job in predicting both labor market success (employment), labor market preparation (work assessment allowance), and labor market withdrawal (disability benefit utilization). Thus, there seems to be something distinctive about the processes leading to continued social assistance recipiency, where randomness could be a more influential force.