Data Portal

Citation Information

Type Working Paper - Southern Africa Labour and Development Research Unit
Title Assessing the incidence of public works programmes: Using propensity score matching techniques to assess the poverty targeting of employment in two public works programmes in South Africa.
Author(s)
Issue 31
Publication (Day/Month/Year) 2009
Page numbers 1-47
URL http://www.saldru.uct.ac.za/home/index.php?/component/option,com_docman/Itemid,32/limitstart,30/
Abstract
This paper explores the socio-economic identity of Public Works Programme (PWP) participants in two programmes in South Africa, in order to establish the incidence of PWP participation, a question which is central to assessing the social protection impact of PWPs, but which is frequently omitted from programme analysis, particularly in sub-Saharan Africa. The paper focuses on an analysis of the demographic and socio-economic characteristics of PWP participants. As in many PWPs, no baseline data on participants in these programme were collected. Therefore, it is not possible to ascertain a priori who the beneficiaries of the programmes are, a situation which fundamentally challenges any attempt to or to assess incidence or the social protection impact of such an intervention. The research interrogates the assumption that the 'less eligibility criteria' central to the design of PWPs (the work requirement and low wages) will lead to participation of the poorest, thereby reducing the likelihood of inclusion errors, attempting first to ascertain who the participants in the programmes are. The question is explored using survey data gathered in 2003 on two case study PWPs implemented simultaneously in South Africa, which adopt different design and targeting approaches. Programme incidence is then considered in relationship to targeting and programme objectives, and the conclusion drawn that in order for PWPs to reach the poorest in a given community, reliance on self targeting through the work requirement and a low wage is not adequate, and explicit targeting measures are needed during participant selection.

Related studies

»