In 2004, South African National Department of Human Settlements (NDOHS) launched Upgrading of Informal Settlements Programme with the goal to facilitate structured improvements of informal settlements and make beneficiaries' living conditions better. With the technical assistance from the World Bank, NDOHS conducted a series of impact evaluations to assess the effects of the Upgrading of Informal Settlements Programme (UISP) interventions in Free State, Limpopo and Gauteng Provinces. The research was designed to reliably identify causal links between the rollout of UISP and the outcomes of interest driven by policy prescriptions (as well as broader concerns) for the program. The study areas chosen allow for four comparisons. In Limpopo, the design allows for estimating the impact of relocating households from an informal settlement with no services (Eastern Disteneng), to a formalized greenfield site with comprehensive services and supporting community facilities (Extension 44/76). Household level survey data was collected from a sample of 432 households from Extension 44/76 (treatment group) and 726 households in Disteneng (control group). In Free State, the relative impacts of being provided with a fully serviced stand (Bloemside) to being provided with a partially serviced subsidized house on the site of the original informal dwelling (Grasslands) are compared. By exploiting the phased approach to the study, estimates can also be made on the long-term impacts of being provided with a subsidized home, by comparing Grasslands II residents who have been living in their upgraded homes for three or four years to the neighboring Grasslands III residents who have had their subsidized homes for one to two years. Researchers surveyed 1,014 households: 370 households from Grasslands II, 289 from Grasslands III and 355 from Bloemside. In Gauteng, the impact of fully upgrading an area compared to a partial upgrade (less than 50% households receiving housing and electricity) can be estimated. The study exploits the phased roll out of Extensions upgrades to compare the extensively upgraded area of Extension 1 (398 household surveyed) to the partially upgraded areas of Extensions 2 and 3 where 905 households were surveyed. Datasets from Free State and Limpopo provinces are documented here.
