Report (R134)

Sustainable Total Sanitation – Nigeria: second CLTS impact evaluation report

Date: 26 October 2017
Publisher: The IFS

In this report, we analyse the impacts on sanitation uptake achieved by the Community Led Total Sanitation (CLTS) intervention, almost two years after its implementation in 2015, in the Nigerian states of Ekiti and Enugu. We report findings from an analysis of data collected in the second rapid assessment of households (RA2), carried out during April 2017, almost two years after implementation. We also discuss these results in contrast to the findings from the previous data collection wave, from December 2015 (RA1), around six to twelve months after implementation. Evidence from the cluster randomised controlled trial (RCT) shows the following.

  • By RA2, CLTS households exhibit 4 percentage points (pp) lower prevalence of open defecation (OD) than control households.
  • However, the small and positive CLTS treatment effects on toilet uptake over the whole sample (around 4pp) observed in the RA1, which were driven by an increase in uptake in treatment areas, cease to be detectable due to a relatively faster increase in toilet coverage in control areas by the time of the RA2.
  • Results on toilet uptake show almost no variation by type of household by the time of the RA2. At RA1, CLTS treatment effects were higher (around 6–7pp), among households with female heads, among households whose heads had low levels of education, and among households with seniors. These higher effects are detectable only for the last group by RA2. Equally, households with children, households with no debts and households with low asset wealth, who also appeared to have reacted strongly to the CLTS intervention by RA1, are indistinguishable from the rest of our sample by RA2.
  • We see no evidence of a change in the share of improved/unimproved toilets in the study sample because of the CLTS intervention.

In summary, most of the impacts observed six months after the intervention (RA1) become undetectable two years after it (RA2). This could be due to programme spillovers within communities that contained both CLTS and control households. Accounting for this in two different ways, we find the following.

  • Estimated CLTS impacts on toilet uptake and functioning toilet uptake measured at RA1 are 7pp and 5pp, respectively, if spillovers are accounted for by a specific methodology, showing an increase in magnitude and significance.
  • Estimated CLTS impacts on functioning toilet uptake become statistically significant and of a magnitude of 6pp by RA2, more than two years after the intervention took place.
  • Future research will aim to shed light on how these spillovers operate, and on their implications for both research and policy design.