Underreporting of high-risk water and sanitation practices undermines progress on global targets

Abstract

Water and sanitation indicators under the Millennium Development Goals failed to capture high-risk practices undertaken on a regular basis. In conjunction with local partners, fourteen rounds of household surveys using mobile phones with a customized open-source application were conducted across nine study geographies in Asia and Africa. In addition to the main water and sanitation facilities, interviewees (n = 245,054) identified all water and sanitation options regularly used for at least one season of the year. Unimproved water consumption and open defecation were targeted as high-risk practices. We defined underreporting as the difference between the regular and main use of high-risk practices. Our estimates of high-risk practices as the main option matched the widely accepted Demographic and Health Surveys (DHS) estimates within the 95% confidence interval. However, estimates of these practices as a regular option was far higher than the DHS estimates. Across the nine geographies, median underreporting of unimproved water use was 5.5%, with a range of 0.5% to 13.9%. Median underreporting of open defecation was much higher at 9.9%, with a range of 2.7% to 11.5%. This resulted in an underreported population of 25 million regularly consuming unimproved water and 50 million regularly practicing open defecation. Further examination of data from Ethiopia suggested that location and socio-economic factors were significant drivers of underreporting. Current global monitoring relies on a framework that considers the availability and use of a single option to meet drinking water and sanitation needs. Our analysis demonstrates the use of multiple options and widespread underreporting of high-risk practices. Policies based on current monitoring data, therefore, fail to consider the range of challenges and solutions to meeting water and sanitation needs, and result in an inflated sense of progress. Mobile surveys offer a cost-effective and innovative platform to rapidly and repeatedly monitor critical development metrics.

Fig 1. Underreporting of high-risk practices.

Main and regular estimates of (a) unimproved water, and (b) open defecation.

Fig 2. Availability of water and sanitation options.

Number of sources used by respondents for (a) drinking water, and (b) sanitation.

Fig 3. Likelihood of using non-high-risk practices.

(a) Reliability of the main water option, when the source is improved. (b) Frequency of using a non-OD source as the main sanitation option.

Fig 4. High-risk practice as a function of wealth.

Local polynomial smoothing function showing (a) unimproved water use, and (b) open defecation, over wealth score in Ethiopia.