Reflections on monitoring and over-reporting in sanitation

Printer-friendly versionPrinter-friendly version

Sanitation is probably one of the sectors with the highest levels of overreporting. It is clearly the case of India, where 4/5 of the toilets reported during the Total Sanitation Campaign were found ‘missing’ in the 2011 Census. The monitoring system, intended to measure the sanitation coverage increase, was found to actually reflect funds disbursement. The perfect combination for overreporting is: multi-level pressures and interests to show progress complemented with a lax verification system. This is widely known, but efforts to tackle overreporting tend to focus only on the second part of the cocktail; smarter technology or new verification procedures that limit the exaggeration of results.

This is happening in the Indian state of Madhya Pradesh, where the monitoring system is now based on photographs. The idea is that in order to change the status of a household to ‘having latrine’ in the database, a photo of the latrine and a household member (such as the one presented) has to be uploaded. The system is said to have already started reducing the level of overreporting. And there are plans to use a more advanced system, where the photographs would be geographically positioned through GPS geo-tags.

As said, as valuable as these innovations are, it is also important to also address pressures and interests to show progress and inflate data: rigid targets, technocratic implementations, corruption, etc. And this is sadly not happening so much in Madhya Pradesh, at least to what concerns targets, as I could clearly see in my recent field visit there.

The state has started a very promising sanitation campaign called Maryada Abhiyan (Dignity Campaign) that aims to be community-led and women-led and focuses on collective behaviour change.

However, monitoring is still construction oriented and the pressure to show progress remains. For instance each block (having around 100 villages) has a paid sanitation coordinator, in charge of implementing the campaign. Recently, all the block coordinators received a letter as the one in the picture, where they were scolded for not advancing towards their targets, and warning them that if they did not get 500 toilets built in the coming month, they would not get their salary.

It is true that there are some block coordinators which are not taking their work seriously, but a letter with unrealistic short-term targets and threats won’t be of much use in changing this. Even more worrisome, it will disturb the work of motivated block coordinators (there are many!!) that are focusing on behaviour change and demand creation and letting construction to progressively pick up.
In some cases, the consequence of the letter –if taken seriourly– may be the rushed construction of 500 latrines, no matter much in whose house and with what construction quality, leading to problems in terms of appropriation and sustainability. In most of the cases, however, it will lead straight to overreporting.

Part of the challenge of monitoring sanitation is that it is a complex issue that takes time and that relates to behaviour and not to infrastructure. However, sanitation policies and monitoring many times do not effectively acknowledge and embrace this complexity, and set too rigid targets and technocratic systems of accountability and verification. As said before, that perspective may give space for making verification more stringent (through smart technologies), but won’t allow tackling the problem of overreporting completely, as interests and pressures for exaggerating results will remain.

Finally, it is also important to reflect in the monitoring system the principles of the policy. For instance, if awareness raising is as important as toilet construction, there should be a balance between indicators of coverage and indicators of activities promoting behaviour change. Otherwise, if coverage is the single cornerstone of the monitoring system, the campaign will drift towards toilet construction mode. Remember: What we don’t count, doesn’t exist. What we don’t monitor, won’t happen.

Date: 24 February 2014
Contributors: 
Region: 
Country: 
India