Tackling open defecation

Within five months of coming to power, Prime Minister Narendra Modi, launched Swachh Bharat Abhiyan, a project to modernise sanitation within the next five years. In light of that commitment, a new study, led by Raymond Guiteras of University of Maryland and James Levinsohn and Mushfiq Mobarak of Yale University, and implemented by Innovations for Poverty Action, assumes special significance for India. Conducted in northwest Bangladesh, over an area where 50 per cent of the population had access to a hygienic latrine, the study tested three different approaches that are commonly used in the development sector for increasing the use of hygienic latrines. Villages either received a community motivation program, subsidy vouchers with the community motivation program, information and technical support, or none of the above. By comparing outcomes in latrine coverage, investment in hygienic latrines, and open defecation between different groups, researchers were able to assess the impact of the different strategies. The one combining subsidies with targeted community allocation seems to have worked best among the communities.

In an email interview with Monobina Gupta, Raymond Guiteras and Mushfiq Mobarak elaborate on their findings and their implications for India’s Swachh Bharat project:

Q: According to the 2011 census in India, only 46.9 per cent of India's 24.66 crore households have a latrine facility, which is in stark contrast to a much lower 3% open defecation in Bangladesh. Given these sharp differences, can you please elaborate on the significance of your study in influencing India’s sanitation policy. 

A: It is true that Bangladesh has lower rates of open defecation than India, although our study area had relatively high rates of OD. There are a number of other important differences as well, just as there are likely to be differences within India. That said, our study is relevant for India in several ways. First, it shows that there may be limits to what motivation and education can achieve among poor populations. Second, many practitioners are concerned that there is a conflict between subsidies and intrinsic motivation. While this may be true in some contexts, our study provides a counter-example -- in fact, in our case, subsidies appear to "crowd in" intrinsic motivation. Third, we show that it is important to conduct rigorous, independent impact evaluations before launching any major intervention. Given the success of information and motivation in reducing open defecation elsewhere in Bangladesh, it would have been quite reasonable to believe that a subsidy-free program would have been successful in Tanore, but this turned out not to be the case. Similarly, it would be very risky to launch an expensive subsidy campaign without understanding the conditions necessary for success.

Q: One of the outcomes of your study shows that education alone doesn’t enhance motivation to access hygienic latrines. But, education plus vouchers do increase motivation and, in turn, access to latrines. Please elaborate.

A: The combination of strategies – behaviour change programming, plus subsidies targeted to the poor is the strategy that works best.  We think there are multiple constraints here: information needs to be provided to the whole community about this community-wide problem, and cost concerns need to be addressed for the poorest segments of the population, for everyone to feel comfortable that they will all jointly move to address the problem.

Q: Another significant pointer from your study is that even those who do not have vouchers tend to buy latrines if their neighbours have one. So should the strategy in India also primarily revolve around vouchers and in allocating them in a coordinated manner?

A: I think it's likely that a carefully designed voucher program could be useful in India as well. However, the specifics of how to allocate vouchers -- for example, to what share of the population, in what amount, and to whom -- is difficult to say. These are questions we continue to study from our Bangladesh experiment, and similar studies conducted in diverse areas of India would provide useful guidance. Certainly, vouchers are much more likely to be successful if they are combined with a sustained motivation and education effort that is informed by local culture.

Q: Would you agree that prevalence of the caste system and untouchability present a problem that is peculiar to India and that it is a hindrance in the way of ensuring total sanitation?

A: Our research results suggest that network relationships in a community are very important determinants of sanitation choices and norms: each household’s investment decisions are inter-linked with other households in the community. This suggests that any characteristic of the community in India that affects inter-household relationships, such as caste identity, are likely to be important determinants of sanitation choices.

There is a lot of excellent research from the r.i.c.e. institute (research institute for compassionate economics, http://riceinstitute.org/) about exactly these issues, and their research suggests that these have been important constraints.
Here are two significant papers from r.i.c.e.:
http://riceinstitute.org/research/culture-and-the-health-transition-understanding-sanitation-behavior-in-rural-north-india/
http://squatreport.in/wp-content/uploads/2014/06/SQUAT-one-page-memo.pdf

Q: Do sanitation strategies have to be country-specific and in consonance with their different cultural and social contexts in order to be effective on the ground? In that case could you please highlight some of the India-specific strategies that could work on the ground?  

A: Some aspects of behaviour are reasonably generalisable: demand curves usually slope down; people seem to be influenced by the behaviour of their neighbors. Some aspects are more specific. It's important to take what we can be reasonably sure is general, combine that with insights gained from local knowledge and experience, and -- most important of all -- rigorously test competing strategies to find out what works and why. 

One example of a strategy that has worked well in India is Gram Vikas's "Rural Health and Environment Program" in Odisha. This is a very intelligently designed program that combines water and sanitation for the rural poor. It is neither easy nor fast - Gram Vikas often has to work with a village for as long as two years to build the motivation and consensus necessary for the program to be effective - but in my opinion it's an excellent example of combining motivation and subsidies. I have a paper with Tom Clasen (Emory University and LSHTM), Esther Duflo (MIT) and Michael Greenstone (U. Chicago) that describes RHEP in greater detail. We find that it was very effective at reducing diarrheal disease, and these benefits were sustained for five years or more. Of course, there are other examples of less-successful interventions.

Paper: "The Short- and Medium-Run Impact of Clean Water and Sanitation on Diarrhea in Rural India", Maryland Population Research Center Working Paper PWP-MPRC-2015-008. April 2015.http://papers.ccpr.ucla.edu/abstract.php?preprint=1125