r.i.c.e.

Here at rice, we write and talk much more about rural sanitation than we do about urban sanitation. This is partially because we know more about and have been more exposed to poor people’s lives in rural India; is partially because local area open defecation is easier to capture in survey data in rural places; and is partially because rural people in India are poorer, on average, than urban people in India. That said, any divide like rural-urban of course averages over many differences within these groups!

Sanitation is still a big problem, even in urban India. Fully 17.5 percent of urban households in India – and 9.6 percent* of urban households with “pucca” or well-constructed walls and floors – defecate in the open, according to India’s 2005 DHS. Moreover, the average urban Indian household is exposed to 17.6 percent of its neighbors defecating in the open, or 13.1 percent for those in the pucca houses. So, even if you are living in urban India and you do not defecate in the open, 9.5 percent of your neighbors do.

Open defecation harms children because they come into contact with fecal pathogens. This means that population density matters: open defecation is more dangerous where population density is higher. That’s why the same difference in open defecation is associated with a greater decline in child height in India than in sub-Saharan Africa. Similarly, again and again in data from countries around the world, the same increase in open defecation causes bigger problems for children’s health in urban places than in rural places. In economists’ jargon, the marginal social costs of an extra bit of open defecation are greater in urban than in rural places.

This means that, all else equal, policymakers with limited power should concentrate their efforts on reducing urban open defecation first. But is all else equal?

Jeff Hammer, an economist at Princeton University, often observes that when policy-makers are choosing among policy options, they must consider two dimensions. First, how bad is the problem they are trying to fix? Second – but equally important – how good is the government likely to be at fixing it, and how much harm might they cause in the process?

We have seen that reducing urban defecation is – pound for pound – more helpful than reducing rural open defecation. But which can a low-capacity poor country government expect to more easily accomplish? At least in north India where we work, reducing rural open defecation is mainly about behavior change. In other words, the binding constraint is not often not having a latrine, but is using it; some people routinely fail to use even free latrines, in part because they simply do not believe that defecating in the open is bad for anybody’s health. As Diane wrote about yesterday in the context of infant feeding, it can be very hard to change behaviors merely by spreading new information.

So, reducing rural open defecation is difficult. Of course, reducing urban open defecation isn’t easy, either. But in many cases, it is mainly a construction project: if you give adequate sewers to people living on top of their neighbors, then they will very often use them. (Finding a place to put sanitation in urban places can be a challenge).

Nobody would say that the Indian state is particularly great at construction, but the government does routinely accomplish construction projects, even if a lot of money goes missing. Indeed, exactly because it creates the opportunity for a lot of money to go missing, construction projects are a large fraction of the “development” that low-capacity governments manage to do. It may be worth spending a little more than things really cost in order to improve the health and human capital of the next generation.

I am often asked by well-meaning policy-makers what I recommend that they do to reduce rural open defecation. My honest answer is that I don’t know, and that we should be experimenting like crazy to find out.

But if it is, indeed, the case that construction projects can – with some perhaps low but non-zero elasticity – reduce urban open defecation, then maybe that is where efforts should be focused? Taking the argument one step further, maybe that is how we at rice should be targeting our not-infinite advocacy opportunities and resources?

I don’t know. There are many empirical questions about cause and effect in rural and urban places. I do not know exactly how much or how little building nice toilets with sewers in urban India would help. I do not know whether this would be very, very difficult, or merely very difficult. However, it wastes opportunities to improve lives if our policy advocacy recommends only things that governments cannot do (or cannot do without making matters worse). Does this principle recommend a focus on urban sanitation?

* An honorable mention to Diane, who guessed 10 percent.