Syndicate content

Why should Governments Spend on Sanitation?

Shanta Devarajan's picture

A puzzle:  Sanitation is one of the most productive investments a government can make.  There is now rigorous empirical evidence that improved sanitation systems reduce the incidence of diarrhea among children.  Diarrhea, in turn, harms children’s nutritional status  (by affecting their ability to retain nutrients).  And inadequate nutrition (stunting, etc.) affects children’s cognitive skills, lifetime health and earnings.  In short, the benefits of sanitation investment are huge.  Cost-benefit analyses show rates of return of 17-55 percent, or benefit/cost ratios between 2 and 8.

But if the benefits are so high (relative to costs), why aren’t we seeing massive investments in sanitation?  Why are there 470 million people in East Asia, 600 million in Africa and a billion people in South Asia lacking access to sanitation?  Why are there more cellphones than toilets in Africa?

This is a puzzle because nobody cares more about children than their parents.  Why aren’t they investing in toilets, so their children will be sick less often, be better nourished, learn more in school and become more productive adults?  One answer is that they don’t know about all these benefits from sanitation.  But the evidence has been around for decades.  Also, we should never underestimate poor people’s knowledge.  We used to say the same thing about education—poor people don’t send their kids to school because they don’t know the benefits—until we found out that illiterate parents know more about the benefits of education than literate ones.

Another response is that poor people don’t have the money.  But they are making decisions every day on how to spend their money—on food, clothing, etc.  The returns to sanitation expenditures are higher than some items that they’re already spending on. 

Even if households don’t invest enough, why doesn’t the government step in?  17 to 55 percent rates of return are better than many, if not all, public projects.

Here is the resolution of the puzzle:  This evidence is not what is relevant for government investment decisions.  Finance ministers are constantly besieged by advocates for one activity or another who say, “My sector has all these effects on growth and poverty reduction, so you should spend more on my sector.”  This argument is wrong.

Why?  Because governments have limited resources.  Any expenditure by government on sanitation will come at the expense of something else (or higher taxes)—such as roads or education or health.

So how should governments decide what to spend on?  They should spend on those things that the private sector will not spend on, or will not spend enough on. The classic example is an externality—where the benefits or costs occur to people other than the person undertaking the action (such as when an upstream firm pollutes downstream villages). Society is better off if government steps in and taxes the externality provider.

Sanitation, or its converse, open defecation, is a negative externality. People who defecate in the open not only harm their own children, but other people’s children.  Their incentive to invest in sanitation is less than the costs.  In Maharashtra, India, a village-level sanitation program showed that children from households that didn’t adopt latrines were statistically better off (taller, etc.).  In rural India, the reduction in diarrhea incidence from others’ adopting sanitation is about half of the effect of your own household’s adopting it.

This is the strongest case for public spending on sanitation.  Furthermore, ignoring the externality argument has serious consequences.  Governments typically overspend on private goods, such as high-end medical care, at the expense of goods with externalities, such as sanitation.  Rich people and medical unions have political power to lobby for spending on curative health care; others lobby for fuel subsidies (another private good).  Poor people’s voice for sanitation goes unheard.  This was also the case in today’s rich countries—in the 19th century.  William Henry Harrison, the tenth president of the U.S., died of enteritis, contracted from fetid marshes near the White House.  Unfortunately, today, many of our cities in Africa and South Asia look like Dickensenian England—or worse.

Another consequence of ignoring the externality is that implementation of sanitation projects can fail.  India’s sanitation program, where they gave out free toilets to households, wasn’t working.  The reason was that people were using the toilets for other purposes, such as grain storage.  Meanwhile, Bangladesh had a very successful program, run by an NGO, financed by government.  The difference was that in Bangladesh, they were giving the toilets to the community, and asking them to allocate to individual households.  They recognized that by giving toilets to people, you weren’t solving the problem of the externality.  But giving it to the community enabled the community to monitor whether people were using the toilets.  The community as a whole had an interest in everybody using the toilets (rather than open defecation).  The good news is that Maharashtra now has a successful total sanitation program.  So much so that you now see signs saying, “We will not give our daughters in marriage to someone from a village that doesn’t practice total sanitation.”


Submitted by Tom B. on

Very true. In the ongoing Rwanda PA, we have a chapter on stunting (which remains exceptionally high in Rwanda). The single main contributor to the high incidence of stunting is not a monotonous protein-poor diet ("unsufficient dietary diversity"), inadequate meal fequency, or even poverty. The single biggest contributor is unimproved sanitation, which explains why one of the poorest districts has stunting rates well below that of the wealthiest districts: For some reason sanitation is really good there.

Submitted by Shanta on

Tom, thanks for the comment. It would be fascinating to study this district's success at reducing stunting and in improving sanitation at some depth. Are there other factors in relation to stunting (for instance, some ethnic groups are taller)? Also, why did this community improve sanitation while others didn't? Was there something like the Total Sanitation Program of Bangladesh at work here? Shanta

Submitted by yayaver on

I was working recently in Ganjam district, Odisha, India (Phalin affected area). Government is investing in sanitation with the scheme - Individual Household Latrines (IHHL)to all Below Poverty Line (BPL) Households and Above Poverty Line Households (APL) restricted to SCs/STs, Small and Marginal farmers, Landless Labourers with homestead, Differently Abled and Women Headed households. In scheme, 90% of funds were given by state and remaining 10% must be contributed by beneficiary to show ownership. Yet the problem of open excretion is remaining in the public. Somewhere, only money spending is not enough. There is a need to raise awareness inside the community to change the behaviour of the people.

Submitted by Shanta on

Yayaver: Your example of Ganjam district illustrates the problem of neglecting the externality associated with open defecation. If the government gives the latrine to individual households, their incentive to use them is limited (compared to the benefits, which accrue to neighboring households too). Perhaps the government should consider giving the latrines to the community organization, and have them distribute it to the households. Then the community, which stands to benefit from everyone using latrines, will have an incentive to make sure everyone uses the latrines. Shanta

Submitted by Patricio Marquez on

Rigth on Shanta!!! Your blog makes a clear case for moving away from sectoral interests and silo projects to solve problems! At the end of the day, the pot of money in government hands is limited and trade offs are inescapable either explicitly or implicitly. As I argued before on a similar blog, a recipe for good health is safe water and basic sanitation, and not only clinics to treat the after effects of drinking polluted water or being exposed to open defecation. For those interested, see:

Submitted by Shanta on

Patricio, thanks for pointing me to your earlier blog post. There is a difference, however, between water and sanitation. The former is not an externality (if I don't drink clean water, the only person who suffers is myself), whereas the latter is--as I explained in my post. I think we often lump water and sanitation together, and assume government expenditure on both are justified. But the argument is really for sanitation. If we want poor people to get access to water, we should give them money, and have them pay the full cost of water. Unfortunately, the way we manage water now--by providing it for free or at reduced rates--leads to diversion, with poor people having to buy water from water vendors at 5-16 times the meter rate. Just as with health, where there are services with a rationale for public spending (immunization, etc.) and those without (curative care, mostly), there is a distinction in WASH. Regards, Shanta

Submitted by Mary Quelle on

The problem in many contexts is that sanitation and water are connected. Having sanitation downstream does not help water quality if there is none upstream. Then there are floods that spread the microbes even farther. Consider today's editorial in the NYTimes on the huge costs of lack of sanitation and resulting poor water quality

Some part of responsibility for clean water is properly for the government to assume in such a context, Shanta. Water utilities should recover costs, but they probably need to be set up by a public (community, city, district) authority to be efficient and effective.

Submitted by Shanta on

Mary: Good point. I agree that someone needs to set up the water systems. If water utilities recover costs, there is a greater chance that they will have an incentive to set up the systems. Put another way, the reason why the water infrastructure is in such poor shape in so many places is, partly, due to the fact that utilities haven't been recovering costs. Shanta

Submitted by Ajith Kumar on

Shanta, The role of the middle class, and hence the mainstream media, in addressing sanitation issues (which does not mean just allocating finances, but getting sustainable outcomes; good example being India) needs to be explored. Monies are allocated in national budgets, and interventions focused on, based on the level of interest and pressure that the middle class in a society places on that issue. Till now, so long as they stay in apartments with 'flush and forget' sanitation facilities, the middle class does not feel a need to stress on sustainable sanitation, especially in rural areas. In addition to the example in your blog, the 'big stink' which forced Parliament to clean up the Thames is another instance of the problem having to hit the right levels to find a lasting solution. Not only sanitation, one does not find solutions to malnutrition, quality education and even poverty, unless there is pressure on Governments to do so. And this pressure is possible only from the middle class and its media.

Submitted by Shanta on

Ajith: Thanks for that perceptive comment. I agree that we need the middle class to support the financing of public goods. But I think the problem goes beyond getting them to recognize the importance of these public goods. It is for government to stop providing private goods to the middle class for free or at subsidized rates. This includes water, electricity, education and curative health care. Most of the government budget is spent on these items, and the middle-class lobby is able to protect that, leaving little left over for the genuine public goods that could enormously benefit the poor. Shanta

Submitted by Dev Flor on

Glad to see that the post ends at community-level execution of sanitation efforts, which is where it sounds like it should be. Also, glad to see that the marriage quote is set within the context of a complete village initiative (some of the PR around the marriage and sanitation approach, at least on the surface, seems to be around individual households, which sounds completely ridiculous - particularly in a country like India.)

That said, is it necessary that government needs to fund sanitation initiatives, even if channeled through NGOs and the community? Are there mechanisms to mobilize community level funds for sanitation efforts (e.g. to purchase subsidized toilets & plumbing)?

Submitted by Shanta on

Good question, Dev. Insofar as sanitation is a local public good, the local community should finance it. But there are variations to this precept. For instance, the externality could spill over to the neighboring community. And the power relations within the community may mean that the community decides not to spend on sanitation and instead rebuild the village head's house. For these reasons, it is probably better if a higher-level government at least monitor and, if necessary, finance part of the sanitation program. Regards, Shanta

Submitted by Dev Flor on

Thanks so much for your response. One quick point of clarification in reference to a line from your post ("But giving it to the community enabled the community to monitor whether people were using the toilets."). I'm curious how the community monitors toilet usage.

Related to this, it seems like upkeep of toilets (whether within the home or for community toilets) is a big issue. The conditions of toilets in villages (particularly communal ones) are such that open defecation often seems like a better alternative, regardless of what the broader externalitites may be. I imagine factors like caste also play into this as cleaning and caring for the toilet would almost certainly not be done by men, and probably not by the women of the household either.

Could government be brought in to help fund the upkeep of toilets?

Submitted by Asif Dowla on

Could you please provide a link to the Bangladesh program? Is it run by BRAC?


Submitted by Shanta on

Asif: Here is the link to the Bangladesh program:
The NGO was the Village Education Resource Centre, a local affiliate of WaterAid. Regards, Shanta

Submitted by Joe on

I am not an economist (or any kind of social scientist), so I'm just throwing this out as an idea: is it possible that people are refusing to co-operate because they perceive that what they are offered is not 'as good as' the thing that they want? I read a study a while ago regarding dry toilets installed in a development in South Africa, and because residents perceived non-flush toilets as somehow inferior rebelled. In the end the developer had to install flush toilets.

I'm not saying this happens everywhere, but it strikes me that there is a potential shame attached to having an 'inferior' sanitation improvement to the extent that it is potentially/actually better to retain a condition of non-santiation (open defecation) 'defiance' on the basis that if one refuses the first bad intervention, eventually someone might come along in the future and offer you what you really want - and you might think (erroneously or otherwise) that accepting an inferior improvement might actually reduce the chances of getting the 'real' santiation improvement that you want, because the funders might think that having *some* kind of improved sanitation takes you out of contention for the next/future round of provision.

Submitted by Shanta on

Joe: Thanks for the question. While the "shame" of low-quality toilets may have something to do with the resistance to use it, that doesn't explain why people were more inclined to use the same toilets when it was provided by the community, rather than directly by the government. So it would seem that the way the program was administered was also a factor. Regards, Shanta

Submitted by Mary Quelle on

No lobby for prevention. You are so right when you say: "Rich people and medical unions have political power to lobby for spending on curative health care;" others lobby for no user fees for health care (another private good). " Poor people’s voice for sanitation goes unheard. "

Would you agree that global leadership could help counteract these lobbies? But the lobby for curative disease-care has already succeeded in making Universal Health Care into the one and only goal for government programs in health, a goal which the World Bank and WHO decided to set for post-2015, judging from the enthusiastic participants from global health organizations at the high-level conference on UHC at the World Bank Spring Meeting. Ministers of health of all countries have adopted it. The UN approved a resolution that UHC is a goal for all countries.

The signals and leadership from global policy-setters on what is the main gap in public services that drives bad health (correcting externalities) therefore differs from what you are recommending here. Or not? In any case, prevention and public health policy are not part of UHC programs, perhaps mostly because they have no lobby.

Is UHC actually conducive to getting the most health for the poor? If not, what post-2015 goal in health should the World Bank promote when there is such chronic low demand for sanitation and other public goods that boost poor communities' health status?

Submitted by Kofi on

On point, as usual, Shanta. This is the one area where Ghana is doing most poorly in. And from my NShoreNa Walkathon ( experience over the past two weeks, I have been seeing the very ugly side of this, as you can see from the facebook page. I have devoted a great deal of my time in the Bank communicating around sanitation, and have seen probably a dozen urban/rural/small town IDA sponsored environmental sanitation projects implemented, and probably a billion dollars of IDA lending to fix these messes, but Ghana today actually seems to be moving in the wrong direction on sanitation... A US$10 million treatment plant funded by the Bank has been abandoned near Tema for more than a decade because Tema Metropolitan Authority would not pay for electricity to operate the plant...

I am taking the liberty to share the article with our economic journalists here...

Submitted by MILAN K SINHA on

It's a common knowledge that defecation in open is fraught with high risk of microbial contamination of water which is a major cause of diarrhoea and other intestinal infections among the children. Health problems pertaining to a large section of population due to the aforesaid reasons has a multidimensional impact on our economy in terms of productivity losses, increased expenses in the name of providing medical treatment by the government. If one calculates the net financial loss to the exchequer for not having the basic sanitation facilities, it would be an astounding figure...

Submitted by Shabel on

Investment in sanitation without adequate awareness building and bottom-up demand creation is literally a waste of money. I agree CLTS programmes have proven to be the most effective but are as yet, even in Bangladesh, inadequately scaled up or institutionalised. Government and government bodies, local and national, are very poor in reaching and interacting with grassroot communities. To scale up, there is a need to ensure partnership between government (policy maker and fund provider), NGOs (social mobilisation and awareness raising) and communities (self-help initiatives supported with partial funds). It is this triangular partnership which can have the biggest pay-off and scale-up potential at the lowest cost.

Submitted by Geoff on

Excellent post, Shanta. Thanks for highlighting the issues of externalities and trade-offs between sectors. Once we agree that governments should subsidize public goods, then two important question emerge around efficiency and equity. As I'm sure you are aware, government financing for household sanitation hardware (even if delivered through an NGO to a community) is but one way to spend on sanitation. As has been pointed out in the comments, community power dynamics may even exacerbate inequities in such an approach. In Cambodia, subsidized toilets intended for the rural poor often end up in the homes of the better-off.

Regarding the efficiency question, I am surprised that your post did not mention the private sector - except for your assumption that they're unwilling to spend on sanitation. To the contrary, I would encourage you to investigate the emerging field of sanitation marketing, where government resources are invested in the enabling environment for sanitation businesses, rather than to procure hardware. This can be significantly more cost-efficient. When local governments exhibit leadership and support for total sanitation, when local supply chains can produce, market, and deliver appropriate and desirable products, when purchase financing is available, and when consumers are empowered to choose how they will adopt safe sanitation, there is a remarkable change in the sanitation landscape.

Having said all that, it's a shame that your post is still utterly relevant and necessary. Instead of needing to be convinced why to make investments in sanitation, it would be preferable if governments were instead focused on how to optimize such investments.

Also, thanks for the great story about William Henry Harrison!

Geoff Revell
WaterSHED, Cambodia

Separately, I'm also curious about your reference to Village Education Resource Centre. The link you gave is actually to the CLTS website, which is an approach that is used globally. But, a key tenet of the approach is that it eschews giving away toilets. Are you sure they were giving toilets to the community to allocate to households?

Add new comment