25 Mar 2010

Water and human rights -- part 4

This post is one of ten in the serialization of my paper on human rights, which is introduced here.

Do human rights deliver results?

Many people support a human right to water, especially with the number of people dying from dirty water and the far greater number suffering from water shortages. But do good intentions lead to good results?

Some data

If a constitutional right to water is going to be meaningful, then it should have an impact. One way to test for such an impact is to compare statistics for “access to an improved water supply" before and after that right is added to the country’s constitution. We begin with countries that guarantee a de jure human right to water in Table 1. In the twelve countries for which we have full data, we see that the share of the population with access increased from 74 percent before the right was enacted (the base year, which varies by country) to 81 percent in 2006 (the most recent year for data). Can we credit this improvement to a human right, or is some other factor at play? Although the dynamics of development are extremely complex, we can get a ballpark answer by matching these human rights countries to other, non-rights countries at a similar level of development in the base year, to see how access changed for each group.

This comparison shows that access in countries without rights increased from 77 percent in the base year to 82 percent in 2006; in countries with rights, access increased from 74 to 81 percent. Although countries that added rights improved access by 2 percent more, this difference is not necessarily due to the existence of rights, which does not have a statistically significant impact in a multi-variate regression. Using a different method, we see that the correlation between access in the base year (when rights existed in none of these countries) and access in 2006 is 94 percent in countries that added rights and 97 percent in matched countries that did not. Put differently, access in 2006 is nearly perfectly correlated with access in an earlier year, not with a change in law that occurred along the way.

Some analysis

Perhaps the main reason that a right to water does not produce access to water (let alone access to clean water) is that such a right is costly to provide. Another reason — perhaps more important — is that water supply requires a functioning government. Most countries in Table 1 have low Human Development Index (HDI) rankings (meaning that their citizens are poor, unhealthy and uneducated) as well as low quality government. While it is clear that clean water can drive HDI, and HDI can improve governance, the opposite causal flow is probably more important: good governance is a necessary (and perhaps sufficient) condition to increase both HDI and clean water. Transparency International (2008) devotes a large portion of its annual report on governance and corruption to the water sector, giving many examples of provision failure and the disproportionate impact of that failure on the poor.

Thus, it is easily possible that politicians make constitutional promises that bureaucrats cannot deliver, because they don’t deliver much, anywhere, anytime. This failure may persist because these bureaucrats and their corrupt bosses can blame outside factors such as the weather, irresponsible customers, lazy engineers, the gods, and so on, diverting attention from their own (in)actions. Also note which countries are missing from Table 1: countries with good HDI and governance values. They do not need to create constitutional rights to something their citizens already have.

So how is it possible to ensure that people get adequate, clean water if not through constitutional rights? Let’s try a different approach: Instead of giving people a human right to receive water from some government organization, give them a property right in water. Give each citizen his share of the nation’s water wealth. Ownership not only makes it more likely that they will get water, it gives them an asset that they can sell in markets. As a useful side-effect, the markets will increase transparency in and improve efficiency in water allocation.

In tomorrow's episode, I argue that people should have a property right in water, and cover the tricky question of personal versus communal ownership of that water


  1. David,

    Thanks for this paper and for your blog. Regarding this section, I'd love to see the right to water compared to the success of a government in providing other positive rights. I'm pretty skeptical of a right to water, but that's largely informed by observing endless subdivisions being built before securing a supply. The situation in, e.g., Bolivia is clearly different.

    Your proposal to provide a base amount of water for consumption and then charge higher rates for excess use is somewhat rooted in the common law of water in the Eastern US, where the amount of water that was deemed "necessary" was not subject to division, even in periods of drought where the appropriation of that entire portion would take all the available water.

    (When I say "rooted," I'm really referring only to the understanding of water supply as consisting of two distinct portions: necessary and luxurious.)

  2. I just want to note that some countries don't have written constitution.

    Secondly, for countries with with written constitution, they do contain nirvana clauses, because that's what's constitution supposed to be: a foundation of ideals.

    I am concerned that the methodology is flawed because it tries to measure a norm's effectiveness through a comparative statistics.

    Consider the issue of slavery and racism. Slavery was abolished in the US in 1865 (through the third amendment). How long does it take, since the third amendment, for a black citizen to be able to vote?

  3. @Eric -- thanks

    @Mova -- a "nirvana" constitution is not worth the paper it's written on. A constitution* is meant to provide RULES that propel or constrain the government, and they should be achievable. Slavery was ended under the 13th amendment; voting was covered elsewhere -- see also votes for women.

    * Wikipedia: "A constitution is a set of rules for government—often codified as a written document—that enumerates and limits the powers and functions of a political entity."


Read this first!

Make sure you copy your comment before submitting because sometimes the system will malfunction and you will lose your comment.

Spam will be deleted.

Comments on older posts must be approved (do not submit twice).

If you're having problems posting, email your comment to me