Research Highlights Podcast
April 9, 2026
Delivering clean water
Fiona Burlig discusses an experiment in rural India that measured the value poor households place on potable water.
Source: Srinivasan.Clicks
More than two billion people around the world do not have safe drinking water at home. Piped water infrastructure remains out of reach for much of the developing world, and cheaper alternatives like chlorine tablets have low take-up rates even when given away for free.
In a paper in the American Economic Review, authors Fiona Burlig, Amir Jina, and Anant Sudarshan explore a third option. Working with a private company in rural Odisha, one of India's poorest states, the researchers ran a randomized experiment across roughly 60,000 households to test the effectiveness of delivering treated water directly to people's doors.
Burlig recently spoke with Tyler Smith about revealed-preference measurements of the value of clean water and steps governments might take toward reaching the goal of universal access.
The edited highlights of that conversation are below, and the full interview can be heard using the podcast player.
Tyler Smith: The ultimate goal here is universal safe drinking water—we want everyone in the world to have access to potable water. How far are we from that goal?
Fiona Burlig: Unfortunately, we're quite far. Around two billion people around the world live without access to safe drinking water at home, which obviously creates a large public policy and public health problem.
Smith: Why has it been so hard to achieve universal clean drinking water in places like rural India?
Burlig: I think there are a couple of major challenges that prevent universal clean water access. The first is that putting pipes in everyone's home is actually a very intensive infrastructure investment. It's quite expensive. Building out piped water infrastructure—even though we've had it in the United States for 100 years or more—is just not feasible in developing countries. You're starting from a low base of capital and need to spend a lot of money to put pipes in the ground and get them out to people's houses. But, of course, just having pipes doesn't solve the problem entirely either. You need to make sure that the water flowing through those pipes is actually cleaned and treated. Here in the United States, we do that through a series of water treatment plants, but that's also quite expensive. And if you're a developing country government trying to balance a series of goals—vaccination programs, education, electrification, along with clean water access—you're making difficult choices with a limited budget.
Smith: Let's talk about the design of the experiment that you conducted. Can you walk us through it and tell us, in particular, what kinds of treatments you delivered to the participants in your study?
Burlig: I'll take a small step back, before we dive into exactly what we did, to set the stage for why we did it. In the context of this universal clean water access problem, we have billions of people around the world who don't have clean water access. There have been two primary solutions that people have put forward to try to solve that problem. The first is the one we might naturally think of—building pipes out to everyone's house. That's in the works in some places, but the reality is that the population is growing faster than we're able to build new pipes, so that alone isn't solving the problem. Another approach that economists have been fairly supportive of, and that we support as well, is point-of-use water treatment, which in many developing countries, is a chlorine dispenser next to a stream. So, when you go and fetch your bucket of water from the river, you can chlorinate it right then and there and bring clean water back home. The downside of point-of-use treatment tools is that they tend to be relatively unpopular with households. Even if you make chlorine tablets or other chlorine solutions completely free, high-quality studies suggest that take-up is only around 50 percent. If we want to close the clean water access gap, we need to try something new.
What we were trying in our experiment is a partnership with a company in India called Spring Health. They're one of a number of companies taking local groundwater, or sometimes river water, treating it in the local village with a small water treatment plant, and then delivering it door-to-door to households in our study. This gives households a convenient source of clean water that is fully treated and carries the same health benefits as the water we have access to in the United States but without the unfortunate taste of chlorinated water. It also means households don't need to take the action of cleaning the water themselves.
With that in mind, the experiment we ran was designed to understand how much households would be willing to pay for this type of clean water access. We partnered with this company, which was delivering clean water for a fee, and randomized the prices of that service. Some households received clean water completely free, and we observed how much they ordered. Others paid the market price, and we had a range of prices in between. That first part of the experiment is essentially a traditional demand curve—how much are people willing to pay for clean water? The second part is what we call an exchangeable entitlement treatment. Instead of saying, “Here's water at varying prices—how much are you willing to buy,” we told households, “We are going to give you 400 liters of clean water for free this month, but for every liter you choose not to order, we'll give you cash back.” The reason we were interested in doing this is that in addition to willingness-to-pay—which you get from a standard demand curve—we wanted to measure something called willingness-to-accept. This is a classic concept in environmental economics where instead of asking "how much do I have to pay to get a good," you flip it to "how much would I have to pay you to give it up?" That's a useful metric in countries like India where there is a constitutional right to clean water.
Smith: Let's talk about willingness-to-pay, the standard measure for economists. What did you find when you varied the prices you charged for this delivery service?
Burlig: We saw a few things that we think are fairly interesting. The first is that, perhaps unsurprisingly, demand slopes down. As you make the product more expensive, households buy less of it. But the way that happens in this setting is actually quite interesting. As we made the water more and more expensive, what we saw is that rather than buying fewer liters of water, households simply opted out entirely. So conditional on buying, they seem to be buying their whole month's supply. But as the price rises, fewer and fewer households make the decision to replace their drinking water at all.
Smith: Your estimates for willingness-to-pay are larger than some previous estimates. Why were the earlier numbers lower, and why do you think your estimate is closer to the true value?
Burlig: Coming up with a value for how much people are willing to pay for this kind of good is actually quite hard because it's not something that's often traded on the open market. If I want to understand how much you're willing to pay for tomatoes, I can go to the grocery store, watch prices go up and down, and see how many people are purchasing. With clean water, it's most often the case that people are just getting it from the local river or stream and not paying much for it. What's nice about our setting is that we are selling it as a product, which makes it straightforward to vary the price. That gives us one advantage—we've simply made the measurement exercise easier. The other thing previous studies have tried to do, where we think we've made an improvement, is that they measured people's valuation for clean water indirectly—by observing how much they're willing to pay for a water purification technology, like a filter or chlorine tablets, or by measuring how far they're willing to walk to reach a cleaner water source. Neither of those is a direct measure of how much you actually value clean water itself. What we're doing in this study is getting closer to a direct valuation, because we're measuring how much you're willing to pay for water that shows up at your home, at your convenience.
Universal piped water access is ultimately the gold standard . . . But we're operating in a budget-constrained world where that can't be made a reality for everyone tomorrow. Given that, we need to think about other ways to fill the access gap.
Fiona Burlig
Smith: Let's turn to willingness-to-accept. What did you find when you looked at the value of clean water through that lens?
Burlig: What we see is that people's willingness-to-accept is definitely higher than their willingness-to-pay—meaning I would have to pay you more, as the government, to take your clean water away than you would pay me to purchase it from me. And actually, the willingness-to-accept measure is only a lower bound on people's true valuation, because within the context of our experiment we couldn't give households enough money for them to say they'd take all the cash and give up the water entirely. That suggests that households actually value clean water quite highly—more so than you might conclude from the willingness-to-pay metric alone.
One reason that having both measures is interesting in our setting is this: the dollar value we get for water using willingness-to-pay turns out to be a little less than what it costs to provide clean water universally on the private market. That helps explain why we don't have a wave of companies solving the clean water access problem—it's more expensive for a firm to provide clean water than households are willing to pay directly. But if you instead use the willingness-to-accept metric, which is arguably more relevant from the government's perspective, you find that households would need to be paid substantially more than it actually costs to supply the water. That suggests providing clean water universally is potentially beneficial to society. And that's before even accounting for the positive health externalities—if I don't have Giardia, my neighbors are less likely to get Giardia, typhoid isn't spreading through the neighborhood, etc. Those benefits don't belong in my personal willingness to pay or accept, because they're externalities, but a government should be valuing them too.
Smith: I want to circle back to the ultimate goal here, which is universal access to clean water. Do you think this approach is the answer? How do we actually get clean water to people?
Burlig: This is a big enough problem that our view is that we need to start throwing new darts at the dartboard and seeing what works. We absolutely view this as an important contribution to the broader set of approaches that people are currently taking. I think there are reasons to believe that universal piped water access is ultimately the gold standard—being able to walk into your bathroom, turn on the tap, and drink straight from it is ideal. But we're operating in a budget-constrained world where that can't be made a reality for everyone tomorrow. Given that, we need to think about other ways to fill the access gap. Chlorine tablets are a great option if you're looking for the least-cost solution. Once you get to slightly wealthier populations or places where you really care about universal take-up—perhaps for public health externality reasons—then adding home delivery of clean water as a third option is probably a good idea as well.
♦
“The Value of Clean Water: Experimental Evidence from Rural India” appears in the March 2026 issue of the American Economic Review. Music in the audio is by Sound of Picture.