[BLANK_AUDIO] In this video, we will shift our focus from household demand for water quantity, to household demand for water quality. We are going to start by focusing on the results from a paper entitled, Trickle down: Diffusion of drink water treatment in Kenya, by Professor Michael Kremer of Harvard University. Note that we will again be looking at findings from Kenya. The research design used in this study involved two different treatments to improve the quality of water used by rural households. The first was spring protection. This involves physical improvements to springs to enable households to collect water without contaminating the water source. The second treatment was a point of use treatment called WaterGuard. WaterGuard uses chlorine to disinfect water collected from sources outside the home, such as springs. WaterGuard was first given to treatment households for six months for free. Then, at the end of the six months, treatment households were required to pay a subsidized price for WaterGuard if they wanted to continue to receive the product. For the WaterGuard portion of the research, Professor Kremer and his co-authors tried to explain whether or not a household used WaterGuard as a function of the price of WaterGuard, health knowledge, social networks, and presence of children in the household. Professor Kremer's research attempted to answer six questions. First, do households understand that dirty, contaminated water is linked to poor health? Second, will households use WaterGuard if it is provided for free? Third, did WaterGuard improve the quality of water in the household? Fourth, dId the use of WaterGuard in treatment households reduce diarrhea? Fifth, do households value clean water? Sixth is behavior, here the adoption of WaterGuard, affected by social networks? Next I'm going to discuss his findings for each of these six questions. So, question one, do households understand that dirty, contaminated water is linked to poor health? Mostly, yes. Households do understand the link between contaminated water and poor health. Households do not like to drink water they perceive to be dirty. Question two, will households use WaterGuard if it is provided for free? Mostly, yes. 79% of households reported to researchers that they used WaterGuard in follow-up visits. But this means that about one out of five households did not want the product even if it was free, but note that this is self reported use. The researchers tested the water quality in households and confirmed that chlorine was present in the water of 58% of households. So, just a little over half of households appeared to be using WaterGuard, even when it was free. Question three, did WaterGuard improve the quality of water in the household? Yes. The research team tested the quality of drinking water in the home and found, on average, 69% less E Coli contamination. They found that the use of WaterGuard had a greater reduction in E Coli contamination than spring protection. Spring protection only gave about 23% reduction. Interestingly, the combination of spring water protection and WaterGuard, did not result in a greater improvement in water quality than WaterGuard alone. Question four, did the use of WaterGuard in treatment households reduce diarrhea? Yes. The baseline prevalence of diarrhea was about 20%. After the treatment, the prevalence was reduced to 12%. This 35 to 40% reduction in diarrhea is typical of other findings in the literature, which we'll discuss in week five, in more detail. But note that this calculation is based on self-reported diarrhea cases. Question five, do households value clean water? Here's where things get really interesting. The answer is no. Unlike household demand for water quantity from piped connections, household demand for improved quality was, in this case, extremely price elastic. Even a small charge, a small price, deters almost all households from purchasing WaterGuard for themselves. This is even after they've used it for six months, and it has reduced diarrhea rates by 35 to 40%. Professor Kremer concludes, quoting, we find no evidence that valuation of the product is higher among households who stand to benefit most from it. Question 6, is behavior, or the adoption of WaterGuard here, affected by social networks? The hypothesis here is that people may be influenced by their neighbors. In other words, if they see others nearby using WaterGuard, they're more likely to use it themselves. The answer is not much. The researchers found that network effects were small in comparison to the increase in uptake when WaterGuard was provided free of charge. Is this an isolated finding? Let's look at another recent paper. The effect of water quality testing on household behavior: evidence from an experiment in rural India, by Amar Hamoudi, Marc Jeuland, and their collaborators. In this study, the sample households were in communities in rural Andhra Pradesh, India. Most had access to potable water from commercially available supplies, but few people, less than 10%, chose to use this safe water. The research team tested the quality of the water supplies in the homes of treatment households. They also provided them with information about the bacteriological quality of the drinking water in their home. But no information was provided to households in the control group. The research team found that households who were told that their water was contaminated, increased their use of safe water slightly. But the majority of treatment households did not switch to safe commercial sources. The bottom line is that the researchers found low household demand for source water quality improvement, just like we saw for Professor Kremer and his team in rural Kenya. You can learn more about this study in Dale's conversation with Professor Jeuland. This is one of the videos available for you to watch this week. Finally, let's look at one more study. This one is by Junaid Ahmad, Bish Goldar, and Sumita Misra. This research team used a stated preference method, contingent valuation, to estimate households' willingness to pay for arsenic-free water in rural Bangladesh. They conducted interviews with 2,700 households. That's a huge sample. They estimated that the average household was willing to pay only ten to twenty US cents per month, that's about 0.2 to 0.3% of average household income, for arsenic-free drinking water. The research team concluded, quoting, the estimates indicate that rural people in arsenic-affected areas of Bangladesh place a low value on arsenic-free drinking water. Costly arsenic reduction technologies may find little social acceptance unless heavily subsidised. Again, we have the same conclusion, very low household demand for water quality improvements. Our review of household demand behavior has left us with a puzzle. Empirical research suggests a variety of low-cost water and sanitation interventions, such as water quality improvements, are highly cost effective and improve health. But household demand for them is low, or in other words, very price elastic. On the other hand piped water and sewer networks are very capital intensive and expensive, and their health benefits are often site specific and somewhat ambiguous. But household demand for piped network services is high, or price inelastic. People can often afford non-piped WASH services, but in many cases they often do not appear to want them. People really want the convenience, time-savings, and quality of life benefits with piped services. But they often cannot afford the full costs of piped network services, but they do want them. Households then demand that governments provide them with piped network services that they cannot afford. This finding that household demand for the health benefits from water quality improvements is low, is not limited to WASH services. Researchers are reaching the conclusion that household demand is low for a variety of preventive health interventions, such as impregnated mosquito nets and vaccines. This photograph shows children fishing with mosquito nets that have been donated to them to prevent malaria. In the next video, we'll look at the effects of social norms on household water use. Thanks for watching this video. [BLANK_AUDIO]
