0:06

Economists often prefer to express the outcomes or consequences of water

Â and sanitation interventions in terms of

Â economic benefits measured in monetary units.

Â In this video, I'll illustrate this approach

Â for just one component of the total benefits.

Â This is the mortality risk to households.

Â 0:25

It's useful to think of two types of benefits

Â that households receive from a piped water and sanitation project.

Â Health benefits and non-health benefits.

Â The health benefits can be subdivided

Â into mortality reductions and morbidity reductions.

Â In addition, positive externalities may spill over and

Â benefit households who are not using the intervention.

Â The main non-health benefits to households are

Â increased income from having more, cleaner and cheaper

Â water, time savings from not having to collect

Â water, and aesthetic, or quality of life benefits.

Â A household can receive the health benefits from non-piped water

Â and sanitation systems, but only some of the non-health benefits.

Â For instance, if they still have to spend time carrying water back to the home.

Â 1:20

The economic benefits from mortality reduction will

Â be a function of four uncertain parameters.

Â The first is the baseline diarrhea incidents in the target population.

Â The second is the reduction in baseline

Â diarrhea incidence due to the WASH intervention.

Â The third is the case fatality rate.

Â This means the proportion of the cases of diarrhea that lead to death.

Â 1:44

The fourth is the value of a statistical life.

Â This is a measure based on how much people are willing to pay ex

Â ante, that is, before they get sick, for a specified reduction in mortality risk.

Â 2:30

This figure is from the Fisher paper.

Â It shows diarrhea episodes per child per year in selected low income countries.

Â The range is about 2 to 4 episodes per child per year.

Â The second parameter is the effectiveness of

Â the intervention at reducing the incidence of

Â WASH-related diseases in the population if a

Â specified number of individuals use the intervention.

Â You can think of the effectiveness of the

Â intervention as the percentage reduction in baseline incidence.

Â So, if an intervention is 90% effective, you have 10% of the number of cases left.

Â We'll assume, for purposes of illustration, that 100% of

Â the people in the target population use the intervention.

Â The instance after the intervention, labeled here Inc subscript after, is 1

Â minus Eff times the baseline incidence,

Â that is the instance before the intervention.

Â The change in the number of cases due to the intervention is the number

Â of cases in the target population before

Â minus the number of cases after the intervention.

Â To calculate the change in the number of cases, you multiply the number of

Â people in the target population times the

Â effectiveness of the intervention, times the baseline incidence.

Â 3:49

For this parameter we go back to the

Â Fewtrell paper that we discussed in the previous video.

Â Here's that figure again.

Â We said a good estimate for the reduction in risk from

Â baseline conditions due to a large intervention was around 30 to 40%.

Â In these calculations, I'll use 30% for the mean case.

Â 4:09

The third step is to calculate the change in

Â the number of deaths due to the WASH-related diseases.

Â You can think of this as the number of lives saved by the intervention.

Â This is done by multiplying the number of cases avoided by the case fatality rate.

Â 4:34

The diarrhea mortality rate is not the same as the case fatality rate.

Â The mortality rate is the number of deaths due to

Â a specific cause for a population during a specific time period.

Â The case fatality rate is the proportion

Â of deaths within a designated population of cases.

Â 5:13

The WHO global burden of disease report estimates

Â the number of deaths by different causes in population.

Â The mortality rate due to a specific cause of death, such as

Â diarrhea, is calculated from the estimates of the number of deaths in population.

Â 5:42

The estimates of the mortality rates in this figure show

Â that very few people in these countries are dying of diarrhea.

Â Even Bangladesh looks pretty good.

Â For Ethiopia, the mortality rate is estimated

Â at about 150 deaths per 100,000 people.

Â That's a 0.15% chance of dying from diarrhea per year in Ethiopia.

Â 6:19

The average diarrhea mortality rate for Africa stands out.

Â Is much larger than for other regions.

Â It is about 115 per 100,000, compared to South Asia which is about 65 per 100,000.

Â Now, let's look at the relationship between the diarrhea

Â mortality rate, the baseline incidence and the case fatality rate.

Â 7:05

The case fatality rate is then 150 deaths divided

Â by 150,000 episodes, or 1 death per 1000 episodes.

Â The fourth step is to multiply the number of deaths avoided by the value

Â of the mortality risk reduction as judged

Â by the members of the target population themselves.

Â The mortality reduction benefits, from an economist's perspective, are equal

Â to the value of the statistical life, multiplied by the

Â case fatality rate, times the size of the par, target

Â population, times the effectiveness of

Â the intervention, times the baseline incidence.

Â 7:44

I should emphasize though that the size of the target population is

Â the only parameter in this equation that is known with much certainty.

Â Our calculations depend on four uncertain parameters,

Â baseline diarrhea incidence, reduction in baseline diarrhea incidence

Â due to the WASH intervention, case fatality

Â rate and the value of a statistical life.

Â Now I want to show what happens to the final

Â result as we make different assumptions about these four parameters.

Â 8:14

This table shows you the assumptions I'll make for the parameters.

Â For each parameter I'll make a low, mean and high assumption.

Â In the baseline incidence column you'll see, I'm assuming

Â 0.5, 1, and 1.5 cases of diarrhea per year.

Â Note that these estimates are an average of both adults and children and are a

Â little different that then two to four episodes

Â per year for children that we saw earlier.

Â 9:12

In this table, we calculate the numerous cases

Â of diarrhea avoided as a result of the intervention.

Â For the mean case, that's the 30% reduction.

Â The second row here, in this table, the number of cases of diarrhea avoided is

Â 30,000 but this varies from 5,000 for the low case to 75,000 for the high case.

Â 9:43

For the mean case, the first row here, the result is 24 deaths per year.

Â This varies from 2 deaths per year for the low

Â case to 90 deaths per year for the high case.

Â Notice how the range of uncertainty is expanding.

Â 2 deaths per year from a population of a

Â 100,00 is very different from 90 deaths per year.

Â 10:06

This table shows step four, the calculation of the economic

Â value of the mortality risk reduction due to the intervention.

Â The total economic value of the mortality risk

Â reduction varies from US $20,000 to US $9 million.

Â That's a huge range.

Â 10:29

The high estimate is equivalent of US $38 per household per month.

Â This is a huge benefit for a poor household in a low income country.

Â But the low estimate of US $0.08 per household per

Â month is not a large benefit, even for a poor household.

Â 10:47

If we are trying to understand household behavior, and how

Â a household thinks about the economic value of an intervention.

Â The difference between these three cases is very large.

Â To wrap up, I want you to think about what

Â these calculations mean for understanding baseline or status quo conditions.

Â One lesson is that the economic outcomes of a water and sanitation intervention

Â will depend on the timing and sequencing of investments, and on local conditions.

Â And inevitably will be subject to a high level of uncertainty.

Â This is because the uncertainty in the estimates increases

Â as you multiply several uncertain perimeters in a row.

Â In other words, this means that uncertainty is not

Â additive, it's multiplicative, and the range of uncertainty expands rapidly.

Â 11:34

This finding resonates with Nobel

Â Laureate Douglass North's observation that, quoting,

Â we should be very tentative about how we understand the world.

Â That doesn't mean you don't do things.

Â You've got to do things, but you've got to recognize that you may be wrong.

Â We will return to the implications of this insight in our second follow up MOOC.

Â 11:54

To remind you, that's where we'll focus on specific policy interventions in

Â the water and sanitation sector, the

Â questions, what works, what should be done.

Â