On Cookstoves, there’s no need for hyperbole

Improved Cookstoves are all the rage. Drawing the kind of A-List backers – from Hillary Clinton and Mary Robinson to Michelle Bachelet and Julia Roberts – that used to gravitate towards microcredit and water wells, cleaner cookstoves are indubitably flavour of the month. And with the dire health and environmental impacts of breathing all that smoke from indoor fires, it’s no wonder the development community is going all in.

Some researchers call the low take-up rates for improved cookstoves “puzzling“.  To me, they’re anything but.

There’s just one problem: a lot of recipients don’t use them. In study after study, demand is weak and adoption rates are dismal. At times even when they’re given away for free, families use them only sometimes, if at all. Obviously, the Third World is a big place, so Clean Cookstoves do better in some places than others. In some studies, improved cookstoves lead to positive health outcomes, but in other settings health impacts prove elusive, with studies often throwing up equivocal, contradictory results.

The long and the short of it is that clean cookstoves seem to be considerably more popular in Washington and Geneva than in Senegal or Orissa.

Some researchers call the low take-up rates “puzzling“.  To me, they’re anything but. Clean cookstoves are perceived as the solution to a problem you may have a looooong time in the future by people whose time horizons are brutally compressed by the daily grind of extreme poverty. Is it really surprising that their value is discounted down, in many cases, to zero?

In some of the wonkier corners of development academia, people are nodding thinking “ah yes, hyperbolic discounting.”

But that’s not what I mean. You don’t need to reach for some exotic discounting function to understand beneficiaries’ indifference for clean cookstoves. There’s, quite literally, no need for hyperbole.

Bear with me a second, see if this makes any sense to you: in the U.S., various government agencies calculate the “value of a statistical life” (VSL) at around $7 million. (That’s about how much a U.S. regulatory agency is allowed to spend in order to save one life.) A U.S. VSL is worth about 177 years of earnings for the median worker.

For someone living on $1.25 a day, 177 years worth of earnings is about $80,800 – we’ll call that the VSL at the global poverty line.

Now, say you’re living on $1.25 a day, and a development worker knocks on your door selling you a clean cookstove. He informs you that without it, you’ll die a gruesome death from respiratory illness 25 years from now. How much should you rationally be willing to pay for that stove today?

The answer to that depends on your discount rate, of course.

We can get some sense of what that might be by looking at the interest rates moneylenders charge in very poor countries. Some studies find that in rural Bangladesh, 103% yearly interest is not that uncommon a rate – to say nothing of the outright loan sharks who make a living in every slum in the planet charging much higher rates. But let’s stick to the non-criminal side of traditional moneylending: what’s the value today of $80,800 twenty-five years from now, discounted at 103%?

If I’m doing the math right, the answer is one fifth of one US cent. That’s how much you should rationally be willing to pay for that stove.

Even at less aggressive discount rates of, say, 60% a year, the current value of that stove to someone on $1.25 a day is still just 65 cents. But improved stoves usually cost at least $15, with more robust models in the neighbourhood of $100. Clearly, we’re not within striking range.

Yes, you could come back and say that the health impacts of indoor smoke inhalation are chronic, rather than all-at-once after 25 years, but then I could come back and tell you that there’s a ton of data implying that in the real world people typically do discount hyperbolically (or, for the pedants, quasi-hyperbolically).

However you want to characterize the discounting function, evidence of irrational present-biased discounting is rife, not just among the Bottom Billion – who face powerful reasons to concentrate on just getting through the next day – but for nicely cosseted NYU students and Netflix customers too.

The discounting functions the poorest people use might not be particularly reasonable, but then, “life under an unreasonably discount function” is as good a working definition of poverty as you’re likely to find. Out of all the things that the very poor just can’t afford, rationally valuing their future is possibly the cruellest.

Which is all by way of coming back to my mantra: if you want to do something about indoor pollution in very poor countries, focus on incomes.

Because, as incomes rise, two things happen in tandem: 1-a poor person’s implicit VSL rises, and 2-his discount rate falls. At $3 a day, 177 years of income is worth $194,050. (Exponentially) discounted at a still high but no-longer-astronomical 40%, that makes the present value of a cookstove that will prevent a gruesome death in 25 years’ time $43. As incomes rise, the math starts to work. It’s not surprising that a number of studies find the single best predictor of adoption is income (sometimes edged out by education).

And it stands to reason. Look in your kitchen. If there’s a smoke-free stove there, I bet it’s not because some NGO gave it to you. It’s because you have an income that made buying it make sense for you. And when clean cookstoves come to the bottom billion, they’ll come in very much the same way: through businesses catering to consumers with the disposable income it takes to demand a clean stove.

To be fair, the Alliance for Clean Cookstoves makes all the right noises about enabling market solutions to the problem. A lot of very bright people are working on the problem, and thank God for that. But as long as you take dollar-a-day incomes as a given you have to innovate around, you’re going to keep running up against some very hard constraints.

Hyperbole aside, there are no shortcuts to increased incomes when it comes to enabling the spread of clean cookstoves.

17 thoughts on “On Cookstoves, there’s no need for hyperbole”

  1. It often comes down to what people ask for and proper incentives for change (and not only what an external donor wants ot considers to be an incentive)

    Just popping in to say I’ve been enjoying your blog. You provide a voice that is interesting and productive, no matter if one (fully) agrees with it or not. Keep it up.

  2. Quico, your calculations of the VSL for someone who lives on $1.25/day is completely wrong. You can’t simply multiply $1.25/day by (365 days x 177 years) because the number of years (177) comes from a VSL estimate in a much richer country. I understand that estimating the VSL anywhere or for any group of people is a very complicated matter but your back-of-the-envelope estimate of the VSL at the global poverty line is a very bad one. On top of this, your are using unreasonably high discount rates!!! Those very high interest rates you observed in very poor countries are not really explained by a very low discount factor (i,e., low preference for tomorrow’s consumption). The high interest rates are probably better explained by high risk premiums, credit market imperfections, and (possibly) high inflation rates. So, I’m afraid that all the conclusions based on your sloppy calculations are not to be taken very seriously.
    It is true that the VSL for people at extreme poverty must be lower than for a typical worker in rich countries. As a proof of this we have the very same puzzle you’re trying to explain: very poor people are not willing to adopt this better home technology that supposedly is going to make their life longer! That’s exactly what the VSL tries to estimate. The VSL estimates the willingness of people to trade off wealth for a reduction in the probability of death. So, why could possibly explain this? Obviously, the cost of adopting this technology is too high. But again, why? One reason is their very low income as you mentioned. Another reason, which is my main candidate, is somewhat related to how they discount the future. The discount rate depends not only on how much individuals value tomorrow’s utility relative to today’s but also on how likely is that they are going to be alive tomorrow. If mortality rates are very high and have very little to do with indoor smoke inhalation, then it’s clear that this improved stove is going to have a very marginal effect on their survival probability. Therefore, they are not going to be willing to pay much for it. As a conclusion, you could also say: focus on public health!

    1. Two things…

      1-Of course 177 years is an arbitrary and high number. If I use a lower number of years, more suitable to a very poor country, then the VSL (and the discounted present value of the stove) is even lower!

      2-Banerjee and Duflo let observed exchange rates charged by non-loan shark money-lenders stand in for discount rates, and if they get away with it, I sure do too!

    2. (P.S.: you know what observed interest rate is too high to let it stand in for a discount rate? The 5% PER DAY that wholesalers in Hyderabad charge buhoneros! But that’s mafia stuff…)

  3. I’m not convinced by this, either. At least, I don’t see that the math adds much.

    Poor people may discount the health benefits which will accrue in twenty-five years for the same reasons that young people discount the long-term effects of smoking. Sure, there are pressing short-term things to worry about, and dying at age seventy of lung cancer seems remote.

    While your “incomes mantra” no doubt plays a role, I would think better education would be just as important. There are lots of public health information initiatives which can affect this kind of decision.

  4. Just to pile on, the discount rate doesn’t apply if the cookstoves are given away free. As you point out, even in that situation people are not using them.

    1. OK, OK, the post has problems!

      Still, my point is that if your discount rate is high enough and you value something at, basically, zero, YOU WON’T USE IT EVEN IF IT’S FREE…

      Defensive? Me? Sigh…

    2. The price might be zero but there could still be other costs associated with adopting this new technology.

      Regarding the discount rate, as I mentioned earlier, this could be high simply because high mortality rates.

  5. A fellow international do-gooder once criticized my criticism of cookstoves as an “a priori xenophobic dismissal of the intentions and products of rich-world technological intelligence.” More here:

    Rather, my concerns are based on wanting to ensure that any efforts to improve people’s lives in the developing world are first based on the locally available resources, before creating additional dependency on outside “expertise,” supplies, or technology. My concerns also include wanting to avoid undermining local economies and local organizations, especially if products such as these are delivered through traditional aid funding mechanisms, with each layer of bureaucracy taking its share.

    Despite whatever trend comes next from the policymakers, development experts, and donors, skilled and experienced development practitioners working on the ground know that no technological initiative in and of itself can offer the full answer to complex problem of poverty in the developing world.

    1. A fellow international do-gooder once criticized my criticism of cookstoves as an “a priori xenophobic dismissal of the intentions and products of rich-world technological intelligence.”

      Man! I bet that looks great on your CV!

      Is it too cynical to see this whole thing as a bit of personal-history-buffing for high powered woman politicians? I really hope so. It’s not like Mary Robinson needs any more humanitarian street cred, does she? What is she, trying to corner the market?

  6. Great post. However I’m not sure the math adds much here either.

    I’m always a bit sceptical of well intentioned formulas and analysis arrived at in the economics labs of universities that attempt to explain human behaviour (see the articles on WEIRD research here too). Are we really meant to believe that people scratching out a living from a small plot of land are doing the complex math behind such grand ideas as game theory, adaptive and rational expectations, VSL and discount rates? I doubt it. This is not how people in the real world make decisions. Including me.

    That said, despite your wrong turn at the math intersection, you’ve still arrived at the right destination. Getting people’s incomes increased is the main show in town. Hanging on the coat-tails of that you’ll find all the other stuff we are looking for: improved health, education, opportunities, democracy, etc etc.

    Besides, if Western, Educated, Industrialized, Rich, and Democratic (WEIRD) liberals such as ourselves can’t figure out the right way to apply the math that we invented, how on earth will anyone else?

    Great blogs! Please keep them coming!


    1. Well, the idea is that the way you behave reveals an implicit calculus, not that you’re literally sitting there with a scientific calculator working out how much to pay for the thing.

    2. Agreed. At least that’s the theory. But isn’t taking out a scientific calculator and working it out exactly what you ARE DOING in the post? :) Like I said, a great post, with or without the math.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s