This is cross-posted from my newsletter The Fitzwilliam, which is at the intersection of policy (particularly Irish and European) and effective altruism.
On a recent episode of the 80,000 Hours podcast, Rob Wiblin and Karen Levy, a development expert, discussed the popularity of global health interventions that charge a small fee for use:
Rob Wiblin: Given your experience, what are two or three aid-related initiatives in Africa funded by Western governments or others that you believe should be significantly curtailed or changed?
Karen Levy: The first one that comes to mind are programs that require user fees. There’s been a lot of work on whether or not paying for something actually increases adoption. And there was a wonderful paper that Rachel Glennerster and Mary Ann Bates and others did called “The Price is Wrong,” which looks at what happens when you put in place even small user fees. The logic behind that always was that if people pay for something, they’ll value it more. And we’ve seen in practice is that, first of all, that’s not true: people do value and use things even when they are given them for free. But also that even small user fees can exclude the poorest and most marginalized — the people that you want to reach the most. And that the cost of levying those fees and collecting them can actually outweigh the value of those fees. So I think programs that have that — whether it’s in the name of participatoriness, or ownership, or cost sharing, or sustainability, or whatever the justification may be — I think that is very, very often misguided. And I would like to see people getting more free stuff. I think that would be a good thing. So that’s one.
Rob Wiblin: Yeah. Is that sort of pay-for-service or pay-something-for-service quite common?
Karen Levy: Yes. I think it is quite common.
Rob Wiblin: Interesting.
Karen Levy: I think sometimes it’s in the very same way that there are people who think we shouldn’t give poor people money because they’ll use it on cigarettes or alcohol. There is this sense that somehow people won’t value something unless they pay for it themselves.
Rob Wiblin: Yeah.
Karen Levy: I also think it comes from pressure from donors who want to see cost sharing . . .
Rob Wiblin: The story I’d heard on that one was that that was a common theory — that it was better to charge people something so that they’d value it and be more likely to use it — but that that had just been tested quite repeatedly, and it had been demonstrated that that was outweighed by other factors. That it was much more important to get as high uptake as you could, and even small fees had a huge influence on uptake. But it sounds like that hasn’t flowed through to all of the different programs where it’s potentially a relevant lesson.
Karen Levy: I think that many people have gotten that message. The bed nets programs have really taken that — I think that it’s had an enormous impact on how mosquito net distribution programs operate. But there’s a whole world out there of programs that are not influenced by these things, and so that is one that I still unfortunately see a little bit too often.
Rob Wiblin: It’s interesting. I’m just thinking about this in the context of the UK’s vaccination rollout, or the US vaccination rollout. Each dose, depending on the vaccine, costs between $5 and $20, something like that. But I think people knew intuitively, as well as from the research, that even if you charge people $2 for a vaccine dose, you would get a big drop in participation relative to it being free. And given the huge social value of people having vaccinations — hundreds, thousands of dollars even — that would be insane. I guess kind of the same principle applies here in other areas: if it’s really valuable for someone to use something, then charging them some nominal fee for it and having half of your population drop out is a really bad move.
Karen Levy: That’s right. And certainly in the case of vaccinations, not only should we not be charging people, but we should be incentivizing them.
Rob Wiblin: Paying them. Yeah.
Karen Levy: Right? We’re paying them to actually get vaccinated. But let’s take chlorine as an example. So how do you get people to chlorinate their water if they don’t have access to regularly chlorinated water? The theory was, if people buy chlorine, then they’re going to be more likely to use it. Well, it turns out that, first of all, it’s really hard to get people to spend money on essentially preventative health types of goods. But also, even if you give them chlorine for free, if it’s sitting in a bottle in their house, it sort of goes up on the shelf and you may not even remember to use it. Whereas the Dispensers for Safe Water program run by Evidence Action places chlorine dispensers at the point of collection. This makes the use of chlorine a social activity. So you see other people doing it, and that reminds you to do it. It’s part of the water collection process. That is what then generates uptake. Even the deworming program, the researchers on that study tried introducing very, very small user fees for the deworming tablets, and the demand dried up.
Wiblin got mixed up here: the fees are intended to make people more likely to use something after they’ve bought it, but you only need to be vaccinated once. I suppose charging for vaccines could make people more likely to show up to the appointments for second and third doses, but the problem is overwhelmingly in getting people to take the first dose.
I understand Levy to be making six claims:
After people have paid for something, they are not more likely to use it.
When people are forced to pay for something – even a tiny amount – uptake is much lower than if it were free.
Even if people were more likely to use things they’ve paid for, this effect is dominated by the lower uptake, and so you should not charge for highly cost-effective health interventions.
Even though charging a fee saves money, it excludes very poor people, and the administrative fees associated with charging it often exceeds the fee itself.
People do not, in general, value free things less, and in fact, there are many social benefits to administering services for free.
Fee-for-service interventions are condescending toward poor people because they undermine their agency and rationality.
I am sceptical of claim (1). Feeling obligated to use something because you’ve already spent money on it is a sunk cost effect, and the evidence that people do this is strong. However, this isn’t important, because claim (3) is probably true.
Claim (5) is more interesting. People certainly seem to value free public education and healthcare highly (“The NHS is the closest thing the English have to a religion”). Many families that send their children to public school could afford to pay tuition, if they had to. But presumably, there are social benefits to living in a society where education is free, like decreasing social stratification.
But what about the bad incentives created by free programs? Wouldn’t it be better if you had to pay a £10 fee to book an appointment with the NHS?
Almost certainly yes. But this corroborates Levy’s account. Even tiny fees significantly dissuade people from doing things, and, at the current margin, many people are going to the doctor too much. Hence doctors’ fees are good. No complicated theory about how people don’t “value” free healthcare needed.
I don’t know if claim (6) is an accurate summary of Levy’s position. But, taking the claim at face value, I don’t understand it. The idea here is that we wouldhave given out some health intervention for free, but we decided to charge for it instead, because we don’t trust poor people to make sensible decisions, and this is condescending. But if the research refutes claim (1), then this distrust is warranted. Poor people really would be making terrible decisions with their health. Maybe claim (1) is correct, but would we have been condescending if the research had come out the other way? What would happen in the least convenient possible world?
My conclusion is that development is hindered when we form beliefs based on what is or is not condescending to believe. It’s either true that user fees make interventions less cost-effective, or it isn’t. It’s either true that giving poor people money causes them to spend it on cigarettes and alcohol, or it isn’t. The less we interject our moral views about poverty, the better.
Making People Pay for Something Doesn’t Cause Them to Value it More
This is cross-posted from my newsletter The Fitzwilliam, which is at the intersection of policy (particularly Irish and European) and effective altruism.
On a recent episode of the 80,000 Hours podcast, Rob Wiblin and Karen Levy, a development expert, discussed the popularity of global health interventions that charge a small fee for use:
Wiblin got mixed up here: the fees are intended to make people more likely to use something after they’ve bought it, but you only need to be vaccinated once. I suppose charging for vaccines could make people more likely to show up to the appointments for second and third doses, but the problem is overwhelmingly in getting people to take the first dose.
I understand Levy to be making six claims:
After people have paid for something, they are not more likely to use it.
When people are forced to pay for something – even a tiny amount – uptake is much lower than if it were free.
Even if people were more likely to use things they’ve paid for, this effect is dominated by the lower uptake, and so you should not charge for highly cost-effective health interventions.
Even though charging a fee saves money, it excludes very poor people, and the administrative fees associated with charging it often exceeds the fee itself.
People do not, in general, value free things less, and in fact, there are many social benefits to administering services for free.
Fee-for-service interventions are condescending toward poor people because they undermine their agency and rationality.
I am sceptical of claim (1). Feeling obligated to use something because you’ve already spent money on it is a sunk cost effect, and the evidence that people do this is strong. However, this isn’t important, because claim (3) is probably true.
Claim (5) is more interesting. People certainly seem to value free public education and healthcare highly (“The NHS is the closest thing the English have to a religion”). Many families that send their children to public school could afford to pay tuition, if they had to. But presumably, there are social benefits to living in a society where education is free, like decreasing social stratification.
But what about the bad incentives created by free programs? Wouldn’t it be better if you had to pay a £10 fee to book an appointment with the NHS?
Almost certainly yes. But this corroborates Levy’s account. Even tiny fees significantly dissuade people from doing things, and, at the current margin, many people are going to the doctor too much. Hence doctors’ fees are good. No complicated theory about how people don’t “value” free healthcare needed.
I don’t know if claim (6) is an accurate summary of Levy’s position. But, taking the claim at face value, I don’t understand it. The idea here is that we would have given out some health intervention for free, but we decided to charge for it instead, because we don’t trust poor people to make sensible decisions, and this is condescending. But if the research refutes claim (1), then this distrust is warranted. Poor people really would be making terrible decisions with their health. Maybe claim (1) is correct, but would we have been condescending if the research had come out the other way? What would happen in the least convenient possible world?
My conclusion is that development is hindered when we form beliefs based on what is or is not condescending to believe. It’s either true that user fees make interventions less cost-effective, or it isn’t. It’s either true that giving poor people money causes them to spend it on cigarettes and alcohol, or it isn’t. The less we interject our moral views about poverty, the better.