If user fees for health have been so vilified (including in comments on this blog), why are we bringing the subject up again? Because new evidence calls into question the prevailing view, namely that removing user fees leads to: (i) increased use of health services and hence to (ii) improved health outcomes. Confirming (i), the recent literature shows that (ii) does not always follow.
Raising the price of a good or service has two effects: it reduces demand and increases supply. In the case of user fees for health, it was thought that paying for a service also makes people use it more appropriately (you don’t go to the doctor for minor ailments) and value it more than if they obtained it for free. The new empirical evidence shows that this reasoning is incorrect [people use bed-nets just as much, even if they get it for free]. On the supply side, though, the evidence is consistent with the view that payments to providers generate incentives for better performance.
Unfortunately, in many African countries, user fees were not paid directly to health facilities or frontline providers. They went into the central-government’s coffers and therefore played no role in incentivizing providers. And since we now know that they didn’t help people use health services better, user fees helped no one. Removing them was a good idea. But instead of replacing them with systems that held providers accountable for performance, they were replaced with—nothing. This neglect was not benign.
Three recent studies, using rigorous empirical methods, provide valuable insights on the effects of changes in user fees on health care use and health outcomes.
1. Jessica Cohen and Pascaline Dupas undertook a field experiment in Kenya where some people had to buy bed-nets (a user fee) and others got it for free. They show that: (a) take-up dropped by 60 percentage points when the price increased from 0 to 60 cents; and (b) families were just as likely to use (hang) the net if they got it for free as if they paid for it. In their words: “We find no psychological effect of price or the act of paying on usage.”
2. In a randomized control trial that looks at both use and outcomes, Tarozzi and others find that free distribution dramatically increased “previous night usage rates”, but had no impact on outcomes. In their words: “Most strikingly, we find that neither micro-loans nor free distribution led to improvements in malaria and anemia prevalence, measured using blood tests.” The authors conjecture that the disconnect may have come from insufficient bed-net coverage, as the outcome benefits begin once more than 80 percent of households start using the product, something that their usage rates fell short of.
3. Ansah and colleagues in Ghana randomly assigned 2194 households with 2592 children into those who continued to pay the normal user-fees and those who received free primary care including drugs. The results?
a.There was a substitution away from informal to formal care (chemical sellers to primary care clinics) for households with lower user-fees. This is the usual evidence for user-fee reductions.
b.There were no impacts on health outcomes. In the words of the authors: “The primary outcome of moderate anemia was detected in 37 (3.1%) children in the control and 26 (3.2%) in the intervention arm…Mean Hb concentration, severe anemia, parasite prevalence and anthropometric measurements were similar in each group.”
In short, use increased but outcomes did not.
What’s going on?
It turns out that the two statements “there was a massive increase in use” and “there was no increase in health outcomes” are not only consistent but likely to be the norm when there are substitutes.
Consider the following example.
Jhanta likes apples and every apple a day he eats keeps the doctor away. At his supermarket, there are two types of apples—green apples and red apples–and he likes them both (exactly) equally. His decision rule is then very simple—he buys whichever one is cheaper.
On Monday of the first week of May, the green apple cost $1 and the red apple $0.99. Jhanta bought 7 red apples.
On Monday, the second week of May, the green apple cost $0.99 and the red apple $1. Jhanta bought 7 green apples.
Note what happened: The price of the green apple decreased by 1 cent from $1 to $0.99. Yet, the use of green apples went up by an amazing 7 units. Replace green apples with formal public health care and the price with user-fees and you have the basic result: A small reduction in user fees (imagine the price went down because the government subsidized green apples) led to a dramatic increase in use. But would this increase health outcomes? Not at all. Irrespective of which week we look at, Jhanta bought (and ate) 7 apples, successfully keeping the doctor away for an additional 7 days in each case.
The point is that interpreting changes in use and their relationship with changes in health outcomes as a result of price reductions depends not only on the service whose price you changed, but the availability of close substitutes. In the Ghana case, households in the study substituted formal care for informal care; the impact on outcomes depends on the quality comparison between the two. Economics (and common sense) tell us that the closer the two types of services were in quality, the more price sensitive people would be. But the closer they were in quality, the less of an impact the switch would have on outcomes.
What do we know about substitutes for formal public care in the health market of low-income countries? Until recently, there was no study of quality among different types of providers in the health market that could answer the following simple question: Suppose the same patient with the same set of symptoms (say, chest pain) went to representative samples of different types of providers, what would be the accuracy of the treatment that he/she received? (This question is conceptually different from many studies of public and private care that focus on the knowledge of health care providers or specific components of their practice.)
A recent study in Health Affairs by Das and others tries to rectify this gap in our knowledge. They trained 22 standardized patients (each for 150 hours or more) to present the same symptoms to providers in the public and private sectors and providers with and without training in rural and urban India. They report that 67 percent of health care providers who were sampled reported no medical qualifications at all, and more surprisingly, 63 percent of interactions in the public sector also took place with clinical staff without any medical training—a consequence of high absence among doctors.
“What’s more, we found only small differences between trained and untrained doctors in such areas as adherence to clinical checklists. Correct diagnoses were rare, incorrect treatments were widely prescribed, and adherence to clinics checklists was higher in private than in public clinics”
Although the resulting small sample of interactions with formally trained doctors in the public sector does not allow them to statistically verify the claim, quality as measured through adherence to clinical checklists was higher among informally trained private sector providers relative to formally trained public sector doctors. They speculate that this is not because of equally poor training, but due to incentives. Informal sector care was not only a close substitute; it was probably better than formal sector care!
By throwing the price incentives of user-fees for providers out with the disincentives of user-fees for users, we may well have thrown the baby out with the bath water. We need to strengthen systems of basic accountability on the supply-side for health care in low-income countries.
This is easier said than done. It most likely involves a process rather than a one-shot solution. A number of different accountability systems—ranging from administrative and peer accountability to pure price incentives—could be introduced, but each will require considerable experimentation, learning and tweaking. Vouchers or simple results-based financing schemes may not be sufficient, largely because they replace the dense information that the market uses for pricing with crude administrative price setting. Although even basic incentives can help improve use and clinical quality, it can also go terribly wrong: For instance, if an insurance scheme prices hysterectomies too high relative to what it was before, health care providers will increase the use of unnecessary hysterectomies.
We can debate how this can be done, but the first step is setting up an institutional framework that allows these experiments to be carried out and for the results of such experimentation to feedback into the evidence-base of the policy maker. Without this basic feedback system, we are likely to get it wrong—again.
Your story of the green and red apples is interesting, but it misses an important aspect of health service utilization - out of pocket payment. There's nothing new in users switching from uncovered to covered services. Many studies found the net effect is roughly zero. However, if people can have the same volume of services but pay less, that should improve their welfare. The literature on health shock and consumption smoothing in developing countries suggests that high out-of-pocket payment could have an impact on nutrition status and other outcomes that we care about. Establishing a rigorous causal story is undoubtedly challenging but it doesn't mean that the effect doesn't exist if 3 studies didn't find it.
Louise, thanks for your comment. You are right that we need to make sure the system works properly before removing user fees. But simply ensuring that sufficient health workers are trained and deployed may not be enough. As we see from the Das et al. work, better trained health workers don't necessarily provide better service. And we know from the absenteeism studies that simply deploying health workers doesn't mean they are actually present (absentee rates in India are about 40 percent). We need to make sure that health workers have incentives to delivery quality health services to poor people. This is a tall order, but one that we should attempt to meet. Shanta
Rather than challenging the view that user fee removal improves health outcomes, does this not just highlight the fact that user fee removal ALONE is insufficient. In order to meet increased demand and ensure quality care, health systems need to be invested in to ensure sufficient health workers are trained and deployed, etc?
You're right that whenever there is a subsidy, people who consume the subsidized good have money left over to buy other things. This is true for green and red apples as it is for subsidized (or free) health care. There is still the question of whether the size of this "income effect" is greater than the size of the subsidy to evaluate overall impact on welfare. I should add, though, that this point is different from the one we're making, which is about the effect of user fee removal on the supply of health care. Systems of accountability are important in the delivery of health care, and if direct prices are not to be used, something should replace it.
Rob: Thanks for weighing in. On performance-based financing, this is exactly the kind of schemes that need to be introduced to strengthen accountability of service providers in the absence of user fees, which is what we're advocating.
Your experience and anecdotes are valuable, but we have yet to see evidence on public v. private systems of care that answer the simple question we posed in the blog: If the same patient with the same symptoms went to different providers, what would be the accuracy of treatment received? We would welcome any reference you may have.
The image of the poor woman with a sick child is evocative, but is tangential to our blog post. We are not advocating private financing. We say two things: (i) evidence of increased use is not evidence of improvements in outcomes (and we suggest why this is so); and (ii) when eliminating prices as an accountability mechanism, something needs to replace them. This could be a variety of systems, so we need to experiment, learn and tweak before deciding on one system.
An interesting blog Shanta and Jishnu, but I am not sure you are raising anything new here and I fear you are muddling up a number of issues.
Nobody is doubting that health workers respond to financial incentives, but I would challenge what would appear to be your overall argument that direct fee-for-service payments to providers results in BETTER performance. They can, as has been shown by some of the Bank’s performance-based-financing work but these successes are associated with improved PUBLIC purchasing of services not private financing through patient fees. See for example the excellent example of Burundi where the Bank is celebrating the stunning success of the Government in replacing patient fees with performance related public financing: http://www.worldbank.org/en/news/feature/2012/09/24/burundi-investing-i…
Why did the government implement this policy? Because it recognized the fundamental market failure associated with healthcare provision – the massive asymmetry of information between the consumer and the supplier. All over the world where user fees are charged health care providers take full advantage of these incentives to oversupply services to vulnerable consumers. From personal experience I remember an employee of ours in Congo being charged $20 for 5 medicines to treat malaria – but which didn't include ACT. This exploitation can reach terrifying proportions (see this brief video): http://www.clipsyndicate.com/video/playlist/5895/1343344?cpt=8&title=he…
In fact the free health care policy in Burundi came about because the President realized that government hospitals were being turned into debtors prisons http://www.hrw.org/reports/2006/09/06/high-price-pay-0
Across the world governments are therefore realizing that, in general, public purchasing of health care (which may involve private sector administration and the use of private providers) is better than private purchasing of care. Sure we can look at including incentive mechanisms into these purchasing systems but public purchasers are much more likely to secure improved efficiency and equity than a poor woman with a desperately sick child clutching a fistful of grubby bank notes
Thanks for this insightful discussion. I hope you don't mind if I add one clarification about the results of my paper with Aprajit Mahajan, Brian Blackburn et al. on ITNs sold on credit. The lack of improvements in malaria prevalence in treatment (that is more ITNs) areas of course has not changed, but in the new, revised version of the paper we also show that self-reported six-month malaria INCIDENCE did decrease considerably in treatment areas. Recall that prevalence tells you the fraction of a population who suffer from a condition, while incidence tells you the number of cases in the population over a longer period of time.
Of course self-reported information is not as reliable as blood tests, but it does have the benefit of concentrating on episodes severe enough to be symptomatic and hence recognized by the households (although it is again not clear that such cases were indeed malaria!).
In the paper we discuss at length why such discrepancy between prevalence and incidence may have taken place. In short, we go back to the idea (highlighted by Shanta) that the low coverage with ITNs is likely at least partly responsible, and we discuss how this is consistent with a number of earlier studies on the effectiveness of ITNs.
Let me also add (and sorry for being pedantic!) that a threshold for externalities/community effects of ITNs to start kicking in has not to the best of our knowledge been conclusively and experimentally established, and indeed a number of models suggest that there isn't a precise threshold, with even some ITNs providing some benefits to non-users. The new version of the paper (that can be found here http://www.econ.upf.edu/~tarozzi/TarozziEtAl2013RCT.pdf) provided lots of additional details and citations in case anyone want to know more.
I would like to share some observations on the studies referenced in the blog as they perhaps merit some further assessment:
--On the Ansah study: "The purpose of Ansah and colleagues' study was not to examine wide-scale national experiences of abolishing user fees, as happened in countries such as Niger and Uganda. Rather, the study was a pilot project on free access to a prepayment scheme in the Dangme West District in southern Ghana, and to verify whether free access through pre-payment schemes would improve members' health. Thus, what was being tested was the effectiveness of pre-payment schemes, more than the abolition of user fees."
--On the Das study in HA, perhaps the interpretation of another World Bank staff member, Adam Wagstaff, is relevant to the discussion: "The results on the public-private differences are pretty interesting. In rural MH both sectors do equally – and very – badly in terms of ensuring the patient gets the right treatment. In Delhi by contrast going to the private sector halves the odds of getting the correct treatment, even though it raises the number of recommended questions the provider asks. It is the latter quality indicator that Das et al are presumably referring to in their conclusion when they say “we observed better quality care in the private sector”. That’s a bit misleading – it’s surely the correctness of the treatment prescribed that matters at the end of the day, not the number of questions asked. This work isn’t exactly a great advertisement for India’s private sector, or for the view that financial incentives will improve quality. But it’s not exactly a great advertisement for the public sector either. It also begs the question: Why does the private sector in Delhi do worse on the correctness-of-treatment indicator while the private and public sectors in rural MH do just as badly as one another? " http://blogs.worldbank.org/developmenttalk/shocking-facts-about-primary…
--On the Cohen and Dupas study - this is what the authors of the study concluded from trial results : "Overall, our results suggest that free distribution of ITNs could save many more lives than cost-sharing programs have achieved so far, and, given the large positive externality associated with widespread usage of ITNs, would likely do so at a lesser cost per life saved."
Also, it is important to note that there are other studies that show the opposite conclusion than the one in the blog: for example, Partners in Health cite a number of studies in support of abolishing fees...http://parthealth.3cdn.net/71d53fd770a3650456_zrm6bt8j6.pdf.
I guess the question then is whether universal health coverage (UHC) is worth it - in the sense of expanding access (assuming that financial protection and peace of mind matter). Can we prove UHC improves outcomes? And the answer has to be yes if the quality of care is good...http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(12)61485….
Your question about whether UHC (people effectively covered by pooled health financing) improves population health outcomes is answered in this excellent publication from Results for Development written by Bill Savedoff and colleagueshttp://t.co/zNdm5VdX . Here research by Peter Smith and Rodrigo Moreno-Sera (in section 2) proves that UHC does have a positive impact on health outcomes. Furthermore on page 11 they show that improved outcome indicators are only associated with pooled PUBLIC financing. “Voluntary [private] health insurance appears to have no measurable impact on population health status”. Furthermore “countries in which out-of-pocket spending is a higher share of total health expenditure have higher adult mortality rates.” ie user fees kill people.
I can’t recommend this paper enough to people interested in UHC and health financing. It’s overall message is that UHC is good for population health and that public financing (driven by economic growth and political negotiation) is the key to UHC.