Name your price: it’s your life

Published online January 25, 2012.

Why small increases in price can lead to a steep decline in demand for essential products

A piece of nylon netting is a useful thing. It can be cast as a fishing net, hung as a curtain, or draped over a seedbed as protective covering. Netting can make a stunningly white wedding dress, or even a make-shift chicken coop.

One can also sleep under it, of course, to keep mosquitos from biting at night. Though insecticide treated nets (ITNs) are routinely distributed in malaria endemic regions, often subsidized by major donors such as the Global Fund, many worry that such campaigns are frequently futile. Anecdotal evidence from the Kenyan shores of Lake Victoria to the alters of Ugandan churches suggest that these bednets are sometimes quite literally cast aside or otherwise misused.

While misuse is certainly problematic from the perspective of those funding mosquito net campaigns, it also raises a broader question, and one with serious implications for public policy in malaria prevention and beyond: Do people value and use things that are given to them for free?

There are two competing arguments used to answer this question. The first argument says that people value more that on which they spend their own money or resources. Furthermore, people will spend some money, when they can afford it, on those objects that they perceive to be useful. A second argument says that if an object is perceived to be useful or of value, people will use that object regardless of whether they purchased it or whether it was given to them for free. The ubiquity of incumbent presidents’ campaign t-shirts in both opposition and stronghold areas is supporting evidence for those in the latter camp.

The mosquito net-cum-wedding dress is a classic illustration of the dilemma of freebies. The protective power of mosquito nets against mosquito bites and thus, malaria, is rather less effective when the net becomes a nuptial adornment or is tossed into a river, much less left in its packaging and stashed in a corner. The creative use of nets is thus often the go-to anecdote for those in the first camp of the freebie question.

Anecdotal evidence, unfortunately, can only get us so far in adjudicating between these two perspectives. Fortunately, a number of development economists have been systematically evaluating the extent to which people use services or tools given to them for free and those provided at a cost. While there is still no definitive answer, and while context matters, much of the evidence seems to suggest that people use many free goods at high rates, and often will not purchase the same products when provided even at very low prices.

A group of researchers at the Abdul Latif Jameel Poverty Action Lab, based at the Massachusetts Institute for Technology (MIT), recently wrote a report summarizing ten studies examining the question of whether user fees and cost-sharing increase or decrease the use of health and education services and products. The majority of the studies were conducted in Kenya, although some were also conducted in Uganda, Zambia, and India. Their findings are striking, and the title of the report says it all: “The price is wrong.”

Time and again, small increases in price lead to a massive decline in demand for products including water disinfectant, deworming medicine, mosquito nets, and soap. For example, one study in Kenya found that while over 80% of people used a mosquito net if they received it for free in a prenatal clinic, only 20% would purchase the net for $.60 (approximately 50 Kenyan shillings or 360 Rwanda Francs). Similarly, another study in Kenya found that while nearly 60% of people used water disinfectant when it was given to them for free, less than 10% would use disinfectant if charged $.30 for the same product. This general pattern appears to repeat itself in different locations and with different products.

Two things are thus evident. First, people are often unwilling to purchase a number of goods and services that promote health and education even at highly subsidized rates. Second, people often use those same goods and services at high rates if they are provided for free. Clearly, receiving something for free does not preclude its use. If we think back to the wedding veil problem however, it is also clear that some products may not be used as prescribed, fee or no fee.

Why are people so sensitive to price when it comes to potentially life-saving goods and services? Individuals and families weigh the costs, monetary or otherwise, of procuring and using goods and services against the expected benefits from using those goods. Bednet wedding veils notwithstanding, in most cases it appears that families perceive some benefit from using goods like mosquito nets and soap, since rates of usage are quite high when the product is free. Some speculate that people may not physically have the cash on hand to buy even very inexpensive products, or that other inconveniences, such as the time it takes to procure a product, may affect their decision. But these are only partial explanations. It is also possible that people do not believe products will be as efficacious as researchers and policymakers think they will be in promoting their health.

Available evidence suggests that people who receive goods and services for free often do use them, although the extent to which they will use them and how they will use them is subject to some debate. Even if there are large benefits to providing free bednets, water disinfectant, soap and the like, products that often provide benefits that extend beyond the individual recipient, the question of sustainability comes to the fore. In the short term, the provision of free goods and services, particularly those that promote preventive health behaviors (like hand-washing) may have large and positive effects on the health of families and communities. But ultimately, we need to better understand why people are often so unwilling to spend even small amounts on products that have the potential to keep their families much healthier.

Explaining health behavior

Pascaline Dupas has an excellent paper in the Annual Review of Economics: Health Behavior in Developing Countries. It’s well worth reading. Conclusion below:

Good health is both an input into one’s ability to generate income and an end in itself. As such, it is not surprising that a relatively vast literature is devoted to understanding the determinants of health behaviors. This literature has recently expanded to the study of health behaviors in low-income settings, for which good data are becoming increasingly available. This review is too short to be exhaustive, but it tries to present the most compelling evidence to date on this issue. The important thing to take away from this review is that when it comes to health behavior in developing countries, there are a substantial number of deviations from the neoclassical model. First of all, people seem to lack basic information, and sometimes have limited ability to process information, because of low education levels. Second, there are market imperfections and frictions, especially credit constraints, affecting people’s ability to invest in health. Finally, there seem to be some deviations from the rational model, with, as has been widely shown in developed countries, a nontrivial share of people exhibiting time-inconsistent preferences as well as myopia.
Overall, this suggests an important role for public policy when it comes to health. Above we identify four important demand-side policy tools: information, mandates, price subsidies, and financial incentives. All appear to have the potential to increase the sustained adoption of preventive behavior. But the success of these demand-side strategies is contingent on the supply side being adequate: on health services and products being available, with delivery and/or enforcement institutions that are effective. The issue of how to improve service delivery in health is outside the scope of this review, but it has been the focus of a number of recent and ongoing studies that will soon need a review of their own.

Science in the time of cholera (and nodding syndrome)

Published online January 11, 2012.

In August 1854 a terrible illness tore through a London neighborhood, killing hundreds in a matter of days. The terrifying disease emptied the body of fluid until vital organs shut down, after which point the petrified soul would succumb to the illness. Death often arrived less than twelve hours after the first signs of an upset stomach. Londoners of the day had a name for this illness, but did not understand its cause. They called it cholera.

Though cholera outbreaks had hit London before the mid-1800s, the Broad Street Pump outbreak of 1854 is now perhaps the best known. It was during this scourge that physician John Snow was able to demonstrate that cholera was not an airborne disease, as was the popular and professional opinion at the time, but rather a waterborne disease. This insight proved critical to improving public health in London and beyond. Londoners had been emptying their waste into the Thames, often just upstream of intake pipes for water companies. Their water and city stunk. But because disease was thought to be airborne, they doused smelly sidewalks in chloride of lime in attempt to purify the air. They made few attempts to purify the water so obviously contaminated with their own waste.

When cholera inevitably struck, they applied all manner of remedies, most of them useless at best. Castor oil, opium, and leeches were all espoused to treat cholera, not just by ordinary folks, but also by doctors. Worse still were treatments such as laxatives or bleeding. The extreme dehydration facilitated by cholera was often “treated” by attempts to further remove fluids from the body.

In hindsight, both the cause and the treatments for cholera are straightforward, if not obvious. Cholera is a waterborne illness that spreads when one person ingests the cholera-infected waste of another person. The treatment for the extreme dehydration that ensues is most fundamentally rehydration – consuming copious amounts of fluid to replace those that are lost. Yet at its emergence, a series of facts and observations did not at first fit together in a single theory about the cause of cholera. When cholera struck a household, sometimes it struck everyone, sometimes just a single person. In a neighborhood, some homes would be hard hit, while others escaped untouched. Whether you survived or not seemed random.

So it is with another illness in our midst – nodding disease. Nodding disease sounds like a folksy and tabloid-inspired syndrome. Its name describes the telltale symptoms of the disease, a rhythmic head nodding in children. The fact that unlike many diseases its name does not betray anything about its likely causes demonstrates just how little we know about its transmission. For example, HIV (human immunodeficiency virus) is named for the virus that causes AIDS. The name malaria comes from the Italian mala aria, meaning “bad air”, so named because the illness we now know is caused by a parasite was originally thought to be airborne.

But despite its odd name, nodding disease is far from folksy or fake. It is often fatal. First reported in Tanzania in 1962, nodding disease has since spread throughout what is now South Sudan, and has been rapidly spreading in northern Uganda as well.

The pattern of incidence of nodding disease and its symptoms are puzzling, as were those of cholera in the early nineteenth century. First, the onset of nodding disease appears to occur almost exclusively in children between the ages of 5 and 15.

Second, nodding is reportedly often triggered by the presence or eating of familiar foods, or when a child becomes cold. Unfamiliar foods, such as chocolate candy bars, do not induce nodding. Third, when untreated, those with nodding syndrome cease developing both physically and mentally. They are often stunted and experience mental retardation. Fourth, most children affected come from very poor families. There are now thousands of children in South Sudan and northern Uganda who experience symptoms of nodding disease, and the incidence of the syndrome appears to be increasing.

Several theories regarding the cause of the syndrome have been mooted, but none proven. For the past several years, teams of experts from the U.S. Centers for Disease Control (CDC) and World Health Organization (WHO) have travelled to South Sudan and northern Uganda in an attempt to better understand the causes of nodding disease, and possible treatments. Their work suggests that nodding disease is a new epilepsy syndrome, and that the characteristic head nodding is caused by seizures that lead to temporary lapses in neck muscle tone.

A vast majority of children experiencing symptoms of nodding disease are also infected with a parasite called Onchocerca volvulus, which causes river blindness. The high prevalence of this parasite in victims of nodding disease means that the most plausible (published) theory about the cause of nodding disease links the syndrome to O. volvulus, but how and why remain unclear. Moreover, there are a number of children both in and outside the region who are infected with the parasite and do not acquire nodding disease, so the link between the two is not straightforward.

So far, therefore, we have accumulated a series of facts about the mysterious syndrome, which have yet to be pieced together in a coherent theory. We have many more tools at our disposal than did the Londoners of the 1800s, but answers to pressing medical and public health questions do not usually come without time and resources. Nodding disease is a terrifying prospect for those living in South Sudan and northern Uganda not only because of the debilitating effect it has on children, but also because families and communities do not understand why their children are falling ill in the first place. A confusing array of facts, theories, and observations are unnerving both to those in the midst of the outbreak, but also those who see its spread as a very serious health issue for the region.

Misunderstanding the causes of nodding disease can have disastrous consequences, as was the case with cholera some 150 years ago. So far, anti-epilepsy treatments appear to be helping children experiencing nodding disease, but supplies of these treatments are often scarce, and determining the ultimate cause of epilepsy in these children should be a high priority for health officials. Cases of epilepsy are often documented at high rates in hospitals in the region, and there is thought to be a link between epilepsy and cerebral malaria as well. In Arua Regional Referral Hospital, in northwestern Uganda at the border with Sudan and DRC, 7 percent of all outpatient children over age 5 in April 2009 were diagnosed with epilepsy. In 2004/05, 74 percent (nearly 4500) of all cases in the Mental Ward were diagnosed as epilepsy.

Clearly, epilepsy, whether nodding disease or otherwise, is a condition that deserves the utmost attention from public health and medical professionals. The sooner we understand the causes of this new breed of epileptic seizures, the sooner we can take steps to both treat it and prevent its spread. In the absence of a compelling theory about its cause, however, fear and futile treatments are likely to ensue.

felled by fever

As my partner-in-crime was felled by a fever this weekend, I got to wondering how often people treat themselves for malaria when they really have a nasty virus, flu or otherwise. If you don’t have the time, resources, or energy, it might seem like a good idea to pop some anti-malarials (assuming you can get them) just in case.

I gave a presentation about health services and malaria in Uganda several weeks ago, in which, among other things, I bemoaned the lack of attention malaria receives from government. Browsing various publications, studies, and policy reports, I mentioned several stats, including the following:

  • Malaria is the cause of 32% of child deaths in Uganda (DHS Child Verbal Autopsy 2007)
  • 42% of children tested positive for malaria during the DHS Malaria Indicator Survey 2009 , compared to 0.7% in Ethiopia, 2.6% in Rwanda, 7.6% in Kenya, and 18% in Tanzania.
  • Malaria is responsible for 30-50% of all outpatient visits, 15-20% of all admissions, and 9-14% of all inpatient deaths
  • Uganda ranks third in the world in terms of malaria deaths

One of the audience members asked about the accuracy of reporting of malaria cases in Uganda. While malaria is undeniably one of the most important health challenges Uganda faces, it is important to acknowledge that the capacity to diagnose malaria is generally weak, and many if not most of the malaria cases and deaths are not laboratory confirmed. There is probably a sizable chunk of these “malaria cases” that are not actually malaria, but rather a flu or some other virus or infection.

The 2009 Malaria Indicator Survey found that of the 3,727 children included in the survey, 44.7% were reported to have had a fever in the preceding two weeks. While 70% of children with fever were taken to a health facility or health provider, only 17% were reported as having been tested for malaria through a finger or heel prick. 60% of children with fever ended up taking anti-malarials, and 15% took antibiotics.

I’m still astounded that 42% of the children in the survey tested positive for malaria (62% were anemic). This figure is especially high when you compare it with other countries in the East African region (see above). Prevalence varies quite a bit by region as well.

Source: Uganda Malaria Indicator Survey 2009, page 61.

The internet has slowed to a crawl, but I’ll post some more links on this soon.