Future Development

The inconvenient truth about public and private health care

Shanta Devarajan

Everyone agrees on the following fact: In low-income countries, the lion’s share of health care is provided by the private sector. In rural India, the private sector’s market share is around 80 percent. Yet, there is disagreement on the interpretation of this fact. One view is that the free, public health system is so underfunded and understaffed that poor people are forced to pay unqualified and unregulated private-sector doctors (often called “quacks”) for treatment. A completely different interpretation is that private-sector doctors, by dint of getting paid by the patient, provide better service—for starters, they may actually be present in the clinic—and that is why poor people choose to bypass the free public clinic and go to the private doctor.

We may never resolve this disagreement since it is based on differences in values (is health a right, like freedom, or a service, like transport?), economic models (what is the rationale for public intervention?) and estimates of parameters (do private doctors respond to incentives?). Attempts at shedding light on the last point are plagued by measurement problems, because public and private doctors see different mixes of patients, and different people choose to go into the public and private sectors, so their differential responses to incentives—including the quality of care—cannot strictly speaking be compared. That hasn’t stopped people from trying to compare public and private health care, with divergent results, most of which are not robust. Out of 182 studies in the literature, Coarasa, Das and Gunnerson find only two that control for differences in the mix of patients and in the types of doctors in the two sectors.

In an important paper published last month, “Quality and Accountability in Healthcare Delivery,” four of my colleagues and friends, Jishnu Das, Alaka Holla, Aakash Mohpahl, and Karthik Muralidharan (DHMM), have undertaken a careful comparison between the public and private sectors in rural Madhya Pradesh, India. They coached a set of “standardized patients (SPs)” to present the symptoms associated with three conditions—unstable angina, asthma, and dysentery in a child (who is at home). They then sent these SPs to randomly selected public and private providers and compared the physicians’ responses against a set of metrics (checklists of questions and examinations, likelihood of making a correct diagnosis, and appropriateness of the treatment) to gauge the “quality of care.”

The results are, to put it mildly, striking. Even though they were mostly unqualified, the private providers exerted significantly higher effort and were no worse in providing the right diagnosis or recommending proper treatment than their public-sector counterparts. And this is in a context where the overall quality of health care in rural India is quite poor.

To isolate the effect of practice type and control for differences in qualifications, DHMM went further and looked at a sample of qualified public doctors who were also in private practice. They find that the same doctors spent more time with patients, diagnosed them better, and were more likely to offer correct treatment in their private practice than in the public clinics. Noting that “free” public health care is not free to the taxpayer, they compare the per-patient cost in the two sectors and find it to be four times higher in the public sector.

Taken together, these and other results in the paper, as well as Jishnu’s previous work with Jeff Hammer, call into question statements such as “health care should be provided free by the government.” Unless the doctor’s pay is somehow linked to performance, there is a good chance that the quality of care in the public sector would be worse than in the private sector.

This is not an argument in favor of user fees. The public doctor’s pay could be provided by the state, but linked to performance, as has been introduced in some results-based financing initiatives around the world. However, such schemes are often resisted by medical unions, not least because they threaten the rents enjoyed by public doctors (such as those in Madhya Pradesh), whose pay is currently independent of even their presence in the clinic. Furthermore, demand for health care rises with income, so giving it out for free to everybody is a regressive (i.e. pro-rich) subsidy. If, to improve the quality of care and avoid subsidizing the rich, governments do introduce user fees, they need to make sure that poor people can afford to pay these fees by giving them vouchers, targeted cash transfers, and the like.

As DHMM note, in poor governance and weak public-accountability settings (which characterizes most low-income countries), market-based provision of health care, with its built-in accountability mechanism, is a viable alternative—not to the ideal public sector, but to the one we find in the real world.


Shanta Devarajan

Chief Economist, Middle East and North Africa Region - World Bank

This blog was first launched in September 2013 by the World Bank in an effort to hold governments more accountable to poor people and offer solutions to the most prominent development challenges. Continuing this goal, Future Development was re-launched in January 2015 at

For archived content, visit »

Get daily updates from Brookings