Friday, November 09, 2012

Blindly sending money down leaky pipes

Proposals to spend more on government programs in India are generally criticised on the grounds that this is sending more money down a leaky pipe. In addition to the problem that the pipes leak, there is an equally big problem that we have no idea about what happens at the other end.

In order to build and refine a system, the first foundation that has to be laid is that of measurement. What you measure is what you can manage. In India, all too often, government agencies and programs start out with lofty ambitions, and embark on spending money to get there. But there is little measurement about the extent to which those objectives have been achieved. Under these conditions, there is little chance of programs being designed properly, and of wastage and theft being checked.

I was reminded of this as I read As Dengue fever sweeps India, a slow response stirs experts' fears by Gardiner Harris in the New York Times. There may be an epidemic of Dengue out there. Or there might not be one. The point is, we just don't know. The statistical system simply does not measure this.

A public goods perspective

What should government do, and what should government not do? The government should work on the provision of public goods and stay out of private goods. In the field of health, what are the public goods and what are the private goods?

When I have a toothache, and I go to a dentist, and I get better, this is a private good. Yet, most government spending is oriented towards building `primary health centres' and hospitals and such like. Even if these worked well -- i.e. even if they were not characterised by theft and incompetence -- they are a bad use of public money as they deliver private goods and not public goods.

A public good is something that is `non-rival' (my consumption of that good does not reduce your access to it) and `non-excludable' (it is not possible to exclude me from benefiting from this good). The best example is clean air. My breathing in clean air does not diminish the amount of clean air available for you. When one more child is born, it is not possible to exclude him from benefiting from clean air.

What are the public goods in health? A few examples that come to mind:
  • Statistics. Measurement of what is going on about health in India.
  • Epidemiology. Tracking down and eradicating Smallpox. Mounting a response to fresh strains of the common cold.
  • Running public systems that measure and ensure that medicines are not counterfeited, are properly stored in a cold chain.
  • Running certification systems. Enforcing against quacks that practice medicine.
  • Getting research done on diseases that matter on India, and releasing the findings into the public domain (i.e. unencumbered by patents).
We in India have this essentially upside down. Health policy in India is unfortunately shaped by the views of doctors, and is low on skills in public economics. We like to focus on Primary Health Centres that are run by the government, and we cut corners on all the five critical public goods listed above.

It is fashionable to say that India should spend more on health. I would advocate spending less on the things that the Indian government does in health. Until the pipes are fixed, we should be closing the taps.

An objectives-and-accountability perspective

The Indian State is in a crisis. The two key factors at work are mission creep and a lack of accountability.

Mission creep has set in because in India, almost any do-gooding is seen as the responsibility of the State. We need to narrow the mission statement of the State to a tangible set of public goods. Clarity of mission, and a controlled and narrow mission, is of essence to obtaining performance.

Consider the principal-agent relationship between you and your contractor. If the contractor is failing to deliver, you would narrow down the specifications given to him, and monitor him tightly to make sure the work gets done. That is precisely what we need to do, in the principal-agent relationship between citizens and the State. The State has failed on a sprawling mission. We need to narrow down the tasks given to the State, and tightly monitor the delivery of results.

Government and government agencies will work well when they have narrowly defined functions and strong accountability mechanisms. In the field of health, absent measurement of health outcomes, there is no accountability.


Is there a Dengue epidemic in India? We don't know.

An information system about the health of the people of India is a public good. It should achieve pride of place in the responsibilities of the State. However, health expenditures in India are squandered on private goods. To add insult to injury, there is theft and incompetence, so even these attempts at delivering private goods do not work so well. But the main point is that running PHCs and hospitals should not be done, even if the Indian State had the ability to run these things well.

In order to reconstruct the Indian State, we need to push on the combination of narrowing the mandate (focusing on a few core public goods) and strong accountability mechanisms.


  1. Cant agree more with you.
    I think whole efforts should be spent to may be start small and start afresh. Doing it right and then slowly increasing it.
    Again important point in here is checking the performance. It shouldn't be done by same dept I would say.
    May be getting direct feedback from patients / employees would do some good.
    Other thought that comes to my mind is to reduce the money factor from PHCs. All the employees get some benefit in some form or other on regular basis which they can encash in other govt offices may be.
    I also see a potential where we can start cheaper medicine prescriptions from PHCs.

    Just tabling some ideas, which again needs to be planned at specific milestones and not to be targeted all in one go.

  2. Same thing with Akash. The recent news articles that accompanied the President's 'launch' of the device mentioned that crores of orders will be placed soon. There are so many thing wrong here that one wonders whether its worth talking about at all.

    1) Mission Creep - why is the govt giving out freebies. I know why it is, but why aren't people/media lambasting them for it, at least in urban India.
    2) Does no one find it odd that the President launched a device created in the private sector, and not manufactured in India? The device isn't an Indian success story, is it? Is the President going to be launching the latest printer that some department now procures for use? How is it any different from any other technology device procured by the govt?
    3) Why has Sibal pussy-footed around with this project? We are the largest democracy. Surely, we could have signed a more comprehensive contract with a much more reputed supplier like Samsung or Amazon or even Apple with a commitment for a long term project with the goal of reducing cost and providing good quality content? Why does the govt go with shady companies like Datawind? GoI could have signed a long term contract and asked Samsung to set up a manufacturing plant in India.
    4) Isn't it hypocritical that while the govt subsidizes this device, it charges high rates of import duties and taxes on all other similar devices? Isn't this corruption where money is diverted from better quality products into a product that the govt has hand-picked?
    5) Why did IIT collaborate in this project. I am disappointed with my alma mater for not having told the minister what a daft project this is. If the IIT director cannot tell Sibal this, then we have a society of sheep that itself points to severe problems.

  3. Dear Sir,

    In this piece you point out that there is little measurement about the extent to which the policy objectives have been achieved. I am assuming that this could be arising out of several reasons, including the one due to lack of sufficient data/information to keep a tab on the implementation status of a particular policy initiative.

    It may be noted that recently the Government took penal action against certain sugar mill owners for failing to submit details of their cane crushing operations. Even while obtaining data to determine IIP, I understand that, the Government leaves no stone unturned in requesting the producers to share the production details. It appears that the Government is taking steps to obtain the required data/information, but due to insufficient response from the public its struggling to achieve the intended goals.

    In light of the above, in your view, what are the other ways in which the Government can take steps to obtain the required data/information, which in turn influences the policy decisions of the Government for a particular sector?

    1. I have a whole section on this blog titled "Statistical system" which has many many posts about the problems of the statistical system.

  4. Hi Ajay,

    Really great post. I very much agree with most of your points and, in particular, your call for better statistics on public health. Sadly, it looks like the government is going in exactly the opposite direction on this by discontinuing the NFHS – currently by far the largest household survey of health in India.
    A few additional points:
    1. I’m not entirely convinced that a heavier government hand is required in ensuring the quality and legitimacy of pharmaceuticals. While incidence of substandard drugs is rife in many parts of Africa, I haven’t seen much evidence that this is the case in India. Maybe it is, but I think that before advocating for another layer of regulation and supervision we should first confirm this. And if there are problems with substandard drugs, it is important to focus on the ones that can do real public health harm if quality is not ensured (e.g. ACT and TB drugs).
    2. While I agree with your “fix the pipes before sending more money down them” approach in theory, in practice this is pretty hard to do. The people at the ends of those pipes tend to get rather annoyed and block the process. A less idealistic but more politically feasible approach is to create a new set of pipes.
    3. While there are certainly a huge number of unqualified medical practitioners in India, the research I have seen suggests that households, even poor ones, are able to distinguish the “Bengali docs” from the others. The fact that they go to the quacks may be surprising to us, but doesn’t necessarily mean that they are unaware of their lack of qualifications.

  5. Great idea. The microeconomic approach to public goods can be very useful. However, it appears that in focusing on public goods, prevention rather than cure will become the guiding principal for public health policy. In a country where a large chunk of population cannot afford the costs of private sector healthcare, is it not the state's responsibility to provide cheaper options?

  6. I admire your courage in writing this despite being in firing range of central govt. I would say that the truth is even more starker - most government programs, and more so in the states, are carefully designed to leak and benefit ruling legislative members in power and to fund the expenses for future elections. Also, for the same reason the governments at the center or the state like to exercise control on enforcement agencies and to let them remain mute on certain aspects of public spending. And the judiciary isn't really as powerful and motivated most of the time.

    1. The role of an intellectual or a leader is as a guide towards the future. How can one guide unless one criticizes the status-quo? In democracy, such writing should be the norm, rather than something needing courage.


Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.