Wednesday, October 12, 2011

In praise of "slavery" or Why education should be equity financed

As an avowed free market libertarian, I am, in general, opposed to Government intervention in the economy. Conceptually, an intervention can be justified only if the nature of the problem is a "market failure" , or if a "public good" has to be provided. Even in such cases, the intervention is justified if and only if the proposed intervention solves the identified problem without inducing unintended collateral damage.

Hence whenever any Government intervention is proposed the tests that I use are a) Is there a market failure? b) If not, are there gaps in financial markets that prevent the market from functioning and then c) Is the cost-benefit evaluation of the intervention better than laissez faire?

The reason I think about financial markets separately is that a well functioning financial market is an unstated assumption in free market theory. Markets for "real goods" can be hampered if people are not able to make intertemporal trade-offs (e.g., borrow money to start a company and then repay it later) or manage risk (e.g., insurance for different "states of the world"- investments may dry up if people think that a single event can wipe out their money)

Having laid out the conceptual framework let me get to a specific example: Recently when I contested the case for Government intervention in education by a friend, he asked me how I would view the problem? Without Government intervention (either by being the provider of free/subsidized education or being the payor by schemes such as school vouchers or Government mandated low interest rate education loans), wouldn't the market for education just collapse- perpetuating poverty since the poor will not have access to education?

The argument seems sound: While education is not a public good in the economic sense (since it is definitely excludable, though it is somewhat non-rival), for a poor person, the benefits of being educated are clear, but how would he ever raise the money to pay the fee?

Is there a gap in financial markets to provide the same? Why have student loans not solved the problem?

For that take a project finance view to education (that is, consider that educating a person is a project. An applicant comes to you (a financier) and asks you for money to finance his education. How would you view this proposal in purely financial terms?). What are the characteristics of this project in financial terms:

1. The payoff is real, but variable: Education definitely leads to an increased economic payoff to the beneficiary (i.e. the applicant). However the degree of benefit is highly variable (see the high distribution of income among all people that you attended school with). Hence the financial requirement is for risk capital, since the financier has to be willing to live with variable payoffs.

2. The payoff is deferred: The returns to education typically accrue after a time gap (for school education the gap is considerable- maybe as high as 15-25 years). Hence the financial requirement is for patient capital, the financier should be willing to defer his expectation on when his money would be returned

Hence from a project finance standpoint, the financier needs to provide patient risk capital. In finance, the capital that has this characteristic is not debt (that is loans) but equity. Equity implies taking a risk, and being content with getting the payback much later. But in equity, when the investment does payoff, you get a much higher return (unlike debt where the returns are fixed)

The missing gap in financial markets is that you cannot buy equity in people (such contracts are generally not allowed in law). If I could sell a part of myself (in financial terms) I may be able to raise the money needed for education.

Markets would arise where we could trade shares in people. Just as in financial markets for companies you have equity investors at various stages of financing, based on their own risk appetites and expertise (for example, angel investors for startups, venture capitalists for fledgling companies, private equity firms for different stages of growth, and the capital markets for more mature companies) we would have various investor profiles for investing in people too.

For example I may make a bet that out of 500 schoolchildren in a village in Bihar,my assessment is that at least 2-3 people may end up as high income earners. If I invest by taking a stake (say 50%) in each of them and pay for their school fees, they will owe me 50% of their income in perpetuity.

The crucial point here is that many investments may fail, but the successes could be so huge so as to justify the investments across the entire portfolio (just as VC and PE firms do with companies).

Such an investor, would at a later stage, offload some of his stake in specific people, to later stage investors. Financial aggregators would warehouse risk across multiple investments (for example, I will create a scrip tha is an aggregation across 10,000 IIT students which I will trade; this may raise the money to finance their MS/MBA in the US in top schools), just as mutual funds and other funds do for companies.

In industries of a "superstar" nature, such contracts actually exist! In an industry like book publishing, or music, the distribution of incomes is very high- successful people earn huge amounts (like  Lady Gaga or a writer like J.K. Rowling), unsuccessful ones end up with no money. Music companies or book publishers typically take up "rights", that is a percentage of the revenue from the sales of the book/ album. This is a equity like contract.

But such a contract would not be possible for the person's earnings as a whole- which is why they don;t exist for other industries.



Look forward to the day when such an innovation would be allowed!

Tuesday, October 11, 2011

Arthashastra and Antibiotics: Alternative approaches to disease control


The latest news in the public health sphere in India is the fear of a new antibiotic-resistant "superbug"- namely the NDM-1 gene http://bit.ly/rdihtz.

In many ways, this is the latest illustration of how we may need to try alternative approaches to the field of communicable disease control.

First, the context: "Superbugs" are newly evolved species of bacteria that are antibiotic-resistant. To defeat the menace of superbugs, we need to understand what fundamentally leads them to arise.

The answer is antibiotics itself! In essence. the mechanism is as follows: "Simple" forms of bacteria infect us, and if their effects are malign (not all bacteria are harmful to human beings), we seek to treat them with antibiotics.

The use of these antibiotics effectively causes the evolution of superbugs.

Now, it so happens that the rate of bacterial evolution is high. The "rate" of evolution is governed by 2 things- how often an organism mutates (scientists still do not understand what governs this) and how efficient natural selection is in weeding out "bad" mutations ("Bad" here refers only to the ability of the mutated organism to survive and reproduce, not to its effects on other organisms)

We don't have control on the first mechanism (that is, the rate of mutation)- it so happens that bacteria mutate really fast. But in using antibiotics, we are ensuring that "weak" or "bad" mutations are weeded out even more efficiently.  In other words, we are helping the superbugs develop faster!

Another way to explain this is to say that our use of antibiotics makes bacteria "want" to develop antibiotic-resistant forms (this is the shorthand that we often use, which is misleading since evolution has no agency, or purpose- it just is. But using this shorthand helps us understand the phenomenon better)

This is an arms race that we are condemned to lose- for example, the rampant use of methicilin  to treat staph infections led to the rise of MRSA (methicilin resistant staph infection), the use of vancomycin to treat MRSA then has led to the the evolution of VRSA (vancomycin resistant staph). We are chasing a moving target, one that moves much faster than we can.

But then, what choice do we have? Clearly, the answer cannot be to leave bacterial infections untreated.

An alternative approach is to actually help the evolution of less malign bacteria- to come to some kind of mutual accommodation with bacteria rather than look to vanquish them.

Let me use a concept from Kautilya's Arthashastra (the seminal text on the use of pragmatic realpolitik to achieve one's strategic goals) to illustrate the point.

In Arthashastra, Kautilya delineates 4 ways to deal with any adversary- "sama" (use of logic or argument), "dama" (gift, bribery or blandishment), "bheda" (sow dissension or confusion) and "danda" (use of force)

Fundamentally, our way of dealing with bacterial disease control is "danda"- destroy the adversary with force.

Now, that is a smart strategy only if the power asymmetry is disproportionately in your favor, and the costs of using force are low.

This is clearly not the case with the use of antibiotics.

So why not try an alternative approach- one inspired by "dama"? Can we bribe our way towards more benign forms of bacteria?

One radical approach in disease control (pioneered by the evolutionary biologist Paul Ewald) actually is looking to do that.

The basic concept to understand is on how diseases spread. Some diseases, like a common cold, are deliberately not virulent- the reason is that by not debilitating you severely, it can keep you up and moving, sneezing all the way so that you can infect other people (translation: help it survive and reproduce). Other diseases such as African sleeping sickness or malaria are virulent since by incapacitating us it increases the probability of its own spread (since the mosquitoes that spread it from one host to another can get to you easier if you are pushed to the brink of death and are unable to move).

So in many ways, the degree of virulence of a disease is determined by the modes of transmission.

Hence, if we were to attack not the disease directly, but the modes of transmissions, we can effectively bribe the infections towards lower virulence.

For example if we spend our medical research money, not on developing antibiotics, but on cleaning water supply or on providing mosquito nets, there is evolutionary pressure on the infectious agents to move towards lower virulence, since for their own spread they need us to be moving around. History also suggests this- when cholera outbreaks spread from Peru to Ecuador (where the water supply was poorly protected), the bacteria became more virulent; when it moved to Chile (where the water supply was much better protected), it became less virulent since it needed people to be alive and mobile, in order for it to spread. (Source: Survival of the Sickest by Sharon Moalem)

Effectively this is a policy of accommodation with the infectious agents- we will not fight you, we will cooperate with you to help you spread, but you need to make your effect on us more benign! By bribing them thus, we may have a far better chance at survival.

I was thinking about this also when I was watching the movie "Contagion". Communicable disease control is really hard- the movie brought out the point not only of the large scale destruction that a single "Black swan" (low probability, high impact) event can bring about, but also that even after a "cure" is discovered, the time to market is so high due to the stringent requirements on testing by the FDA, as also the scale of production and distribution required.

We need multiple approaches to deal with this problem, and the wisdom from the Arthashastra of how deal with adversaries may have useful pointers on what kind of approach we may consider!