The trillion-dollar question

Written by Nirvikar Singh | Updated: Dec 19 2010, 06:17am hrs
At the beginning of 2010, management guru CK Prahalad gave a figure of Rs 2,50,000 crore for the annual cost of corruption in India. He got to this number by adding up the costs of elections and assuming a 10X return on those risky investments. That sounds like a huge number, so let me reduce it by converting it to US dollars (after all, still the worlds reserve currency). That becomes about $55 billion. That still seems like a lot an annual cost that exceeds Mukesh Ambanis known wealth. But lets put it into perspective. Indias GDP is about $1.2 trillion. So Prahalads estimate is only 5% of GDP. That does not sound too bad.

Unfortunately, the Prahalad number does not get to the heart of the matter. Bribes are transfer payments. If that money gets recycled into productive uses, then the cost of corruption would be lower than the transfer payments. The industrialist bribes the politician, who also becomes an industrialist by investing the bribe (almost certainly through some relative or other front, rather than directly). As long as the money gets used as productively as possible, the main impact is on distribution some people get rich by illegal means, others by legal means, and perhaps some talented, but honest people do not get as rich as they might have.

But this is just the beginning of the problem. An income and wealth distribution based on corruption does not encourage the kind of effort that builds real wealth. Instead, effort goes into bribing and otherwise influencing corrupt decision makers. Effort and personal investment decisions are distorted. But it gets even worse. Suppose that corruption leads to inefficient allocations of rights (perhaps to land, minerals, spectrum or other assets). The ones who get the rights do so because they are most corrupt, or most able to bribe, not because they will be the most efficient users of those assets. Resources are misused, and output is lower than it could and should be. If the distortions are just at the margin, perhaps the cost will still only be a fraction of the transfers. But suppose that some things do not get produced at all as a result, or that innovation is stifled. Then the cost of corruption can be a multiple of the transfer payments. Maybe the cost of corruption is ten times the transfer payments estimated by Prahalad, or $550 billion. That seems very high, approaching 50% of GDP. But even if India were 50% wealthier through eradicating corruption, it would be only a third as rich as China. So maybe this is not an unrealistic figure.

It could be that the cost is even higher. This years Union Budget includes Plan expenditure of Rs 3,73,000 crore, or a little over $80 billion. Plan expenditure goes for all kinds of things. Some expenditures that are good for development (including on health and education) are not part of the Plan. And states spend a lot on these things too. Yet much of the money does not reach the intended beneficiaries. Mothers and children continue to die unnecessarily, to be sick more often than they should, and in the case of children, to be malnourished. Some people think that 85% of the money does not reach who its supposed to. But suppose that only half of the government money is wastedmaybe not through corruption alone, but also incompetence. These are not part of the bribes counted by CK Prahalad. But I bet his factor of 10 is a reasonable one to use to calculate lost output. That would be $400 billion of annual income lost because India does not adequately invest in the basic health and education of its people. That seems large, but again, add it to the previous figures and we still have a total less than half of Chinas.

You might argue that investing in people without building the capital they need to be productive will not work. Thats true, but the capital is also not being built up because of corruption. You might argue that there is more to wealth creation than some marginal improvement in existing capital and human capabilities. That is also true. There are also institutions and knowledge. But if India had institutions and a capital stock to match the United States, its GDP might be 20 times what it is now, or even more, not just double, which is about what I get from imagining removing corruption. So my estimate of $1 trillion (rounded up) as the annual cost of corruption shouldnt seem too fanciful. Of course, if corruption were instantly eradicated, GDP would not instantly double. The impact would take time to work its way through the system. But perhaps growth could be 12 or 13% a year rather than 8 or 9%, for quite a few years. That is not fanciful either.

One can also assert some non-monetary costs of corruption. It is unpleasant to pay bribes, even if they do not distort allocation decisions. It is stressful to drive on roads where corrupt drivers have purchased licences from corrupt officials, without learning how to drive or passing a road test. This is so even if ones productivity or lifespan doesnt suffer from the stress. The increased happiness from reducing corruption would also be substantial. Its difficult to put a monetary figure on it, and no countrys GDP measures happiness, but Indians often remark on the joys of being able to do things without being hit up for bribes in other countriesthis is separate from any differences in monetary incomes across countries.

If corruption is so costly, why does India stay so corrupt One could ask the question of other corrupt countries too, but India is very spiritual, has great moral traditions, and so on. In fact, India seems to be getting more corrupt as it grows. I think the latter trend is an illusion. Its just that rapid growth generates greater opportunities for corruption. Look back in history, at the US a hundred or more years ago, and corruption was rampant as it grew into the worlds biggest economy. But it battled corruption, and controlled it. Countries that do not control corruption tend to also become, or stay, very unequal in their distribution of income, and growth tends to fizzle out.

India does have some relatively strong institutions, which might help control corruption. Indias national government has been trying to improve the sharing of the fruits of growth. But it has been rather slack in dealing with corruption, at least till now. Many state governments have been no better. This is natural, since politicians are not angels, nor are they more honest on average than the general population. To control corruption, the government needs to invest in monitoring and enforcement institutions the police and judiciary, auditors, ombudsmen, and so on. It needs to allow itself to be investigated thoroughly when called for, and to practice greater transparency. It needs to keep its hands off the media, which enjoys exposing scandals of all kinds, because that increases media profits. The media is important for letting voters know who is dishonest and should not get their votes. The government also needs to better insulate bureaucrats from political pressures. Paying bureaucrats and politicians high legal salaries, comparable to the private sector, would also help. And the government needs to liberalise further, so that opportunities for corruption that arise from government controls are reduced.

None of this is easy, even if there is a political will to do it. One has to get to a tipping point, where the corrupt politician and bureaucrat become rare enough to stand out, and not be lost in the crowd. So a concerted effort is required, never easy in India. Business people will also have to change. It used to be that doing business in India seemed to require being corrupt. That has changed somewhat, but not enough. The nexus of business people and politicians in some industries remains to be broken. This will require more competition, and therefore a better competition policy that promotes freer entry and growth of productive, innovative businesses, not just successful bribers.

Again, none of this is easy. But it is doable. Other countries have made transitions from endemic corruption to societies where honest hard work pays off reliably. Reducing corruption will not guarantee growth, but it seems to be a necessary condition for sustained growth. Recognising how costly corruption is may be the first step in mobilising solutions to the problem. If the cost of corruption is 20 times what garnered headlines for CK Prahalad, perhaps that will make people sit up and take notice even more than they did a year ago.

The writer is professor of economics, University of California, Santa Cruz