I had recently had the chance of researching the subject of corruption with my former colleague Ajay Subramanian*. The questions to ask included whether all corruption is the samethe everyday kind that we face from traffic policemen, road inspectors, tax refund clerks, railway officials or the grand larceny of the kind that the likes of Raja and Kalmadi and former senior judges are now accused of Which one is worse for the nation, and therefore a bigger danger And finally, what circumstances bring about the best outcome Is there a point where the costs of reducing corruption further are more than its benefits Given that it is unlikely that we can ever build a nation with zero corruption, how much of our resources spent in detection would lead us to the optimum level that maximises the nations welfare
In our view of the world, the entire national income-corruption tradeoff is driven by the alignment between three distributionsof productivity, power and compensation. Let me explain. Not all jobs are equally productivethe farmer directly produces output, while the policeman does not. But that does not mean the farmer should get all that he produces, for then there will be nothing for the policeman, and crime will go through the roof. So, clearly, the farmer must support the policeman in his own interest. So, productivity and compensation structures will be different, taxes are necessary. The nature of the job matters too. The schoolteacher has little power to extract rent (particularly if the examiner is external and cannot control admission) that the policeman enjoys. So, professions and hierarchies create a distribution of power in society as well. So it is possible to think of the entire society in a three-dimensional spaceof productivity, power and compensationwhere each individual chooses his level of individual corruption (between 0 and 1) depending upon the risk-return tradeoff he faces. The higher the chosen level, the lower his output and greater his actual compensation if he goes undetected. Of course, society suffers more from a corrupt PM than a corrupt constable, so societal corruption level has to be power-weighted. Societal output is the sum of all individual outputs. We simulate various such 3-D societies to investigate the income-corruption relationship in them.
Our findings are a mix of intuitive and surprising results. For instance, the more aligned compensation is to power, the better it is from a corruption-reducing viewpoint. For instance, when our top bureaucrats making less than R9 lakh a year (OK throw in the Central Delhi flat and a car to be more precise) hold the key to cash-flows amounting to thousands of crores of rupees, bureaucratic corruption should not surprise anyone. Ditto for judges (just talking
about incentives here, not alleging any judicial wrongdoing). So the Singapore model of paying CEO salaries to ministers has its merits. Not surprisingafter all, despite its high-corruption neighbourhood, Singapore is among the five least corrupt nations (and the only one outside Scandinavia).
Somewhat surprising is the trade-off between petty corruption (of the everyday variety) and high level corruption that institutions like the CVC, CBI, CAG and Lokpal are supposed to catch. Our lowest corruption societies are marked by less high level corruption than petty corruption. So it appears that going after the big fish does produce better bang for buck for reducing corruption than disinfecting the microbes. So the Lokpal debate is actually keythe Anna brigade is certainly not barking up the wrong tree.
It is important to realise that corruption is not the monopoly of the government. Countries like Canada and Finland have large public sectors and low corruption whereas unfettered government monopoly over resources has created kleptocracies in sub-Saharan Africa with active private sector involvement. Our model suggests that scaling back government power reduces petty corruption while reducing private sector power leads to a reduction in high level corruption. Little surprise, then, that liberalisation has perhaps taken us from one kind of corruption to the other, potentially more harmful, kind.
The author teaches finance at the
Indian School of Business, Hyderabad
* Chakrabarti, Rajesh and Ajay Subramanian, Power, Compensation and Corruption: Theory and Evidence Working Paper, SSRN