Defence land: A key dimension of India's privatisation problem

As Vijay Kelkar has long emphasised, India's privatisation question should be viewed as a question about the portfolio of the State. For each Rs.10,000 crore of shares of Air India that the government owns, it is forgoing 2,000 kilometres of highways. The State needs to ask itself whether it is better to own 2,000 kilometres of highways as opposed to owning the same shares of Air India. On this subject, also see Section 4.3 of this paper.

The second key dimension that should shape the discussion on privatisation is that of improving GDP growth. When assets are moved from public control to private control, the translation of capital stock and labour into GDP growth generally becomes more effective. Through this, India would reap GDP growth by better utilisation of existing resources.

Both these issues have, so far, been largely seen questions about the control of public sector undertakings (PSUs). But both issues are broader: they are about asset ownership by the government more broadly. Given India's socialist background, government has frequently and wantonly grabbed assets, far beyond those required for the production of public goods. Hence, the problem of selling off assets is much bigger than just PSUs.

As an example, this article says:

The defence ministry is the largest state landowner, holding 80 percent of the 7,000 square kilometres of government land, much of it now prime real estate, according to the CAG report released Friday.

Here is the CAG report referenced there.

There are two interesting dimensions to the problems of defence land. First, while the Ministry of Defence undoubtedly needs large tracts of land on which it can run exercises, training, experiments, etc., it certainly does not require prime land in cities.

The heart and soul of a city is the dense interactions between top decile people. What makes a great city is greater interactions within a greater density of higher talent people. By definition, military installations are aloof from the general population; they do not interact with the main citizenry. Hence, large military installations are like black holes in a city: they increase distances for everyone else, and they stand aloof. The presence of big defence lands in cities is not just about wasting fiscal resources (you'd be better off selling that land and retiring public debt) but primarily about increasing the quality of the cities.

There is a case for placing defence research labs in other innovation hubs of India - e.g. Bangalore, Poona or Bombay. This is justified because they would serve to increase the density of scientists thus enhancing the quality of these cities, and because these defence labs would benefit from interactions with civilian scientists. Barring research labs, there is no case for placing defence facilities in cities.

There may be a role for one big HQ in New Delhi, but I don't see why dozens or other defence installations are required to be in Delhi. The governments stands to raise trillions of rupees by selling this land and shifting these organisations to locations in the boondocks, where land is roughly free. And there is a further kicker: When defence holdings in places like New Delhi or Poona are moved off into private ownership, India's GDP will go up. So this is a win-win at two levels: First, India's fiscal problem is eased by selling defence land and writing down debt, and India's GDP is increased because the land gets put to productive use.

Similarly, I don't see why anything connected with the Indian Navy needs to be in Bombay. It's perfectly feasible to create naval bases at boondocks locations on the coast and thus free up the space used in high marginal product land.

The second interesting dimension is that of the Ministry of Defence as a creator of new cities. If you start off with land in the boondocks, on day one, nobody wants to go there. The Ministry of Defence has the ability to solve the coordination problem. It can engineer the synchronised movement of a large number of distinct pieces that are required to create a new cantonment town. Once this has been put into motion, within 20 years or so after starting up, it would be wise for the government to sell this off and start over. For MoD, there is little difference between being in a mature cantonment town versus a brand-new one. But for the exchequer, enormous value is created through this process. And for India, this works well because new cities can be steadily created in this fashion.