Cross-subsidizing inefficiencies Vs direct subsidy transfers

Livemint reports that the Indian Railways is considering purchasing stressed power assets since they are an attractive opportunity to buy power assets at a cheap price. The argument for such purchases is that it would kill two birds with one stone - provide captive power assets for Railways at a cheap price and thereby de-stress the asset and help banks eliminate their non-performing loans. But does the cheap price alone make the asset attractive?

Consider this. The vast majority of power generation projects currently distressed are so because either their tariff is prohibitive or because they are unable to access assured fuel supplies. The former can happen because of high construction cost, high fuel price, or inordinate delay which cause cost over-run and accumulation of interest during construction. Since the contributory factors are not easily mitigated, if at all, the majority of these assets are most likely to remain high-cost generators. The lure of the cheap price would generally be more than off-set by the high life-cycle cost and sunk-cost effects (such projects generally require further large capital investments). After all, if the assets were really that cheap, wouldn't they be equally attractive to those with far bigger pockets, the infrastructure funds and other asset managers, leave aside other competing power generators. 

This brings us to the issue of efficiency and public policy. Governments are attracted by the prospect of such apparent free-lunches that typically ends up cross-subsidizing inefficient public entities. They are everywhere - mandating that government officials travel by Air India, forcing LIC to purchase shares during disinvestment, regulating that power generators can buy coal only from state-owned coal miners etc.

The Railways would surely have to incur higher transaction costs in running power stations and pay more (than if it were purchased from the market) to purchase that power, apart from suffering greater unionization and resultant politicization. The former would crowd-out resources available for its core activity, while the latter would exacerbate the bureaucratic inefficiencies of the organization. The cumulative costs (which are effectively another form of subsidy) of all these are most likely to be far more than the simple cross-subsidy (either as a tariff or investment subsidy) that would have enabled the project to begin generation. All this would force an otherwise efficient entity to now bear the cost of some one else's failures, with attendant knock-on effects on its operations.   

This raises the question whether governments should abjure from such stealth cross-subsidy to prop-up demand or keep running inefficient entities, with its numerous distortions, and instead directly finance them through its budget resources. Reinforcing the second welfare theorem, conditional on the political economy constraints that necessitate such choices, direct budgetary support is a cleaner and efficient approach.