By Rajesh Chakrabarti, Executive Director, Bharti Institute of Public Policy, ISB

This article was first published in the Economic Times on Oct 31, 2013

The current controversy over the coal block allotment is perhaps the latest and one of the most sordid examples of inefficient allocation of national resources, but the malaise is hardly new.

In fact, in an interview back in 2010, Raghuram Rajan had identified favoured allocation of resources as the driving force behind the fortunes of many billionaires in India – a nation that ranked second only to Russia in terms of the number of billionaires per trillion dollar of GDP.

How then to best allocate national resources – mining rights and telecommunication spectrum – in an efficient manner? Auctions come to mind immediately as competitive and transparent mechanisms of allocating resources. They allocate resources to the highest valuer, ensuring both socially most productive allocation as well as maximum revenue gathering by the exchequer.

Still a Surprise Element

But governments around the world have long used auctions to allocate resources and yet the mechanism has at times thrown nasty surprises. Auctions can be of several types and the success of the mechanism depends critically on its design.

Take, for instance, the 3G spectrum auction in six European countries in 2000. While in UK and Germany, the auction yielded prices over 600 per person, in Switzerland, the yield came to only 20, about 5% of the predicted value. In 1999, in Germany, as spectrum was being auctioned in 10 blocks, Mannesman and T-Mobile colluded to split the spectrum between themselves, each picking up half the offering at a very low price.

A successful allocation mechanism should ensure widespread participation, a prerequisite for competition. It should also prevent collusion so that each bidder reveals close to its true reserve price for the item or items being auctioned.

Sealed with a Dutch

Broadly, auctions follow two formats: ascending, or “Anglo”, approach where an auctioneer keeps raising prices till there is only one bidder left (there is a descending version as well); and sealed-bid, or “Dutch”, auction where bids are simultaneously revealed. The former has the advantage of transparency, price discovery and revenue maximisation, but, in the presence of one or more “strong” players, may end up discouraging participation by smaller players, in the absence of which the supposed “stronger” player can get away with offering significantly less than what it truly values the resource at.

A mix of the two, “Anglo-Dutch”, is sometimes resorted to that runs the “Anglo” way till only two bidders are left in the fray, and then switches to “Dutch” with the floor of the last price of the Anglo auction.

Another deterrence for auctions is the infamous “winner’s curse”, referring to the fact that the actual winner is likely to be the party that has overvalued the resource the most. The issue is particularly significant when the resource being auctioned, like mining rights, whether land or offshore, is difficult to value. This is where valuation becomes critical.

“Common-value” resources like oil, that are valued pretty much the same way per unit by all parties, but whose supply per block is something each bidder is uncertain about, differ from “private-value” items to which bidders assign different values irrespective of the valuation given by others. The “winner’s curse” problem is strongest in the former.

Tailor-Made for Scam

A standard solution to this problem is the “second-price” option where the highest bidder gets the resource but pays the next highest bidder’s bid. A common result: less hesitation to bid on the higher side. But even this can throw uncomfortable surprises. Consider the embarrassment of the New Zealand auctioneer – and the immediate suspicion of foul play had it been in a country like India – when a winner bidding NZ$7 million walked away paying the next highest price of NZ$5,000! Clearly, the presence of a reserve price is essential to avoid such situations.

While under certain, rather unrealistic assumptions, all major auction designs would yield the same results, in reality, the suitability of the design to the resource being auctioned, the nature of information about it and the nature of competition among bidders hold the key to the success of the auction process.

Then, one has to pay attention to other auctions as well. Turkey auctioned two mobile telephone licences in 2000; sequentially, with the condition that the price of the second auction could not fall short of the first. The result? The winner of the first auction bid (and paid) a price so high that it could never be worth with two rivals in competition. The second licence, naturally, went unclaimed!

Rules and punishments are not beyond question either. The design issues apply not just to minerals – coal or offshore oil, for instance – but also to PPP projects where we have often seen re-negotiations and cost escalations. Auctions are better than the first-come-first-served or government-determined allocation, but their outcomes depend critically on the appropriateness of their design.