Logo


Telecom spectrum auction: How this year's Nobel-winning idea failed India

The bridge that Milgrom and Wilson built was the 1994 auction of telecommunications spectrum by the Federal Communications Commission

Idea buys Videocon's spectrum for Rs 3,310 cr
The more a company pays for spectrum, the lower its profits and the less it has left to invest in new infrastructure.

Economists like to think of themselves as mathematicians — or, if feeling momentarily humbler, as physicists. This year’s winners of the Nobel Prize in economics, however, seem to conceive of themselves more as engineers. Like Alvin Roth and Lloyd Shapley, who won the prize in 2012, Paul Milgrom and Robert B. Wilson are specialists in “market design,” a field which, as Roth wrote in a famous paper, calls for “an engineering approach.”

Roth argued that market designers should take bridge-builders as their exemplars: “Engineering is often less elegant than the simple underlying physics, but it allows bridges designed on the same basic model to be built longer and stronger over time, as the complexities and how to deal with them become better understood.”

The bridge that Milgrom and Wilson built was the 1994 auction of telecommunications spectrum by the Federal Communications Commission, later called “the greatest auction in history.” In its 50th anniversary volume, the National Science Foundation used the $7 billion in revenue the auction generated as a justification for its years of support of game theorists. Since then, auctions have become the gold standard for the distribution of all sorts of natural resources, from exploration permits to mining leases to railway franchises. It is almost taken for granted that, if properly designed, auctions will find the ideal balance between efficiency and revenue generation.

But, it turns out, economists don’t actually work like engineers. The example set by the FCC auction has in many ways turned out to be very unhelpful, particularly in emerging markets.

There are lots of reasons for this. One is built into the notion of auction design itself. In much of the early academic literature on auctions, economists tried to maximize the sum of state revenue and consumer surplus. But setting goals is the task of regulators and politicians, not economists.

Policy makers may have multiple other considerations, complicating market design considerably. South Africa had three failed auctions between 2010 and 2017, for instance, because regulators sought to design them to further the government’s broader economic inclusion agenda. To general amusement, the government announced last month that it was trying again.

It’s even worse when countries try to maximize just one variable, because for bureaucrats and politicians that variable is usually government revenue. Economists generally don’t object because revenue is easier to measure than consumer utility, making their job simpler. In India, for example, the government has grown addicted to using telecom spectrum revenue to help finance its deficits.

But the more a company pays for spectrum, the lower its profits and the less it has left to invest in new infrastructure. In India, high fees have led to high levels of debt. Constant demands from the government for cash have caused telecom providers to look for the exit. And spectrum scarcity leads to low quality service: In 2019, India had 50 times as many subscribers per MHz of spectrum as did Germany.

Indians at least should have known this would happen. The country’s telecom revolution — which drove its years of high growth in the 2000s — only took off after the government moved away from auctions and started assigning spectrum to licensees in return for a share of their revenue.

Worse, for those who imagine that auctions designed to maximize government revenue would at least maximize government revenue: The new system brought in twice as much in fees as the auction bids would have.

And suppose that, in order to ensure government revenue was robust, auctions set their reserve price too high, as has happened in Ghana and Bangladesh? Would consumers be served well by a poor-quality oligopoly or even a monopoly? Would that benefit the broader economy? Can Milgrom, Wilson or their successors design an auction model that takes into account the effect on overall economic growth of vibrant and cutting-edge telecom infrastructure? Maybe they can. All I can say for certain is nobody has.

That doesn’t mean they should not try. This is a well-deserved Nobel prize not because the bridge these economists built can be erected over any kind of river, but because it got built in the first place. The true miracle was the transparent, inclusive, academically informed process that led to its design. It ensured that the specific complexities of that market at that time were clearly reflected in the eventual auction design.

Market designers can’t give up because they haven’t been able to replicate that Nobel-worthy success. They need to work harder and think bigger.