Telecom auctions have failed India

https://www.bloombergquint.com/global-economics/economics-nobel-prize-celebrates-auctions-that-failed-india

84 89 6 years ago

teej• 6 years ago

The book Discovering Prices by this year’s Nobel laureate is really fascinating. The very first chapter opens up to explain the boundless political, mathematical, and computational complexity involved with the FCC Broadcast Incentive Auction. I recommend it just for that alone.

samsonradu• 6 years ago

Here’s another book of his on the topic - Putting Auction Theory to Work

http://www.econ.ucla.edu/riley/271/Milgrom-Putting%20Auction...

jjoonathan3• 6 years ago

How well do the FCC auctions actually work though? If you pop up a spectrum analyzer just about anywhere, you'll see that the spectrum is vacant and deserted except for broadcast radio and the bustling ISM (free-for-all) bands holding approximately all of the traffic. If you have a really good spectrum analyzer, you might be able to capture brief blips of cellular activity, and if you're near a radar you'll see that of course, but by and large the spectrum looks like a ghost town of empty apartments with absentee owners using them as capital assets while the plebeians cram into the designated slums. The stark contrast between the wildly over-utilized free-for-all bands and the wildly under-utilized auctioned bands really looks like a strong argument against spectrum auctions.

I'm sure you could find an economist to argue that the present state of affairs is natural, good, and maximally value producing (producing ROI for investors, that's what life's about!) but from where I'm standing the FCC auctions look like an idiotic forward propagation of last century's engineering limitations, no matter how sophisticated and lucrative the auctions themselves are.

nickff3• 6 years ago

I doubt that many economists would "argue that the present state of affairs is natural, good, and maximally value producing", they'd say it was a legacy of "last century's engineering limitations" which the bureaucracy adopted, and subsequently ossified around.

I think most economists and business people would agree that real-time auctions similar to what happens with internet ads would be much better.

stefan_• 6 years ago

I generally agree, but there is of course a third category of frequency band besides FFA and auctioned, which is various "legacy interests", military and others pretending they need each about a gigahertz worth of prime bandwidth fully alone or the sky will fall.

anthropodie• 6 years ago

It’s even worse when countries try to maximize just one variable, because for bureaucrats and politicians that variable is usually government revenue.

When this happens, all other variables like environment, humanity, morality, future of mankind and the greater good goes down the drain and I am not necessarily talking about India or its telecom sector. No wonder we are making mess of our planet.

amadeuspagel8• 6 years ago

Government revenue is a pretty good variable to maximize in an auction. The government can spend that money on "other variables like environment, humanity, morality, future of mankind and the greater good".

marcosdumay• 6 years ago

It's a problem if it destroys an industry. Or would it be better if the government spent on industry subsides so that the revenue generation action doesn't destroy it?

I don't have any good answer. Auctions work best when the market is predictable, but telecom is about the opposite of predictable. Maybe the government could detach the spectrum from the service somehow and make the spectrum revenue predictable.

I don't see any alternative to an auction that is actually better around here, and don't know of any either.

upofadown• 6 years ago

They haven't worked super well anywhere, have they? The effect where the auction sucks all the money out of the industry occurs everywhere, it is just that in some places there are entrenched monopolies with lots of money willing to pay the kickback for exclusivity.

tryptophan• 6 years ago

Its what the USA uses, and I guess it works fairly well here. I guess India followed the law of how to run these auctions, but did not follow it in spirit, messing things up(I'm guessing politicians fiddled with it) and led to the bad outcome.

toast04• 6 years ago

We've only got 3.5 networks to choose from, competition on price is mediocre. It's more competitive than residential networks in most of the country, but that's not saying much.

bluedevil2k4• 6 years ago

They work extremely well in almost every country in the world. The situation described in the article in India was an unusual circumstance, and obviously the bidders overpaid for the spectrum given that the ultimate customers have very limited financial means. It’s possible this may repeat itself in other similar countries with high population and low income levels.

That definitely doesn’t mean it doesn’t work elsewhere. I’ve personally been involved in Singapore, Belgium, Sweden and Mexico’s spectrum auctions and the results have been very successful. As further proof, if countries felt they weren’t successful we would see the number of auctions decreasing. It’s the exact opposite, the expected number of spectrum auctions in the next 2-3 years is extremely high. It’s a quick way to make some money for governments struggling to raise money right now.

(I provide the software used in many spectrum auctions in the world)

haltingproblem6• 6 years ago

The Tragedy here is that spectrum auctions have been lose-lose-lose.

1. Telecom companies and their shareholders are burdened by debt and either go bankrupt or have horrendous returns.

2. The govt (and taxpayers) do not realize the revenue from spectrum auctions

3. Finally, the subscribers suffer under poor network infrastructure, towers and upgrades as the companies are hamstrung by spectrum debt and unable to raise cash.

The lose-lose-lose Nobel winning work does not stop the columnist Mihir Sharma, who is also an economist from playing fast-n-loose with his triumphant conclusion:

"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."

Economists can never be wrong and even when they are they should be commended for their efforts even though other solutions exists - e.g. revenue sharing - which work very well. Top down market design is the highest virtue of all even though hundreds of millions of very poor people might have suffered for academic theories of a few.

blueblisters2• 6 years ago

Why was this downvoted? Seems like a fairly well-written argument - would love to hear counter-arguments instead of blanket downvoting.

xhkkffbf• 6 years ago

I'll give it a go. It's a nice argument, but it's needlessly negative, almost for show.

Consider the claim that's it's a loss if the company shows poor returns. If it had good returns, the anti-auction folks would claim some small group is getting rich off what used to be a public good. Poor returns for the company means that the auction is limiting the plunder and that sounds like a win to me.

And why is this a loss for the government treasury? If the returns are poor, there wasn't much more juice to squeeze from the orange, right? So it sounds like the government (and often the taxpayers, sort of) are getting as much as they can for very little work.

I'm much more sympathetic to the deeper arguments about structure of auctions and how they might be improved. Why not auction the spectrum off for a shorter time? Maybe use smaller chunks of spectrum? The auctions are just a competitive algorithm and they can become more accurate with finer resolution.

KorematsuFred2• 6 years ago

OP suggested markets work. HN folks downvoted without presenting arguments.

donavanm2• 6 years ago

Its a statement, not an argument, as I read it. Theyre ignoring other confounding factors, the agency of both telecoms and policy makers, and basing the lose-lose-lose on a limited subset of auction markets. Additionally theres no other suggested mechanism besides a vague wave at “revenue sharing” where you have the same problem of determining what the appropriate rate is.

amadeuspagel8• 6 years ago

Top down market design is the highest virtue of all even though hundreds of millions of very poor people might have suffered for academic theories of a few.

Can't stand this snarky, melodramatic rhetoric. I might as well talk about millions of very poor people suffering because they lack clean water and healthcare which the government could have provided with the revenue from spectrum auctions.

haltingproblem6• 6 years ago

Cellphone service has markedly improved the lives of the very poor something most economists did not predict. In fact when India deregulated rates from some of the highest in the world (voice was around $.10 to $.20/min) most economists argued against it saying cellphones were a thing of the rich and the tax-losses to the govt from telecom revenue were not worth it. Deregulation was called a give-away to the telco companies and the rich etc. Instead, what came about was the biggest uptake in cellular and data usage and one of the cheapest rates in the world.

The poor benefit disproportionately in relative terms from access to cheap communications. Studies show cellphone and data usage to be highly correlated to improved markers in health, economic well being, education, etc. Poorly designed top-down auctions bankrupts companies and causes under-investment in infrastructure do directly cause suffering to hundreds of millions of very poor people. Maybe you find it snarky to point that out but I see it as a tragedy.

True, you can also argue that the auction revenue would have helped clean water and other developmental activities but that is an indirect effect.

Spooky232• 6 years ago

Give a man aid and he's fed for today. Give him a cellphone and he has a job.

Economists peddle wild ass guesses as science. Missing the boat on cellular is par for the course.

hcknwscommenter5• 6 years ago

1. Many telecom companies in other countries that use auctions have pretty decent returns. See, e.g., ATT and VZ among others.

other solutions exists - e.g. revenue sharing - which work very well

The terms of the revenue sharing need to be worked out somehow. That "somehow" is typically via some sort of competition for scarce resources (spectrum), e.g., an auction. Sure, the gov't could just set a price, but the inefficiency and lack of price discovery inherent to that approach is precisely the issue the auction seeks to address.

bluedevil2k4• 6 years ago

Your claims are unfounded though. The example given in the article, India, is an anomaly. Most countries are very happy with auctions and the results, both financially and coverage-wise. If they weren’t happy, why would they all keep running them? There are dozens of spectrum auctions planned over the next few years around the world, because every country realizes it’s the best way to raise money, the fairest way to distribute spectrum (ie they won’t get sued), and the best for consumers.

Now, is every country perfectly set up for a successful auction? No - but in those cases you’ll see special rules meant to promote competition in the market. New entrants may receive discounts, or certain blocks of bands may be set aside for new entrants only. Big companies may be capped in how much additional spectrum they can purchase is another common case. These auctions are almost never straight up “take what you can get”. You can imagine in the US if that was the case then AT&T and Verizon would own all of it.

My company writes the software for these spectrum auctions, and I’ve personally been a part of dozens of them.

tpetry• 6 years ago

He is correct. In Germany the auctions for umts licenses costed 50 billion euros. These insanely high auctions (because every telco believed they need these frequencies and outbid everyone) has negative price impacts today, something which happened 20 years ago! Other countries had better outcomes, but you cant foresee how such an auction will go and which impact it will have.

surajs• 6 years ago

ironically, an industry still using outdated modes of revenue measurement can't be said to be thinking "bigger" or better.

donavanm2• 6 years ago

Without disagreeing with the lose-lose-lose premise how do you draw the blame to the auction mechanism? The telecoms are in control of their bid values. The government policy makers are in control of minimum bids, related contingencies, and how the proceeds are spent. And yes, consumers are saddled with the confluence of the two.

With those controlling factors or problems Im not clear how a different mechanism, say fixed rates and allocations, is above the same failings?

amadeuspagel8• 6 years ago

But the more a company pays for spectrum, the lower its profits and the less it has left to invest in new infrastructure.

Strange argument. Does india not have banks? A bond market? If that new infrastructure is going to pay for itself, it should be possible for companies to raise money to build it. The government shouldn't have to give resources away, so that companies have more money for investments.

robocat2• 6 years ago

The very next sentence after your quote is “In India, high fees have led to high levels of debt” which somewhat answers why more debt might not be the answer.

amadeuspagel8• 6 years ago

No it doesn't. So it has led to high levels debt. So what? What's the problem with debt?

blueblisters2• 6 years ago

It's very hard to forecast revenues from emerging markets - bondholders likely won't be able to stomach the risk of large infrastructure investments without any idea if/when they can breakeven.

amadeuspagel8• 6 years ago

It's their job to figure out what risks to take, how much interest to charge, how to hedge.

robocat2• 6 years ago

“The country’s telecom revolution only took off after the government moved away from auctions and started assigning spectrum to licensees in return for a share of their revenue. The new system brought in twice as much in fees as the auction bids would have.” [shortened].

It seems such a slippery slope to tax infrastructure: although I guess it has the advantage that tax avoidance is difficult!

haltingproblem6• 6 years ago

The companies still found ways to avoid taxes, but I guess that is a tautology ;) Say you are selling a wide menu of choices like an enterprise offering or better a restaurant. You can offer that appetizers are free as long as each individual orders at least two drinks. Make your appetizers salty and fried and it creates a (un)virtuous cycle of ordering more drinks.

Companies tried this exact strategy in India. They made the bandwidth cheap and moved revenue to ancillary services - ringtones, song libraries, ... Telecom authority ruled that all revenue falls under revenue sharing and the supreme court finally ruled against the telcos after a long drawn out battle.

hcknwscommenter5• 6 years ago

"The new system brought in twice as much in fees as the auction bids would have.”

If the new system brought in twice as much in fees, then by definition it was more burdensome on the telcos no?

colinmhayes2• 6 years ago

The article claimed that the initial debt from the auction was more burdensome because telcos rely on a critical mass of infrastructure that they weren't able to create due to debt.

avmich• 6 years ago

Economists like to think of themselves as mathematicians — or, if feeling momentarily humbler, as physicists.

What, those proud masters of the universe, physicists, are humbler that out-of-their-mind, impractical looneys, mathematicians?

tsjq3• 6 years ago

Not to forget the huge scam how Reliance Jio got their spectrum.

https://archive.is/Ggvmi

TLDR : that 4G spectrum auction was for Data-only. Reliance bid through a benami company, won all zones across the country, and then got the rules modified to data+voice.

triceratops• 6 years ago

What's a "benami company"?

tibbydudeza• 6 years ago

A front company.

rational_indian• 6 years ago

shadow company. Benami = nameless = anonymous.

perryizgr84• 6 years ago

I'm happy Jio did it. They singlehandedly brought about the digital transformation in India. It's truly amazing.

tsjq3• 6 years ago

You're happy they did a scam , and nearly bankrupted other telcos (all their NPA will come back biting us in the form of taxes for recapitalisation )

c06n• 6 years ago

It has worked very badly in Germany as well. In the year 1999, the UMTS licenses were auctioned based on Milgrom and Wilsom's ideas. The government's revenue was awesome, but it came at a high price. The telcos had to take up large credits, hindering future investments in their their mobile networks. Today, Germany has one of the worst mobile networks in the western countries.

article in German: https://www.golem.de/news/mobilfunk-umts-versteigerungstakti...

nashashmi• 6 years ago

Well said. Blame the act of crediting being used for auctions. It hurts the losers, winners, and users. Only govt (sort of) wins.

hcknwscommenter5• 6 years ago

I mean, maybe? It is a huge stretch to tie a specific event (auction and the related debt) to the ensuing 20 years worth of telecom infrastructure investment. This implies (incorrectly) that there are no other significant reasons Germany might lag in mobile network investment.

watt• 6 years ago

It fits, though. It's like the opposite of subsidizing an industry: saddling it with a crippling debt instead. And yes, Germany has awful telecom infrastructure, and lags behind the likes of Spain or Czechia horribly.

All because telcos here need to recoup the initial "investment" and are not getting to the newest tech while everybody else is.

KorematsuFred2• 6 years ago

For anyone interested in this topic (and Indian government in general) I highly recommend Dr. Arun Shourie's book "Governance and the sclerosis that has set in". It gives a detailed look at the telecom mess and the kind of mentality that causes it.

Dr. Shourie was a cabinet minister in early 2001 and was single handedly responsible for Telecom reforms that made even these auctions possible. He wrote the entire book on this experience.

Few points:

1. India had several telecom "circles" based on the political borders of the states and handful of players (I think 6 or so). All 6 of them had lawsuites against each other and with government so around 50+ lawsuites over petty things. Dr. Shourie worked hard to clear every single of those lawsuites without having to wait for the courts to resolve it.

2. Indian regulations requires the cell towers to not extend their range beyond the political boundary. People who wrote the rules refused to acknowledge the fact that you can't stop radio waves based on political borders.

3. You had to apply different licenses based on how you will use the spectrum. For example for voice you had to apply one license and for SMS another license, for data another license. Also, a separate license for inter-circle communication. Which means if you want to allow your Mumbai customers to be able to send SMS to Delhi customers you required a separate license for the two circles and separate license for Mumbai to Bangalore communication. Also, separate licenses for ISD, international communication.

4. Indian government also insisted that they get to control how telecom company will roll out their towers. So if you erect a tower in Mumbai you must also erect a tower in some village even if you don't have any customers there. If you fail to do so, you will be fined.

5. Indian government also wanted (and actually did) to ban services like Skype because they would enable voice over data thus making a separate voice license pointless. For many years VoIP was illegal in India.

The body responsible for these hair-brained regulations was a technical body called DoT.

Not only Dr. Shourie worked hard to resolve most of these regulations and replace them with something sensible, he insisted that if telecom companies are not interested in even bidding for some circles we should give them the spectrum for free. No telecom company was even interested in applying for spectrum in many parts of India (something unthinkable today) which were then given on first come first serve basis.

Later when telecom revolution happened and spectrum value exploded through the roof, some people wanted to investigate Shourie for corruption for giving this obviously valuable spectrum for free. Even the supreme court of India had trouble understanding why the spectrum that sells for billions today was not worth anything 20 years ago.

Having known Dr. Shourie personally, he is one of the smartest and competent individual Indian government was thankful to have working for them.

amadeuspagel8• 6 years ago

There's a general sense in the article and the comments here that how much a company invests depends on how much money they have. This is wrong. If there's a profitable investment to make, the company is going to make it, if they don't have the money, they're going to loan it. If there's no profitable investment to make, the company isn't going to make any investments, if they have money sitting around, they'll return it to their shareholders. Yes, there are some frictions involved here. If a company has some crazy genius idea, it might not be that simple to loan money. But for infrastructure investments, it should be.

gruez• 6 years ago

If spectrum auctions are a bad way to allocate the radio spectrum, what are better ways? Wikipedia says:

Alternatives to auctions include administrative licensing,[clarification needed] such as the comparative hearings conducted historically (sometimes referred to as "beauty contests"),[citation needed] or lotteries.

Those seem worse than auctions. They seem very prone to corruption, and everyone ends up having to spend money on lobbying, essentially turning it into an auction[1] in disguise.

[1] https://en.wikipedia.org/wiki/All-pay_auction

bluedevil2k4• 6 years ago

Not lobbying, you’re thinking too Western. It’s given to family or family friends.

ohazi• 6 years ago

Monopoly rights to spectrum should have ended once cellular modulation schemes went digital.

The modulation techniques used in 3G/4G/5G systems are already agile enough that they really wouldn't have a problem with all the carriers sharing all the spectrum all the time. Carriers and other license-holders would have to do a token amount of coordination among themselves, probably similar to the settlement-free peering arrangements that the tier 1 Internet networks use.

In return, we would all get more efficient real-time spectrum allocation, and carriers could instead spend their billions differentiating themselves with infrastructure and capacity improvements rather than locking out anyone who didn't spend billions on spectrum 10-20 years ago.

Sniffnoy• 6 years ago

This article doesn't really explain why these mechanisms would work in one place but not another. What assumption was violated / was different, for instance?

LatteLazy• 6 years ago

The article seems to think that if the proceeds of an auction are misused, that means the auction is bad.

Then it goes on to nitpick: auctions that maximise revenue discourage investment, auctions that don't maximise revenue waste money. So all spectrum allocation is damned?

I'd be happy enough to agree that auction design is hard (and I think really interesting) and that combining it with politics is also difficult. The issue here (which the author never states but I think would not be disagreeable to them) is that you have to be careful what you wish for: maximising revenue then misusing revenue won't help. But that's not the aauxtion/economists fault any more than you're mechanic is responsible for you intentionally crashing.

dredmorbius• 6 years ago

Paywall: https://archive.fo/C7JDc

known2• 6 years ago

This Nobel-Winning Idea Has Failed India https://archive.is/C7JDc due to Cartels aka Casteism in India https://archive.vn/VWFSC

Which Caste is looting India ? https://archive.is/5zz1b

known2• 6 years ago

India is unique https://archive.vn/pOmij

It has Producers, Consumers and Leeches https://twitter.com/0x101/status/1317685559414972418

marcinzm2• 6 years ago

I wonder if it's the case that the first time you do any kind of novel distribution approach (auction, revenue share, etc.) at large scale it has a good chance of working. The second time, however, everyone has had time to learn how to game the system (including the government) no matter what the system actually is. Then you can play a game of whack a mole to patch things but you rarely come out ahead.

Telecom auctions have failed India

Related Stories