Brazil announces new spectrum pricing rules. Why not throw them out completely?

Posted on Tuesday, July 24, 2018

Today the Brazilian regulator, ANATEL, announced a new spectrum pricing framework. A general article (in Portuguese) can be found here. The regulatory document is here.

According to the news article – no I am not going to read the actual document, sorry – the formula tries to capture the value that an operator will derive from the spectrum so it takes into account the location, population, the service for which it will be used etc. All good economic theory to set prices based on underlying value and thus avoid an inappropriate, uneconomic allocation of resources.

Except I disagree with spectrum pricing, period.

Neither do I believe it should be free. That would encourage the accumulation of unused or underutilized spectrum which would indeed be a waste of a country’s scarce resources.

If that sounds like a contradiction, I hope to show that it is not.

Spectrum pricing came about initially for good reasons: to prevent frivolous requests of a scarce commodity; to prevent spectrum being acquired by those unable to develop it or for merely speculative purposes (as often happened early in the history of US spectrum auctions).

Instead spectrum was made expensive so that only well-financed and so presumably serious actors would be awarded it.

Except that Ministers of Finance saw a windfall opportunity for their challenged public coffers. Astronomical figures paid by deep-pocketed American carriers and (to a lesser extent) European ones, only reinforced the notion that monetizing spectrum was a way to fund more stuff for voters without having to annoy them by raising taxes. Ministries of Communications became profit centers and some of the largesse was able to stick to ministerial budgets, vastly improving the staff’s ability to attend conferences or have better laptops and maybe permitting the Minister the occasional upgrade to Business Class.

But this tax was a transfer from the shareholders of the operator to the taxpayers of the country. Maybe not an inappropriate tax to be levied, but it was money that, frankly, did nothing to enhance the experience of users; money that would be better spent by investing it in the network.

Thus, my modest proposal.

It starts by presuming that the purpose of a Ministry of Communications is not to make money but to ensure continuous upgrades of the country’s communications infrastructure; to ensure the country is competitive technologically.

If a company spends a ton of money on spectrum that necessarily means less for investing in infrastructure. High spectrum fees mean slower deployment of coverage, slower increases in capacity, slower achievement of the expected benefits from the spectrum being auctioned.

To me high spectrum prices are counterproductive for everyone except perhaps the Minister of Finance.

And low spectrum prices do not fulfill the economic purpose of preventing uneconomic purchases by either incumbents or new entrants.

So, the answer is to do away with them completely.

Instead, those wanting to get spectrum should propose more aggressive deployments than those contemplated by the Ministry’s minimum deployment plan: more coverage, more secondary and tertiary cities, faster download speeds etc. (This technique is often used in auctioning oil and gas exploration blocks. The main criteria for deciding who wins is the company who makes more above-and-beyond commitments than its competitors.)

Thus, the money would have been spent (needlessly) on spectrum would instead be invested in the country’s infrastructure, which (as above), I assume is the fundamental reason the spectrum is to be auctioned in the first place.

Certainly, the Minister of Finance will be very unhappy with this arrangement as will those who think the country should get something material out of private investors use of a public good.

My solution for that admittedly legitimate concern is to raise revenue-related royalties or create a special income-related tax on telecom companies. That way operators have the upfront money they need to deploy infrastructure quickly but they share the economic benefits of the industry’s success with the spectrum’s underlying owner.

This is more complicated without doubt. There has to be an objective way to compare proposals that compete on different dimensions, for example, one that proposes faster deployment in major cities versus one that expands the number of cities to be covered within the initial deployment. That would not be easy, but I know a consultant that could be hired to help.

A more serious charge is that such a system favors incumbents. With an existing network, it is much easier and cheaper to propose accelerated deployment.

But let’s be frank: how many new entrants have we seen in Latin America – or anywhere else for that matter – lately? And what has their performance been? How many still survive as standalone players?

Governments always make great promises about attracting new investors to a spectrum auction but it rarely if ever happens even when new entrants get artificially low spectrum prices or reserved bands. All of the companies that I can remember that recently received this benefit are now part of AT&T – hardly a ‘charity case’.

Instead spectrum auctions – at least successful ones – are a case of ‘round up the usual suspects’: the same players as always.

My argument is reinforced by the current problems with Colombian spectrum auctions. These keep being delayed because the current operators say they have no money. The Colombian regulator has been diligently destroying operator profitability (i.e. increased transfers from shareholders to consumers) and the outrageous license renewal fee charged to America Movil and Telefonica reduced these players’ cash flow considerably. How can the Colombian government expect them to pay inflated prices for spectrum after having charged them US$4B for assets they already own? (Don’t ask!)

Finally, if Latin American governments want to see 5G faster than some of us are predicting (later in the next decade) then they need to seriously rethink their spectrum pricing strategies. The 5G business case in Europe is dodgy. The US has decided to ‘build it and they will come’ – not always the most successful strategy. With Latin American incomes and Latin American ARPUs, operators will face significant resistance from their Boards of Directors when it comes to 5G.

This task would be made easier if spectrum was cheaper.

Free would be best.


No Comments »

Leave a Reply