There are many loony numbers being thrown around in support of the great US mission to steal 120 MHz of spectrum away from the broadcast industry and two new reports out this week highlight the lack of thought that some of them are based on. They both focus on the overall effect of the Broadband Plan to make available some 470 MHz of spectrum, but in each instance the largest slice is the 120 MHz that the broadcast industry is supposed to lose voluntarily, but this can only happen if new legislation is passed.
While the Brattle Group is a respected economic group which says that the auction of this spectrum and other identified spectrum would generate government revenues of $64 billion, a figure which we sincerely question, including the $21 billion that it calculates for the TV spectrum. The other figures in a paper entitled Mobile Future from the Analysis Group goes further and says that all of the new contemplated spectrum will create 500,000 jobs – 300,000 in the next five years – with an anticipated 1% increase in internet penetration that this investment creates. The overall FCC broadband plan will also create $75 billion in network investment and about $230 billion in additional GDP in the next five years and another $155 billion in the five after that.
The Analysis Group bases much of the argument on the facts that capital spending in the wireless industry in the US was $185 billion and created 420,000 jobs between 2002 and 2010. We’d like to ask if this report takes account of the complete lack of investment that AT&T has made during the past five years in its own wireless network, only just being addressed, and even then at the lowest commercial level possible?
2002 was very early days for the cellular industry. Today it is the dominant telecommunications force. So if those years were the transition from 2G to 3G, this is the transition from 3G to 4G ‘ which of course will not cost anything like as much, because cellular operators have learned from past upgrades, and because of that equipment suppliers are in tune with making the network equipment fit in more with multiple tiered networks. If you don’t have to install a different base station top to toe for each piece of spectrum, much of the capex goes out of the equation, and what is spent could be going out of the US, to overseas suppliers.
Almost everyone now has access to the internet either at home, at work, at school or in the public library. New services will not come just because 1% more, the poorest 1% in the US, begin using the internet a little more often. Sure we are on a course, ecommerce is still in its early days, but for every dollar that is spent on online services, for the most part this leads to a dollar not spent on retail services. It’s not a net win for the overall economy, but a shuffling of the cards.
The core contention of the Brattle Group’s piece is that spectrum prices have risen since 2006 and will continue to rise at the same rate and that there are more, not less people bidding for spectrum ‘ and it includes all the generic industry players like Google and Microsoft, as well as opportunistic players like Dish Networks, and tier 2 players such as LightSquared.
It also based all of its assumptions about spectrum value on AWS-1 spectrum pricing, which of course has very different physical characteristics to 700 MHz spectrum and uses an average MHz per pop calculation, which could lead to vastly changed build out economics ‘ in 700 MHz the signal can travel for miles and miles, in 2.1 GHz the spectrum travels far less distance and more base stations have to be built. So each piece of spectrum will have a different reach for a given spend, and some spectrum will attract lots of bidders and other spectrum almost none. It says that it has taken this into account in its pricing but we suspect that a complete absence of bidders will occur a) in some spectrum and b) in some regions and c) depending on what spectrum is already sold but not built out in any given region.
It also places a value on Pairing and says that all the FCC spectrum should be paired to get best results, and today that is true. But under LTE, and in particular LTE TDD, this will be unlikely to remain true, depending entirely on who develops which capabilities, such as carrier aggregation, in which chips, and for how much. In the end the advantages of pairing may evaporate during the timescale of this forecast.
The same goes for Wireless Broadband revenues ‘ they are expected to go up and up ‘ which is what spurs the auction values ‘ Credit Suisse says that revenues for wireless services in North America will increase to $216 billion in 2015, up from $178 billion in 2010, growing at 6% in 2011 and 2012, 3% in 2013 and
2014, and 2% in 2015 through to 2020. This is even if ARPU falls per subscriber a little. But of course those sums don’t add up if ARPU turns a corner and falls more rapidly.
Why would that happen? Well it might happen purely if this IS a more competitive market. So the forecast is not taking into account the forecast itself. So on the one hand more suppliers of broadband wireless services will lead to more rapid price erosion, while if the opposite happens, that AT&T gets its way and buys T-Mobile USA, and other consolidations happen in the industry, and perhaps Clearwire falls by the wayside and LightSquared loses out to the GPS factions ‘ then prices will remain high, and usage will turn down low in a semi monopoly market, then who is going to invest in a) spectrum and b) build out. Predicting that ARPU will fall by around $1 a year at most, you are heading for a fall during a period in which the US is supposed to reinvent its broadband.
During this no-one who has spectrum to sell but can’t sell it, such as Nextwave and
Clearwire, has been consulted on the price of US spectrum ‘ it clearly is a moveable feast based as much on the timing of availability and previous results, as on history.
And what of the effect of rural fixed line broadband initiatives, which may get a boost from plans such as the one launched this week from AT&T, CenturyLink, FairPoint, Frontier, Verizon and Windstream to build out uneconomic regions using the Universal Service Fund ‘ shifting more broadband to fixed line and removing an opportunity for broadband wireless.
And also what of the white spaces broadband initiative, which may actually reduce the requirement for paid broadband? No, not mentioned.
What it does mention is what it calls the McCaw effect ‘ where Craig McCaw put up $33 million in the PCS auctions to give him bidding rights, and how this drove up spectrum prices by some $825 million. That’s going to happen all over again Brattle says. But there aren’t many McCaws out there and right now Craig himself is fairly disillusioned after Clearwire, so it may not be him.
So that’s covered the auction and the capex, now let’s think about those jobs that this is going to create. So in five years 300,000 jobs will emerge ‘ Well these are not supposed to appear within Cellular operators, at least not all of them. In an industry that has to find the money to buy spectrum, build networks, advertise and promote that network to gain customers ‘ as if they haven’t enough things to be spending their money on.
300,000 people at a cost of employment of $50,000 a year will cost some $15 billion a year in new revenues to pay for. Over a five year period that totals $75 billion of any new revenue generated.
So let’s be clear ‘ these are not jobs at the operators ‘ Verizon (not just Wireless) staff fell by 14,000 from 210,800 a year ago to 195,900; while AT&T fell from 272,450 to 258,870, again by 14,000 over the same period. Do they mean that the wireless arms of these companies will employ most of the people laid off from the fixed line arms ‘ and that’s growth? We don’t think it can be that.
No the suggestion is that 300,000 jobs simply materialize because more people have broadband and that drives $billions into the US economy, but we don’t think that any rules established about the internet and ALL the jobs it has spawned, is solid ‘ because, as a phenomena it’s too recent to be predictable.
We won’t argue that countries which have access to better broadband have a more technology literate workforce, and export more services ‘ we think that’s true, so perhaps we can accept the GDP part of the argument ‘ but that’s all.
Anyway the upshot from all this is that the Mobile Broadband Enhancement Act of 2011which will make all 470 MHz of auctioned spectrum possible without further government sanction, will go through. Any other action will be doomed to go up against the commercial interests of the major operators and look retrograde.
But simply because the Act passes doesn’t mean that anything that was used to excuse it into existence, will come to pass ‘ so don’t expect more jobs, don’t expect better services and don’t expect build out deadlines from the FCC on any licensed spectrum to be intimidating and don’t expect them to be adhered to ‘ Broadband Wireless in the US is weak because companies will not build out the spectrum they already have, because it’s not the ‘right’ spectrum ‘ and neither is much of this.