There was a fine article on the Adam Smith 'blog pointing out the merits of auctioning off radio spectrum to the mobile 'phone companies (the only alternative is giving it away, i.e. corporatism).
The second commenter said this: "A lot of people think the service providers pay this amount but, surely, in the long run, it is the user who pays?"
Nope. It's all got to do with commonsense/economic concepts like 'sunk costs', 'marginal costs', 'revenue maximisation' and 'marginal profits'.
1. Let's ignore discounting and assume (i.e. these are made up figures to illustrate the point) that the companies estimated revenues from users for the next twenty years (the duration of the licence) at £50 billion and future costs (satellites, mobile 'phone masts, advertising, providing handsets, billing, maintenance, etc) at £30 billion.
2. Once the cheques for £20 billion have been handed over, they are a sunk cost, they play absolutely no part in future pricing policy.
3. Next, the companies spend another £20 billion on long term capital expenditure (satellites, mobile 'phone masts etc). Once completed, these are also 'sunk costs'.
4. The balance if £10 billion is 'marginal costs' i.e. things which relate to current or future income (advertising, providing handsets, billing and maintaining the physical infrastructure). Let's assume that these are fairly fixed at whatever level of usage, and cost £0.5 billion a year for twenty years.
5. These companies are profit maximisers and it makes sense for them to stay in business if their current/marginal revenue exceeds their current/marginal costs. They can only do a trial and error thing to work out which prices maximise their revenue (i.e. if they cut prices by 5% and unit sales go up 9%, it's better to cut prices; if they can increase prices by 5% and unit sales only fall 1%, then it's better to increase prices; there's no right or wrong answer, and what one company does depends on what all the others do, and so on).
5. Provided that they can get more than £0.5 billion a year in revenue, this covers their marginal costs and it is worth while for them to stay in business. Perhaps, try as they might, they can only get £2 billion a year in revenue, so over twenty years they get £40 billion revenue and have spent a total of £50 billion.
6. Overall, they will have made a loss, but the user is not paying for the original £20 billion for the licence, he is paying for the service provided. It is the mobile 'phone companies who paid that £20 billion for the licences.
7. Sunk costs play very little part in all this. Imagine that a few years in to the project, one of the companies with a licence goes bankrupt and is snapped up for £1 nominal by a new entrant. The new entrant has effectively paid £1 for the licence and the infrastructure, for which the predecessor paid £billions. This new entrant has the same marginal costs and the same revenue maximising decisions to make as his competitors, and these is no reason to assume that they will be very different, therefore he will be charging very similar prices for mobile 'phone calls as his competitors.
8. What this boils down to is that the price paid for the licence was a wild guess at a balancing figure, and businesses cannot 'pass on' wild guesses or balancing figures to their customers.
9. To take a simple example, let's imagine you're in the business of retailing apples. A man in the pub tells you he has a lorry load of apples he needs you to shift, he'll take £500 for the lot, and just for a giggle, you buy them. The next day he turns up at your market, and the lorry is full of ten thousand absolutely tip-top apples, which you can sell for 20p each, hooray, your revenues are £2,000 and you've made a handsome profit. Or maybe the lorry only has a hundred apples in it, that are just past their 'best before' date and you can only sell them for 10p each, total revenues £10 and you have made a loss.
10. The £500 you paid him was a wild guess of the price you thought you could sell the apples for (minus your normal daily running costs, which we will call £nil for simplicity) so the £500 is in fact your estimate of the profits you think you might make on the apples, and once paid, it is a sunk cost and has no impact on the prices you charge your customers - you cannot say to your customers: "I'm sorry that these apples are a bit over-ripe, but I paid £5 each for them so I have to sell them for £5 as well."
Are you all set?
5 hours ago
17 comments:
"I'm sorry that these apples are a bit over-ripe, but I paid £5 each for them so I have to sell them for £5 as well."
Mark, no doubt you are as depressed as I am that a majority of people think that that exactly how things are priced.
Since bandwidth is limited it will be rationed in some way. If not by buying licence then by the strength of broadcast signals used to deown out other people wanting to use it or by government allocation on crony ot political approvement. All of these will ultimately cost the customer far more in either money or reduced service than auctioning licences.
B, yup, most people resolutely refuse to accept that "Value does not equal cost" except where it suits them.
NC, correct, but ultimately it will always be 'the state' who rations band width, i.e. they are the only ones who can arrest pirate broadcasters, shoot down satellites etc. Conversely, private broadcasters provide much better content than state broadcasters (or mates of the people in government). Ergo, best of all worlds is the auction system. It's a bit like...
It not a like for like comparison to compare this to an apple sellers as the mobile ops are oligopoly and there isn't a startup with zero cost in UK.
The beauty of economics is that it is about expectation. As far as I can see, the 5 UK operators and their investors EXPECT to get that sunk cost back. When they all expect the same thing, they behave the same way (think H3G tried to be different, but have now backed down anyway) - and so the pricing is likely to include this un-sunk cost.
Further, OFCOM allows the sunk cost to be included in the cost component of the mobile termination charge:
Quote:"The model assumes that 3G costs13, in particular 3G spectrum costs are relevant to the
assessment of cost-orientated MTR."
http://stakeholders.ofcom.org.uk/binaries/consultations/mobile_call_term/responses/euro.pdf (page 4 of 6)
Also see here:
http://www.nuff.ox.ac.uk/users/klemperer/biggestsept.pdf
So..when economic theory does not meet reality...what would you do ?
Anon, proper capitalist economics is about free consumer choice (1) and about risk and reward (2).
1) Seeing as consumers can use email, Skype, landline or old fashioned letters with stamps, and seeing as we managed perfectly well without mobile 'phones until ten or twenty years ago, there is a strict upper limit to the prices they can charge for mobiles. Although 3G revenues were unknown ten years ago, the operators made their best guess.
2) The operators worked backwards from their best guess from (1) and reasonably well known actual costs to arrive at probable value of 3G licences. They bore the whole risk at this stage (assuming the auction was not rigged, and I have no reason to assume it was). And their reward is, if they guessed correctly, they will make a profit (hooray for them) and if they guessed wrong they make a loss (so what?).
The fact that they are legally entitled to amortise the cost of licence is irrelevant - it is possible that it simply wouldn't have worked very well on a technical level, sunspot interference or Chinese jamming the signals or something - and maybe the system could only have handled thousands rather than millions of calls.
In this case. to get their money back, they would have to charge hundreds of pounds per minute rather than a few pence, which no consumer would have been prepared to pay and their income would have been zilch.
Two points
1. There was an element in the 3G auction of "we have to do this or be left behind". The phone companies themselves got caught up in the hype. Rational considerations of getting a return on investment were only part of the picture.
2. A call made in a network has all the costs borne by that network and therefore all the revenue. In contrast, calls between two networks have two parties to share the costs. Since the revenue is earned by the caller, not the receiver, rules exists to share the cost - there is a cost of call termination. There are quite complex rules set by the regulator to control the cost of call termination. The reason being that incumbents like BT have fully depreciated networks, serve the majority of callees and could notionally strangle competition from new entrants by levying a very high termination charge. This is even more complex for VoIP players such as Skype (Skype-in and Skype-out service) and Vonage who own even even less of the transmission network.
TDK,
1. Indeed! So to some extent it was a 'tax on giant egos' or a 'tax on stupidity'.
2. Sure, which is why I chucked the companies into one pot and treated them as one thing for the purposes of these calculations.
If you put in £20 billion for the licence 1. You had the money and have now lost the revenue from that money. 2. You borrowed the money and have a cost servicing that money. How is either a nil operating cost?
PW, where did I say that it wasn't a "cost"? Of course it's a cost!
Did I not explain at length that while clearly it is "a cost" it is a particular kind of cost called "sunk cost"?
I was explaining that "a sunk cost" does not affect the price charged to consumers (if we exclude TDK's point 2).
B, yup, most people resolutely refuse to accept that "Value does not equal cost" except where it suits them."
Or the amazing piece of loony leftie nonsense.
Profit is the difference between the price and what the price should be!!
Licensees didn't have to pay all the money up-front. Assuming that they are able to walk away from any remaining obligation, does the outstanding license fee affect pricing?
F: "Licensees didn't have to pay all the money up-front."
If that is true then I'd have to re-write the article. But if there is an option to walk away, then surely the original bids would have been far, far, higher? If the price is fixed with an option to walk away, then they might as well not fix the price...
Having done a bit more research (you have to root around a bit but the auction site is http://www.ofcom.org.uk/static/archive/spectrumauctions/3gindex.htm), it turns out it's a moot point anyway.
There was a deferred payment option - 50% upfront, the remaining 50% spread between years 6 and 10. This means that the last payments would all have been made this year.
F, thanks for that, but the post was about 'sunk costs' and whether they can be passed on to the consumer rather than the finer contractual details sifted out of 10,000 pages of small print.
Either way, sunk costs are sunk costs. Maybe the telecoms companies overpaid, maybe they underpaid - the point is that the price you are prepared to pay, today, to make a mobile 'phone call is unaffected by a decision taken - in a highly febrile environment by nutters who had never done this before - ten years ago.
In which case, I'm intrigued as to how not having to pay it all upfront would necessitate a rewrite of the post :-)
The reason I ask is because I see the 3G auction as basically being an example implementation of LVT principles. Spectrum is the same as real estate (can only be occupied, not created or destroyed), so I'm pretty sure there are lessons that the auction experience can teach us. I'm just not sure yet what they are. The FT for example is reporting that the licences won't be auctioned again. Ofcom will just set a price, and I can't decide if that's a good or bad thing.
But to restate, my original question is moot. All the money has been paid (and there was no get-out clause) so you don't have to rewrite anything :)
Fraggle, provided that the auction was for the spectrum rights for a fixed period, it is a case of LVT principles in action. However if it was selling them for all time, it's not. LVT is a ground rent and if you sell the freehold without making any provision for ongoing charges, you are losing the incentives which lead to efficient use of the spectrum. In a truly LVT-based auction, companies would have been making rental offers for a fixed period rather than purchase offers in perpetuam.
F, thanks for updating your update. Yes of course it's like LVT (that's the usually the sub-text round here) but I was also trying to address the notion that "landowners will pass on all the LVT to the tenant or add it to the selling price".
D, the auction was for twenty years or thirty years, so it is like LVT.
Post a Comment