Of course immediately after the "successful" auction telecoms couldn't afford to do anything with the acquired licence and telecoms went into recession.

3G rollout and services took a ridiculous time to arrive.

Same in Germany.

The market for mobile Internet access here still suffers from this...

Intestingly (I think) the UK was very slow in adopting 3G, but 4G has come a lot more quickly and I believe we're one of the countries with the higher average 4G speeds now.

That's not true at all. 3UK had one of the world's first operational 3G services.

4G LTE rollout in the UK was and is painfully slow. The only good thing was that a lot of networks had very good coverage of HSPA+ and DC-HSPDA which is pretty close to LTE responsiveness in day to day use.

From a technical perspective, yes the UK had early 3G, however uptake was slow compared to 4g.

Iirc one of the reasons the US has poor 4g is because the telcos there branded HSDPA as 4g there while it was still branded as 3g in Europe.

So when real 4g (LTE) arrived, European providers had more incentive to roll out for marketing reasons.

This might be half remembered bs though.

Very misleading title, especially for a scientific paper. While it was the biggest auction up to that date, a similar auction in Germany just a few months later had a significantly higher revenue, nearly twice as much as the British auction. [0] The paper was published in 2002, so the title was already incorrect when it was published.

[0] https://en.wikipedia.org/wiki/Spectrum_auction#Germany

And after that, hopes for a big auction at your friendly neighborhood the Netherlands where high! Expecting similar numbers, the action became a big flop with just under 2.5bil euro's.

The paper acknowledges this right there in the first page (footnote 2).

"The German telecom auction subsequently raised even more in cash terms (although less per head of population) ..."

I don't understand why the licenses seem to be 'forever'. Surely a 25 year license which is then sold again at auction would have been a better design.

The 'forever' licenses also can't be resold (although the entire company owning them can merge). That means if in the future another technology which completes with 3G so incumbents don't want to get involved comes along and could make great use of that spectrum, it won't get use.

The article mentions that there licences were to last only until 2021. Not sure what happens in 2021 though.

Edit: after a bit of googling it looks like the licences will be re-tendered in 2021

I saw a talk by Ken Binmore, the organiser of the auction and author of the paper.

He found it very funny how each participant turned up with teams of Nobel prize winning economists and mathematicians.

He said the correct way to approach such an auction is to decide up front what the value is to you and then bid up to but not beyond that figure. That is all. There are plenty of foolish things you can do but nothing better.

> decide up front what the value is to you

That's the tricky part ;)

I remember that the designer of the auction did various mock auctions beforehand getting maths PhD students to do the bidding pretending to be the various mobile companies. I was a postdoc at the time and didn't take part myself but some of my friends did.

ahem http://www.reuters.com/article/us-fcc-auction-idUSKCN0ZV2C3

With respects paid to Paul Milgrom who designed this auction!