Sunday, February 26, 2006


We went to see 'Munich' last night - Steven Spielberg's movie about an Israeli vengeance squad tasked with assassinating those claimed to be behind the murder of 11 Israeli athletes at the Munich Olympics in 1972. The Wikipedia article on the massacre suggests that the movie is not a documentary. A very fine film which we did not forget the moment we walked out of the door.

From matters of blood, honour, homeland, life and death to economics. The cinema tickets cost £6.50 each. We were given vouchers worth £1.50 each to see our next film, so that our next tickets would cost only £5.00 each. Why did the cinema do that?

A first thought might be that by lowering the price, the cinema was stimulating demand, and that this would somehow increase its returns. However, why issue a voucher? Why not just decrease the price to £5 per ticket for everyone. No, only people who had already bought a ticket were eligible for the reduced price.

This is an example of 'diminishing marginal utility'. By buying tickets for Munich at £6.50, we had already established that this was a fair market price for the film we most wanted to see. Had there been another film we valued more highly than Munich, then we would have seen that one.

Given that the portfolio of films changes fairly slowly, the cinema was bidding for our custom for our next-preferred film. As we certainly valued that lower than Munich, the cinema therefore gave us a lower price (£5.00). (Note that the marginal cost to the cinema is effectively zero, so they always make money).

Always nice to see effective price discrimination in action.

Friday, February 24, 2006

Checking out the neighbours: hypertelescopes

Topical item: telescopes capable of resolving earth-like planets around other star systems. According to the current New Scientist, creating a distanced swarm of telescopes in space could create the resolving power to image even quite small features on an alien planet.

While not a near-term prospect, in a few hundred years of progress, why not a a whole bunch of reflectors at the earth-sun Lagrange points L4 and L5? With around 200 million miles separation, you should get a pretty good image. (You might have to wait for the right time of year, for the L4-L5 axis to be orthogonal to the direction to the star system).

Suppose they were hostile. Could we take them out, or could they take us out easily? One weapon the military apparently like is the relativistic impact weapon. If you crank up a projectile to ever closer to the speed of light, its apparent incoming velocity to an observer at the target tends to infinity (the projectile catches up with its own light). Doesn't give much time for defensive action, and the energy delivered goes up by much more than the square of the velocity. Perhaps a combination of laser drive initially and then Orion to really pump the energy up? Terminal guidance might be challenging - from the kinetic weapon's point of view, the target is approaching at speeds which also tend to infinity.
Enough of this bad-taste military SF. I blame the cold weather.

Sunday, February 19, 2006

Age discrimination

I'm 55 and increasingly, the senior clients I deal with are younger than I am.

People in their fifties and sixties complain of age discrimination in employment, and consequentially lobby for regulation. Governments try to raise the pension age for all the well-known reasons, which promises to make the problem of ageism worse. But perhaps market mechanisms can solve the problem?

Suppose a senior worker is just as productive now as they were twenty years ago. But for 'irrational' (= non-economic) reasons they experience prejudice today, monetarised as $P. What does this mean?

If a thirty-something worker gets paid $W, the fifty-something worker will have to accept pay of $(W-P) for the same job. If they are prepared to take a wage-cut, then problem solved.

Or perhaps not. Maybe seniors don't want to suffer a purely prejudicial wage deficit? Can economics remove the $P differential?

If the job is customer or client-facing, and customers don't like dealing with oldies, then the answer is no. There is evidence that better looking people earn around 15% more in client-facing positions.

If the job is back-room, where real performance rather than bodily-age is important, then employers preferring to hire seniors at $(W-P) should have a preferential cost advantage, which ought to bid $P away in a competitive market.

However, if the source of prejudice is younger fellow employees, then the thirty-somethings may require extra pay to compensate for having to work with or amongst a bunch of geriatrics. This nullifies any price-savings on pensioners, and can perpetuate the discriminatory wages.

The model may be simple, but the lesson for employers is to fight age prejudice amongst their own staff, having first assured themselves that there is no performance basis for it!

There is another angle to this. It is widely noted that over a career, workers are first paid below the true value of their work, while at the end of their careers, workers are overpaid. Various semi-plausible reasons have been advanced for why this is the case: perhaps it serves to minimise employee churn where experience counts? However, in a volatile economy, perhaps the rationale weakens, and older workers will simply have to accept their real market rates. Their peak earnings may then come in mid-career. Another nail in the coffin of final salary pensions.

Note: this discussion is taken from the excellent "Microeconomics the Easy Way”, pp. 265-267, by Walter J. Wessels. It was stimulated by a special article on senior working in The Economist.

Monday, February 13, 2006

IMS for VOD?

Could, or should an operator use IMS to manage a Video-on-Demand service?

IMS supports the following relevant functions.
  1. User authentication via the HSS.
  2. User service profiles via the HSS.
  3. Billing based on the user service profile.
  4. Bandwidth allocation via the Go interface between P-CSCF and the first service routing device (GGSN).
  5. Session admission control (via many components, most notably the P-CSCF).
  6. VOD server functionality via IMS Application Servers.

On this basis, there is no question of the adequacy of IMS to manage and bill the VOD service. However, IMS does not directly support the encryption and key management requirements of a Conditional Access (CA) system. Unfortunately, the CA system is quite tightly coupled to user authentication and billing, so unless CA systems are re-engineered in a modular fashion to interwork with IMS, this appears to be a significant roadblock.

Most organisations which have a history of ‘pay TV’ will already have systems which do many of the middleware functions which IMS abstracts and modularises. This will impede take-up in the short-to-medium term. However, if a company wishes to broaden its portfolio, e.g. introduce a ‘quadruple play’ of multimedia-sessions + data + IPTV/VOD video + mobile, then the increased generality of IMS may justify its deployment.

Saturday, February 11, 2006

Interweave Consulting

Following a rethink of business model, I have resigned from Mentor and restarted Interweave Consulting, with effect from February 1st 2006. I am continuing to do client work through Mentor, as an associate.

My immediate business plans include conducting a number of interviews with senior staff at some of the UK's major carriers and media companies for my book on NGNs (link to the right). The material needs to get to the publisher in June, and this initiative should provide the remaining chapters.

Progress will be reported here, and on the web site.

Thursday, February 09, 2006

Why have friends?

We are all partial to family. Our calls, visits and gifts are proximately driven by family affection, and ultimately by shared genes. Conversely, business relations with genetically-unrelated business colleagues are explained by standard 'reciprocal altruism' theories first outlined by Robert Trivers.

Friends are a problem, though. Why do we invest significant personal capital in people who are not kin relations? Friends get a level of personal commitment which extends far beyond that which lubricates routine business or social transactions.

Tooby and Cosmides, pioneers in evolutionary psychology, suggested it was a form of investment. Favours in the favour bank for when you really needed a favour back. This certainly has the ring of truth, but how to select the candidates for friendship? - After all, friendship is underpinned by emotion, not (just) by rational calculation.

In a recent article*, Krebs suggests that friendship is a 'mis-firing' of kin-preference mechanisms. Animals cannot directly detect the genes of kin conspecifics, and need proxies such as appearance and geographical proximity. Many friends resemble each other and friendships bloom via close proximity. The language of friendship is the language of kin.

This suggests a number of research issues. Whether friends mostly inhabit sibling-like relationships, or whether other familial relationships are also expressed in friendship modes. And whether having siblings (and of which gender) correlates with the number and type of friends (controlling for personality variables).

A final thought. Real friendship involves significant personal investment and costs - as they say, it's only when you're down and out that you find out who your real friends are. Most people would be horrified to discover just how few friends they actually had.

* Donald Krebs, 'An Evolutionary Reconceptualisation of Kohlberg's Model of Moral Development', chapter 9 in "Evolutionary Perspectives on Human Development" 2nd Edition. Robert L. Burgess & Kevin MacDonald (Eds.). Sage Publications, 2005.