Where I live at least, the rush-hour traffic seems to be significantly less bad on Fridays. Is it because enough people simply take Fridays off? Or is it because people are just a bit more flexible on Fridays? The boss probably isn’t going to get really mad if you’re a little late on Friday morning, or if you leave a little early on Friday afternoon. So people are a little more relaxed about their commuting time on Fridays, and the times that people hit the road are probably spread out a bit more, which reduces the peak-hour crush.

I’m only guessing but I wonder if this principle could be applied to congestion pricing. Instead of charging a price for travelling during certain times, charge a price for flexibility (like airlines do). Say you allocate everyone a random time window in which they can travel for free during the rush-hour. Outside this window they can still travel but they have to pay, so those who need flexibility can pay for it. At the right price, enough people will choose to stick to their allocated time window, the random allocation might smooth the traffic flows enough to reduce congestion. I guess an experiment would be needed to test this and see whether it works and whether it’s more or less efficient than standard congestion pricing.

Freakonomics has a detailed discussion of the congestion pricing scheme was recently rejected in New York. I read on someone’s blog (I forget who or I’d link it) that the lack of acceptance of congestion pricing is a marketing problem. I agree entirely. The name congestion pricing is bad. You’re not paying for congestion, you’re paying for uncongestion. You’re not paying to use the road that you could have used for free, you’re paying for extra time. You’re paying to keep other people off the road. Don’t market congestion pricing as a punishment payment for the congestion that people cause. Market it as a payment for valuable extra time.

Whoever came up with the name “congestion pricing” must have been an economist …

Here in Japan it has been virtually impossible to buy butter at the supermarket for the past month or so. There is some sort of shortage and the newspaper blames increasing feed prices, China and substitution of milk for cheese production instead of butter. First of all, I don’t know why supermarkets don’t just put the price up instead of allowing stocks to sell out so fast. I guess consumers would complain about getting creamed (sorry) but I think they’d quickly realise that being able to buy butter at a higher price is better than not being able to buy butter at all.

Actually I’m a bit confused about how this situation has arisen, since the supposed causes sound like relatively long-term things, and it was perfectly easy to buy butter earlier this year. The newspaper says that milk producers have reduced production because milk consumption dropped suddenly, which left them with an excess supply. But if they could cut production so quickly, surely they can increase it just as quickly again too. Unfortunately not, says the newspaper, it will take up to two years to breed more cows to increase production.

Adding to my confusion, as far as I know, there is also no butter shortage in any other country. Now I know nothing about about Japan’s trade policy towards butter, but I’ve never seen imported butter on sale. So I wonder if protectionism is causing consumer suffering in this case while domestic dairy farmers are milking it (sorry, couldn’t resist).

How much extra would you pay for an iPod that has unlimited access to all the music on iTunes?

Personally, I think I’d be willing to pay up to 50% more than the price of the iPod alone. I reckon this could be a pretty successful business model. Unfortunately, I recently bought a new iPod. If this rumour is true, I hope they offer an upgrade for existing iPods as well as selling it with new ones.

I’ve been reading How to Price by Oz Shy. The book describes itself as “A guide to pricing techniques and yield management”. In other words, fancy forms of pricing compared to just setting a single price for everyone. The following diagram sums up the basic idea:

At the monopoly price shown, you make profit of the green rectangle. However you’re leaving some surplus on the table. The top triangle is surplus that consumers who bought your product get because they were willing to pay more than the price you charged. The lower triangle is profit you could have made from consumers who were willing to pay more than your marginal cost of production but were excluded by the price you charged. How to Price is basically about how firms can capture these two triangles as revenue. It goes through the standard things like bundling, menus of two-part tariffs and peak-load pricing, as well as more unusual topics like advance booking, overbooking and refund strategies.

The book is well written and clear, but I found it a little disappointing. Each chapter covers one pricing technique and starts off with a numerical example to illustrate the basic idea, which is fine. This is then followed by a “general formulation”, which is not really general but just the special case of linear demand or some other very strong assumptions. This generates a lot of hairy-looking algebra but doesn’t give any further insight into what’s going on beyond the numerical example. I would have preferred to see a more general exposition, although that would raise the difficulty level considerably. I found myself skipping over the “general” parts of the book and just trying to learn from the examples.

One interesting feature of the book is that it sketches out computer algorithms for implementing the various pricing techniques. Again these are not terribly general as they assume some input data in a particular form, but it does give you some insight into how you could write a program to set prices.

Another issue is that the book lacks material on empirical techniques. Obviously implementing complex pricing strategies requires some fairly detailed knowledge of your customers’ demand characteristics, which I suppose could be estimated with some fancy econometrics. Shy is very up-front that the book does not cover this material (he says so in the very first paragraph of the preface). He suggests that since preferences vary so much over time, econometric techniques are not very helpful, and trial-and-error is a better approach. Personally if it were possible I’d go for econometrics first, and then fine-tune things with trial and error.

Anyway, overall it’s not a bad book if you just want to get the basic intuition behind various pricing techniques, and then you can follow up the references for the gory details, and find another book about econometrics of demand estimation (anyone know a good one?). It’s also not very expensive at around US$35 for the paperback. I’ll give How to Price 3 stars.