% fortune -ae paul murphy

Who's responsible?

One of the odd things about the Giffen good idea explored yesterday is that economists who write about this stuff don't want to give up the idea of rational choice - and that creates a conflict because choosing a Giffen good is almost definitionally irrational. As a result they've come up with some rather twisted examples of how buying a Giffen might actually be rational - the Freakonomics article cited yesterday, for example, includes this rendition of one such convolution:

Imagine you are extremely poor, just barely able to afford enough to eat. And for simplicity, pretend there are only two foods: a basic, staple food like bread that gives you a lot of calories and fills your stomach at a relatively low cost, and a luxury food like meat, that tastes good (indulge me, vegetarians) or adds variety to your diet, but is very expensive, offering few calories per dollar.

So, if you're really poor, you'll eat a lot of bread to fill your stomach and get your calories - then with whatever money you have left over, you buy a bit of meat to make yourself happy.

You're going merrily along like this, until the price of bread goes up. Now you can't afford the same bundle of bread and meat you were buying before. You have two choices:

  1. Eat less bread and more meat.
  2. Eat more bread and less meat.

Actually, if you enjoy being alive, you really only have one choice: option two.

The problem with option one is that if you cut back on bread, you lose a lot of calories and a lot of bulk to fill your stomach. And because meat is so expensive, you get very few calories from the small amount you add to your diet. So, since you were just barely getting enough to eat before, you would end up with too few calories and a grumbling stomach. Eventually, you might even end up dead.

But if you instead cut back on meat and eat even more bread than before - while you may enjoy your diet less - you'll at least get enough calories and fill your stomach. Really, you have little choice. So you break the Law of Demand: the price of bread goes up, and you end up eating more of it.

This is clearly explained, but it's hard to see how that process would work out in real life - and, in fact, it didn't work out for Jensen et al because, as he points out in Part Three, what their data showed wasn't a Giffen good but a policy failure because the subsidy at issue distorted choices and produced, on net, a poorer diet for more money.

Luckily the rational explanation for the choices that make Wintel expense a Giffen good is much simpler: senior management's typical refusal to face up to its IT responsibilities has generally left the people who benefit from selling Wintel in charge of Wintel purchasing - and from their perspective every price increase offers both direct and indirect incentives to work harder at selling more.

Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.