Tim Harford

The Logic of Life

They gave their rats the choice of two drinks, each of which had its own lever. One was root beer, a longtime favourite with your average lab rat. The second was water flavoured with quinine—tonic water, in other words. Rats don't like its bitter taste, but the researchers had made the servings of quinine solution much more generous than the servings of root beer.

Think yourself into the rat's position for a moment. You're thirsty. The root beer is delicious but it's expensive, so you compromise, slaking your thirst on the nasty quinine solution but also enjoying some root beer. You don't press the lever at random.

Now, what happens when the price of quinine goes up a little—that is, when the servings become less generous? To an experimental psychologist, the answer is simple. You're getting less for your money from the quinine lever so you should press it less frequently. That seems sensible. But it happens to be irrational, as an economist could attest and a rat instinctively grasps. As a smart rat, you drink more quinine when it gets more expensive, as long as the servings are still larger than those of the root beer. That's because you're responding to your budget as well as the price. The total consumption of liquid—root beer plus quinine water—is what's keeping you alive. Quinine water is still cheaper than root beer, and because the experimenters have made you poorer by raising the price of quinine, you are obliged to drink less of the expensive root beer and slake your thirst by consuming even more of the nasty quinine water, which remains relatively cheap.

Battalio, Kagel and Kogut showed, quite convincingly, that this is exactly what rats do. By consuming more quinine when the price of quinine rose, the rats had solved a conundrum that went back to 1895—do 'Giffen goods' exist? A Giffen good is a good like the quinine water, one that is such a wretched necessity for the poor that when the price rises, demand rises too, because the price rise creates more poverty and the poverty creates more demand. As an impoverished economics student, I imagined my staple diet of baked potatoes might be Giffen goods: if the price of potatoes rose, I would not be able to afford the cheese or tuna fillings and would buy larger potatoes instead. Over the years, economists had suggested, but never proved, that foods ranging from potatoes during the Irish famine to noodles in rural China are Giffen goods. Battalio, Kagel, Kogut and the rats provided the first incontrovertible example: quinine water.

Yet the real significance of Kagel and Battalio's experiments was not to settle obscure Giffen goods wagers in economics departments across the world. It was to establish that the rats showed surprising intelligence and responded to their full range of options, including the way that their present choices would restrict their future choices. Given the chance, even rats can be rational.

This isn't really about the rats, of course. It is about the way that rational decisions can be made without conscious calculation. I've already drawn an analogy between rational decision-making and the fiendishly difficult differential equations that describe the trajectory of a ball in flight; ask a typical cricketer to solve them with pen and paper and he's not likely to do much better than your average rat. Yet give him a glimpse of a flying ball and he will turn, sprint and then twist round in just the right spot to make the catch; some part of him was solving the differential equations after all.