Lazy economics

Or, how to assume your conclusion and pretend you have unique insight into policy.

Christine Li promises to explain the merits of different approaches to helping the homeless "using basic consumer theory". Warning bells should be going off. In general, using basic economic theory is a good way of oversimplifying issues and getting wrongheaded conclusions. But homelessness? It takes a special level of gall to think that basic consumer theory would shine any helpful new light on something as complex and difficult and pervasive as hundreds of thousands people forced to live on the streets. And unsurprisingly, there's no such insight.

You can read the post for yourself, because it's short and any gloss I wrote would be uncharitable. The conclusion, though, is that it might be more beneficial for homeless people to give them cash transfers than in-kind benefits. This is not necessarily a wrong conclusion. But the argument has massive problems.

First, it isn't actually an argument for the conclusion drawn. The post depends on a probably untenable distinction between "primary" and "secondary" goods. The goal seems to be to compare consumption of essentials and non-essentials. That's probably okay, and you can use a stylised distinction like that to bring forward a point. And the point being made is that just primary goods aren't sufficient to increase utility, so continual in-kind transfers of them - things like shelter, clothing, food - don't really leave homeless people better off. Which, if true, is an excellent argument for not just giving primary goods to the homeless. But it's not an argument for giving them cash! It just means that if you care about improving the welfare of homeless people, you should try to make sure they get things which are leisure or 'luxury' items as well as the essentials. And as it happens, this is indeed a thing. So, tip one: know what your argument actually is; and tip two: know a bit more about the reality of how a problem is addressed before you try to critique that approach.

The second, and more pressing problem: given that the pointy end of the argument is that increasing primary goods alone won't increase welfare, it would be pretty bad if it turned out that that result was built into your case as an assumption. But it is. Compare the setup:
Two sets of consumption goods can be differentiated: primary goods (necessities: food, shelter, clothing), and secondary goods (everything else: education, skills courses, luxury goods). We can assume that like everyone else, rough sleepers would want to consume these two sets of goods in fixed proportions, for example a bed (secondary) with each hostel room (primary), or a T.V. show with each meal. This makes primary and secondary goods perfect complements – the two must be consumed in fixed, but not necessarily equal proportions to give a consumer maximum benefit – or, as economics students would say! – maximum utility.
 to the conclusion:
something is problematic. Under this system, the homeless are stuck on the same indifference curve no matter how many more primary goods they consume. In other words, their situation changes very little despite receiving extra food stamps from the Salvation Army, or collecting marginally more loose change from panhandling (which would likely go towards the cheaper primary goods anyway).
Emphasis is mine. In the first quote, we're told to assume that essential and non-essential goods only increase utility in fixed proportions, so increasing just essentials doesn't leave the homeless better off. In the second quote, we're brought the conclusion that, um, increasing just essentials doesn't leave the homeless better off. This would be an interesting result, but it's nothing more or less than the assumption made at the beginning.

Is there any reason to believe this assumption is reasonable? Not that I can see. It's pretty intuitive that homeless and marginalised people will be better off if they're able to access more of life's essentials, even if they don't also get to consume more non-essentials. Would they be even better off with non-essentials? Sure. But even without them, increasing the primary goods is likely to improve quality of life. Let's hew as closely as we can to the presented argument. We're being asked to assume for the argument that a homeless person given an extra night in a hostel, or an extra meal, is no better off unless they are also given comfortable bedding or the chance to watch a TV show. Does that sound sensible?

A caveat. This site is run by and for students. It's not a sign of massive problems in the economic profession. But my microeconomics lecturer also frequently urged us to think of the subject as a method, rather than as having its own subject matter, and maintained that it could be applied to lots of issues not normally considered as within the scope of economics. And that's true. But encouraging the application of really basic and simplifying economic theory to really complex social issues can lead down a troubling path. Because if microeconomics is a method, it's one which depends in a lot of ways on assumptions. If it's going to be helpful, you have to pick assumptions which simplify the problem at hand whilst also being plausible. That's a really difficult balancing act! The only way to know which assumptions are really plausible is to analyse issues very closely. That takes it out of the realm of basic theory. You can't just bring indifference curves and budget sets and extreme assumptions to any social issue and expect to get helpful results.

(There are also some strange incongruities between what the post's text says and what its indifference curves and budget lines actually present. I might go back over them in another post to pretend I'm revising.)

Comments