Why Economics Doesn’t Understand Inequality

Note: This post is far from perfect, and in places probably flat-out wrong. This is a blog for BHAGs and controversial opinions, not for being careful. Also: this post is about within-country inequality, not cross-country inequality, and will unabashedly feature large block quotations created with mac default fonts and colors. Prepare to be visually assaulted.

Growing up I was your typical well-off wannabe do-gooder. I was sensible to the appeals of charity, into the idea of "social justice," and thought that wanting to make money when you grew up was terribly gauche. I knew there was something wrong about a society where the top 1% have nearly 40% and the bottom 40% have less than 1%. I decided that in college I would study economics to figure out the structural reasons why some people became rich and other people became poor, and from there choose a typical do-gooder career that would help make the system as a whole more fair.

Of course I knew that some economists claimed that the "invisible hand" essentially meant that the poor deserved to be poor, but I had always assumed that those people were either idiots or bigots, just as I had decided those who cried "exploitation" were missing the point.

I was sure that my study of economics would reveal a reason for inequality, and that, armed with understanding, I could dive directly into the equally difficult but considerably more straightforward task of application. (This isn’t just imaginative backfill, by the way… if you dug up my college application essay, I say the same.)

Yea, something like that.

I was disappointed. I waited through micro, macro finance, growth. I took advanced courses. I read on my own. No one could tell me why the system wasn't fair. They could tell me how many widgets I should produce to minimize cost, how many gizmos I should buy to maximize utility, and what Mary should do if she happens to be simultaneously unable recall if Bob said he would meet her at the opera or the football game and that cell phones exist. They could tell me that the poor are getting poorer, and that poor people are bad for growth, and they could even suggest some strategies that might help the poor become less poor, but they couldn't tell me WHY the poor are poor, except in cases, like disability, discrimination, or disaster, where it was rather obvious. I started to believe that economists were just a bunch of classist assholes after all.

Then I finally got it. They don't know.

Or rather, they can only explain it by saying that it is someone's fault. You see, economists generally make the world out like this: there are a bunch people who make stuff and trade with each other, and these people choose what to make and what to trade in order to optimize their own personal happiness, or utility. Things get a bit more complicated when you add time, uncertainty, and money to the mix, but that’s the gist of it. If you work harder, or happen to be particularly talented, you make more, trade more, and therefore get more. If don’t work hard, it’s because you get more utility from leisure—slacking off—than from consumption, or because you’re just not very good at making stuff. In other words, if you’re consuming less, it’s your fault.

Accumulation across generations again makes things more complicated; just because your father preferred to nap all day doesn’t mean that you’re also slacker at heart, but he might not have been able to give you the leg up that a more diligent parent could have. Because of this nearly universally recognized Rawlsian injustice, we have passed many laws and created many public institutions that seek to create equality of opportunity. But yet inequality persists, and exists in such exponential relief that often even the enormous shortcomings in the quest to create equal opportunity can't explain all of it, and at the root of it was still daddy's fault. So we get this kind of thing: 

Source: Defending the One Percent. Mankiw, you're great and I
get it, but I'm still not sure you can say that in polite conversation. 

I think that most economists start out sympathizing with the poor. They know from observing, like I did growing up, that something is unfair. Then they go to school and learn the basics; supply, demand, price, competition. All that is interesting, but still pretty detached from the real world. Then they learn about utility, externality, elasticity, and statistics, and they start to think they know something. They apply rigorous models to real world situations. They apply some other models and add their own tweaks. They get peer reviewed. And they realize that all of the models spit out the same result: inequality, at least in the beginning, is somebody’s fault. Hence the slow transformation of the well-off wannabe do-gooder into the classist asshole by the most sublime of all evils: logic.

I'm exaggerating of course. Most academic economists are quite progressive, like the academic demographic as a whole. Unequal opportunity really does account for a lot of inequality—there are rich literatures that model and empirically verify the effect of various types of advantage, such as health and education, on welfare. Most academic economists agree that life is not fair. Still I think that if you ask an average economist to tell you why people may become poor in the first place, you'd be hard pressed to find an answer that's not "some people contribute a greater marginal product" or "some people value leisure more than others." And sometimes inequality IS someone's fault.

I didn't really who GK Chesterton is, but he confirms my bias. THUMBS UP!

So is that it? Does the logic tell the truth? I don’t think so. I think it’s missing something. Here’s a metaphor: imagine that a woman goes to the pet store and buys two fish, but when she gets home finds only one fish in her tank. She returns to the pet store and tells the manager that there is a problem. The manager counts the fish left in the store tank and says no, she got two fish. The beneficiary, the manager, thinks it’s the customer’s fault. The have-not, the customer, thinks it’s the manager’s fault. The outside observer, the economist, agrees that it must be someone’s fault. But they are all missing a key piece of information: sometimes fish eat fish.

Great thinkers simultaneously hold on to the customer’s and the manager's perspective, and try to look for a way to change the model to explain how one fish could disappear when no one is lying, or how income distribution could become so patently unfair when not all poor people are lazy or stupid. Marx tried to fix the model, and he came up with the Industrial Reserve Army and a really complicated theory that most acolytes of the god of logic agree is bullshit. Keynes tried, and he came up with a really complicated theory that seemed to hold promise for a while but that herr logica eventually cut down to sticky prices. Stiglitz is trying, but as far as I can see the farthest he’s gotten is a well-written book that spends a long time describing the fish in the tank and a long time explaining why having only one fish is so bad. But why is there only one fish in the tank? WHY?

Drummmrrroolll please... I think it's choice.

Right now it's very in vogue to think of choice as a bad thing. There’s this TED talk with a guy complaining about how much he hates buying jeans. There's this web 2.0 bible called "Don't Make Me Think!” There is article after article bemoaning the complexity of the modern world and wishing we could all just go live deliberately in the woods for a while. (Pssst... nothing's stopping you!) But choice is also power. When two people apply for one position, the interviewer, who might in all other respects be equal, becomes a god. In retail, they say that “the customer is always right” because the customer, unlike the merchant, can choose to walk away. In high school, the queen bee is the girl who can afford to slight friends because she has plenty to choose from. In politics, voters have power because they can choose their officials; but as the number of candidates to choose from decreases, so does their power.

Having choice is important.

Choice is already an essential concept in economics, but in a slightly different sense. Companies are always minimizing cost by choosing how many widgets to produce, and consumers are always maximizing utility by choosing how many gizmos to buy, and Mary is always choosing what strategy she should play because she is still having trouble communicating with Bob. Economics is very good at telling us which choice we should make, given defined set of options, but it is less good about explaining the effect of different amounts of choice.

I’m still an amateur in economics, so I may be missing or misinterpreting a wide swath of literature, but from my reading the economist to come closest to formally incorporating the idea of different amounts of choice into an economic framework is Amartya Sen, with his concept of capabilities. Instead of using utility as the ultimate measure of welfare, like most economists do, Sen uses “capability,” or what an individual is capable of. In his book Development as Freedom, Sen emphasizes the difference between someone who is starving, and someone who is fasting—both outcomes are the same, but only the latter has a choice.

There have also been some investigations into the importance of the amount of choice using the traditional utility framework. Game theory can do a good job of demonstrating the advantage of choice in specific situations; in a dictator or ultimatum game, for example, the "proposer" has a greater amount of choice and therefore more power. But it’s a long jump from a game with rigid rules to the real world. I’d like to believe that we can build up a general understanding of choice from micro models in game theory, but I think that’s mainly wishful thinking. It’s the same reason micro and macro split 80 years ago; micro is too slow for practical application. A more heuristic but less scientific approach like Sen's capabilities may lead more quickly to a working understanding of inequality.

Last but not least, I've been interested for a while in the connection between transaction costs and inequality. In 1960, Ronald Coase wrote a paper called The Problem of Social Cost that proved that, if transaction costs are sufficiently low, negotiating parties will reach an efficient outcome no matter the starting distribution. Coase’s larger point, however, was that transaction costs are basically never low enough for this result to occur. For example, if two people are disputing a land title, the claimant with greater access to authority will probably win the dispute; transaction costs in negotiation become a key factor in resolution. From another angle, you could say that the claimant with greater capability in negotiation, or a greater choice of options to support his case, had the advantage.

Another example: imagine two shopkeepers, one who speaks Spanish and English, and one who speaks only Spanish. Both of them negotiate with suppliers to stock their shops. The shopkeeper who speaks both Spanish and English can negotiate with both English-speaking and Spanish-speaking suppliers, but the shopkeeper who speaks only Spanish can negotiate with only Spanish-speaking suppliers. Which shopkeeper is more likely to succeed? Here language is a transaction cost and a capability that gives one shopkeeper more choice.

Unfortunately "The Problem of Social Cost" was one of the greatest academic backfires of all time; since 1960, the “Coase Theorem” has been abused in economic papers and legal cases to show that outcomes are “efficient,” implying fair, when they are really neither. I wonder what economics would be like today if instead of pretending that we live in a world without frictions, economists had gotten serious about incorporating those frictions into their models.


  1. "pretending that we live in a world without frictions."

    If that's what most economists were doing, the theory literature would be tiny. Most papers are about how various frictions move results away from the frictionless case. It has become very popular to criticize a caricature of the econ profession in which everyone is going around solving bare-bones RBC models, but that caricature is not accurate. There is a large literature exploring various drivers of inequality other than choice. See, eg, Quadrini 2000, Terajima 2006, Cagetti and Di Nardi 2006,2009, Guvene's stuff, Heathcote, Storesletten, & Violante 2010, and so on. Economists care a lot about inequality, even if we have yet to arrive at a consensus about all its causes and consequences.

    That said, I like this post; but I think to be convincing you need to decide who is your audience. If your audience is people outside the econ profession, you may be persuasive as they will also be relying on caricatures. I don't know how persuasive it will be to people inside the profession, especially those who have spent a lot of time on this issue.

  2. Yes definitely oversimplified there, as in many other parts of the post, and thanks for the paper recommendations! Also yes, this blog in general suffers from a lack of audience definition... I don't really want to write completely popular stuff, because that is like all caricatures, but I don't feel quite enough a part of economics yet (going back grad school in the fall) to write for economists. Mainly it's for me to flesh out my own ideas :)

  3. Works for me! Very thought provoking in any case.