So You Want to Learn About Economics

The coolest papers I know

Economics can be intimidating to get into from the outside. I want to show you why I get so excited about economics. I have 11 papers that make me go “woah”, and notes on each of them. Before anything else, here’s the list:

1. Friedrich von Hayek, “The Use of Knowledge in Society”, 1945 (*)

2. R.A. Radford, “The Economic Organization of a POW Camp”, 1945 (*)

3. Klein, Crawford, and Alchian, “Vertical Integration, Appropriable Rents, and the Competitive Contracting Process”, 1977 (*)

4. Clark, “Why Isn’t the Whole World Developed? Lessons from the Cotton Mills”, 1987 (*)

5. Bulow and Roberts, “The Simple Economics of Optimal Auctions”, 1989 (*)

6. Kremer, “The O-Ring Theory of Economic Development”, 1993

7. Romer, “The Origins of Endogenous Growth”, 1994 (*)

8. Stiglitz, “The Contributions of Information Economics”, 2000 (*)

9. Acemoglu, Johnson, and Robinson, “The Colonial Origins of Comparative Development”, 2001

10. Bloom, Eifert, Mahajan, McKenzie, Roberts, “Does Management Matter? Evidence from India”, 2013

11. Donaldson “Railroads of the Raj” 2018 

These are quite biased to what I, personally, am interested in; an equivalent from someone more interested in monetary policy, or education, or lab experiments, might come to a very different conclusion. If one wants a truly comprehensive macroeconomics syllabus, the reddit user integralds put one together ten years ago, which I found helpful to read through. He recommends ten critical works for macroeconomists in the front.

If one wants to understand the rudiments of economics, I suggest reading an introductory textbook, or taking a beginner class. MIT opencourseware has an intro econ class here; I cannot recommend the class from experience, but have found them to be an excellent resource in other topics. 

I read a lot of papers. What I have found is that the right way to read economic papers is to not read them cover to cover, except in uncommon cases. I have marked with a “(*)” those readings which I would recommend you read beginning to end, with the remainder being ones which you can skim through some of the more chalky data presentation parts without losing much. 

And with the prefatory remarks out of the way, onto the notes!

The Economic Organization of a POW Camp: 

The author was imprisoned in a Nazi camp after being captured in Libya in 1942. While the soldiers did not have to work to get sustenance, they nevertheless found trade highly advantageous. The Red Cross sent parcels, which contained cigarettes, margarine, tinned carrots, and other such luxuries. (Alright, tinned carrots might be stretching the definition of “luxury”). One could improve their allotment of rations by barter, and so we see how trade makes people better off.

Of course, finding someone to trade your bundle of goods who has the same wants is difficult. In order to reduce the cost of transactions, money arises — in the prison camps, cigarettes. Even the people who did not smoke cigarettes found it advantageous to deal in them, as they are durable, uniform, and commonly wanted. In short, we see how free markets arise as a natural product of people’s wants. It’s been a staple of introductory syllabi ever since, and it was on mine. 

The Use of Knowledge in Society: 

We want to know how to maximize output. If we knew all the facts, we could do it; but it is the nature of the world that information is dispersed among many people, and is never known in its entirety by any one person.

To Hayek, prices are simply a way of revealing information. They are not an arbitrary imposition from the gods, but reflect all the possible other uses of those resources. Hayek stresses the myriad ways in which things can be used, and the impossibility of eliciting all of this knowledge in a way which is legible to any central planner. This is deeper than simply incentivizing people to reveal honestly. Much knowledge is tacit, and cannot be expressed to anyone. A central planner would never be able to maximize welfare in the way that markets can.

There is still much room for government action, which, while speculative and certainly not optimized, may still be useful. No one could say what the precise social cost of pollution is, but hardly anyone would deny that attaching some cost to pollution is better than no cost at all. When property rights are clear, though, there is hardly any room for improvement. To have good things happen, it is necessary for us to not act.

Vertical Integration, Appropriable Rents, and the Competitive Contracting Process:

It isn’t necessarily obvious why firms exist, or given that they do exist, why everyone hasn’t become one big firm. There has been a lot of work on this, of which I chose this as the best combination of accessible to the beginner, and up-to-date factually. Ronald Coase kicked things off with “The Nature of the Firm” in 1937, which views the firm as a method to economize on transaction costs. Grossman and Hart got the last word in 1986, which argued that firms are due to unspecifiable contracts. (Moore has some remarks on this paper, which you will find either charming or deeply obnoxious depending on your taste for economist navel-gazing). Klein, Crawford, and Alchian (whom I will refer to as KCA) are the line of the second.

The paper is really quite simple. Suppose a firm is considering whether to buy some products, or own the company which produces those products. If the products are highly specialized, then the parties are extremely vulnerable to opportunistic behavior by the other, and they would prefer to be owned by the same owner. If the products are not specialized, then it makes more sense to use the market as usual

That’s it! The rest is merely illustration. It gives us a clean, simple rubric, though, for understanding why it is that 

Why Isn’t the Whole World Developed? Lessons from the Cotton Mills:

This is perhaps the most out of place article on the list, being a history paper in disguise. But what a history paper! To me, economics is not really a field of study – though it can be – and not even a collection of tools – though it can be – but an ethos. Anyone who thinks really seriously about what caused what is an economist. 

It is also an unusual argument structure. Most economic papers advance some proposition, give a statistical test which would favor or disfavor the hypothesis, and report how it goes. Clark’s thesis is that cultural differences were responsible for large gaps in productivity between countries. Culture is not something easily observed, so instead he argues why every other possible explanation is wrong, which he does convincingly. It couldn’t be that their capital was worse, or that there were barriers to capital, or that the skills were difficult to learn, or that the management was worse – in India, textile firms were run identically to firms in Britain, at a far lower level of productivity. Clark argues, with support from many accounts and descriptions of textile practices, that workers did not want to be as productive as in Britain, even with substantial raises in pay. Why? That cannot be directly observed, but one wonders what other explanations are left. 

The most intriguing part, which if you don’t read the whole thing you will miss, is that it might explain immigration. Immigration is honestly kinda weird. Surely it is easier to move capital than to leave your home country, go to a place which speaks an entirely different language, and subject yourself to years of risk. Yet, people do it. Clark thinks the most ambitious are leaving their stagnant cultures behind. Think about this section when you are reading the O-Ring Theory of Economic Development later – they are as peas in a pod.

The Simple Economics of Optimal Auctions:

This is what kicked off my interest into auction theory. It is an explanation of Myerson 1981’s paper “Optimal Auction Design”, which I have linked here for you to read. The significance of it was lost on me when I first read it, though, and I suppose it was for many other economists when it came out. Bulow and Roberts analogize their subject to a monopoly engaging in third degree price discrimination, which is more familiar ground for economists.

And what exactly did Myerson find? Suppose a monopolist wants to sell a single good a single time. What is the optimal way to do this? Assume for convenience everyone has a value for the good randomly distributed between 0 and 1. The correct method is to have a second-price auction (highest bidder wins, and pays second-highest price – the reason for this is that it makes it optimal to bid your true valuation. This is isomorphic, by the way, to an ascending auction – the kind you might be most familiar with) with the seller bidding at .5. With one buyer, the expected value is .25, with 2 5/12ths, and so on. If we break the assumption that everyone has the same distribution, and instead say that each person has some known but different distribution, the seller sets a reserve price for each individual at the midpoint of each person’s distribution. 

We can think of this as maximizing an equation. Suppose people’s value is between 0 and 1. The expected profit of each person takes the form x(1-x) – at a price of 1, you have a 0% chance of it being taken, at 0 it is taken every time but you earn no profit. The derivative of that is 1 - 2x, and that crosses 0 at .5. The derivative line is marginal revenue, and just as in a monopoly engaging in third degree price discrimination, you set the price equal to where marginal revenue equals 0. In fact, you can think about price, as in a grocery store, as really being a reserve price in a one on one negotiation. 

Myerson also discovered the principle of revenue equivalence. If an auction method assigns a good to the same bidders, then their expected revenue to the seller is identical. One must assume risk-neutrality and that buyers are not uncertain of their private valuation, but given that, the form of the auction doesn’t end up mattering. It’s a cool paper, and Bulow and Roberts wrote the best explanation of it. 

O-Ring Theory of Economic Development: 

I have written about this paper before here; Kevin Bryan wrote an exegesis here as well. I think about it every single day, and there is probably no paper which has influenced me more. It raises and answers so many questions. 

• What would happen if production required multiple steps, an error in any of which destroys the whole product? Error rate would have a non-linear impact on income, and so we can explain the differences in income between countries as a function of skill. 

• Why should we expect immigration to increase human capital in the less skilled country? Because it also implies multiple possible equilibria of skill attainment, most of which are sub-optimal. Allowing immigration changes people’s expectation of what they can earn, and so causes them to invest more into human capital, even if they don’t end up moving, as I wrote about before.

• Where should we expect raw materials to be made? And where are the finished goods made? In the poor countries and in the rich countries, respectively.

• Where will the highest skilled workers be? With the other high skilled workers, regardless of job. We should expect the best paralegals to be paired with the best lawyers, and so on.

In the original paper, Kremer includes a fair bit of formal modeling of these ideas. Don’t be intimidated if you cannot understand it. He is simply restating his propositions in the language of mathematics. If you can understand the logic behind them, you understand them, whether or not you can read the notation. This happens a lot in economics. As a novice, if you can’t understand the notation, do not slog through it! You will simply waste your time. You’ll pick it up eventually, but for now, read for the ideas. 

The Origins of Endogenous Growth: 

This is the culmination of several papers by Paul Romer on endogenous growth, and is the best introductory reading for the macro modeling of growth. He sees it as coming out of two strands of thought. First, old models which took technological advance as an exogenously given function of time implied that countries would converge, with growth rates being faster in the poorest countries and slowest in the rich. The places with the least capital should have a much higher return, and capital should flow from rich countries to poor, as Lucas pointed out. This has clearly not happened. Convergence, while clearly not impossible, is also not inevitable. 

At the same time, economists began moving away from modeling competition as perfect competition. We moved toward monopolistic competition, where each individual firm can affect price through their actions and goods are differentiated. In practical terms, this means that price needn’t equal marginal cost. Starting in the late 70s with Dixit-Stiglitz, we got tractable models of monopolist competition, and could now deal with it in a mathematical way.

The way forward was to relate the rate of technological development to elements of the model. It is possible for places to diverge now; if the rate of return on technological research is lower than the rate of return on capital, then no growth ever occurs in the long run. We can have divergence, and we can have imperfect competition, and we can model it precisely. I wrote about macroeconomic modeling here, in what is rather a long article – hopefully you can find it enjoyable. 

The Contributions of Information Economics:

I cheated a little. This is not a paper which advances science in and of itself, but is instead a history lesson and a review of the literature. It shows all the ways in which our standard conclusions break down once you drop the assumption of perfect information. Asymmetric information is not like a little sand in the gears, but can completely reverse our findings. Highly worth reading in full — I read while walking laps around campus in April of this year, and I legitimately think it changed my life.

The Colonial Origins of Comparative Development: 

This paper is here not for its conclusions – which, while not necessarily wrong, do not have the data to support them in this paper – but for its methods, which were revolutionary. We want to know why poor countries are poor. Acemoglu, Johnson, and Robinson (who just won the Nobel Prize) think the answer is “institutions”. Institutions are defined as the “rules of the game”, more or less – rule of law and democracy and respect for private property and everything else that’s good. But how can we tell that those things are actually causing better outcomes? People might get rule of law because they want it for its own sake as they get richer. 

What they use is an instrumental variable strategy. They use the mortality rate of colonists, which, they argue, affected what kind of institutions a place had in the past, without having much of an effect on present day growth rates. In places where people expected to die quickly from tropical diseases, they aim to extract as much as they can and leave. In places where people expect to live, they create inclusive institutions which protect private property and are in it for the long haul. Institutions persist over time, and in the present day those better or worse institutions lead to different levels of development.

Later work argues with it. Glaeser et al (2004) argues that actually, everything was endogenous to human capital to begin with, and the skill of a country (remember Kremer?) is more important in finding its later outcomes. Places with better mortality rates attracted more capable people. Albouy (2011) shows (over the unconvincing replies of the authors) that the mortality data for the instrument is unbelievably shaky. Honestly, what else would you expect of trying to show the mortality rates of Africa in 1800 but a lot of imputation and guesswork? An even more fundamental objection is that they measure institutional quality by giving it a rating for each country, and then interpreting those numbers as meaningful. There is no reason to think that doubling your rating means the government is now twice as good, yet, that’s how they treat it.

AJR 2001 (as it is known) isn’t here because of that. It’s here because it inspired me, and many others. We can answer the big questions in history and development, and we can do it with the data we have. All we need is some sort of exogenous difference. 

Does Management Matter?:

Firms in poor countries are smaller and less productive than those in the developed world. Why? This paper argues that it is about information – firms could choose better techniques, they just don’t. To support this, they have a randomized controlled trial, in which some firms were given management advice and others weren’t. Incredibly simple management advice like “record inventories” led to a 17% increase in profitability, and what’s more, these gains stuck around years later. 

I think these constraints of imagination are a really big deal! I wrote more about them, and other papers, here (in what is and may always be my most popular article). Bloom and van Reenen, with various collaborators, have an extremely cool literature on the impacts of management, which I highly suggest reading. Like culture, management is extremely hard to quantity, but nevertheless they have managed.

Railroads of the Raj:

This paper had an extraordinarily long gestation, having been worked on since at least the mid-2000s. It was worth it, though. It is everything modern economics can be. I do not think I can surpass Kevin Bryan’s writeup of the paper here, and so I will yield gracefully to comparative advantage. I have distinct memory of reading this, though. He uses the price of particular kinds of salt which are only manufactured in one district, but are sold all over India, as a way to infer transportation costs. I stood up and jumped around the room, I was so excited.

You may not have taken economics in school, or you might be set upon other subjects. That needn’t stop you from learning economics! The knowledge is out there, and not too hard to pick it up. I’m not especially bright, and could do it – so can you!