## Friday, August 18, 2017

Armen Alchian

Armen Alchian - nicknamed by his friends "the Armenian Adam Smith" - was an economist at Stanford, UCLA and the RAND corporation. He participated in all the economic revolutions of the time - general equilibrium, economics of information, theory of the firm, utility theory and evolutionary economics. Though ideologically attracted to libertarianism, Alchian was a devout pluralist in his methods. He was a great writer - unlike most clear writers, Alchian comes by his clarity honestly rather than by covering up difficulties. Alchian did not participate in the usual tedious academic point scoring games.

The economist who Alchian most resembles is Frank Knight, whose insightful but somewhat mystical Risk Uncertainty And Profit hovers behind much of Alchian's more innovative thinking. Alchian's thinking also engages with the even more mystical writing of Friedrich Hayek, especially his famous essay on knowledge.

Alchian also wrote many important papers on inflation (that is - unpredictability in price level) on resource unemployment and therefore macroeconomics more generally. However, I haven't really read and engaged with these papers, so I will not be able to describe them here. I would recommend that you peruse Glasner's blog for that stuff.

"... automobile makers may buy such things as finished seats from outside suppliers because their inspection is relatively easy. But automobile makers are hesitant to use outside sources to supply sheet metal parts, which are ordinarily only discovered to be out of tolerance only when a car body does not fit together correctly."

- Armen Alchian on what is internal and external to the firm

There used to be two major schools in thinking about the economics of organizations.

There was the Lange-Coase school of high transactions costs. The main take away of this school is that the market cannot see into the firm because it is just too damn expensive to negotiate every little thing out. This school gave economists two avenues to improve life: 1. "Mechanism Design" of institutions with low transactions cost - from carbon markets to fight global warming to Lange's surreal technocratic communism 2. Use the courts to allocate goods optimally outside the market (this is regarded as an important insight - why has never been adequately explained to me)

On the other side was the Mises-Pigou theory that the point of the firm was to organize the production process in a way that "internalized externalities" - made the production process as efficient as possible. This school too leaves us a bifurcated road 1. Reduce legislation so that firms can organize themselves as efficiently as possible 2. Use taxes to discourage firms from doing bad things.

As the above quote shows, Alchian was on the side of Mises & Pigou  in this conflict. The choice of what is internal to the firm is a market choice - the firm keeps close that which is expensive to monitor and lets loose what is cheap.

Alchian's insight here is invaluable, but I find myself in mixed agreement with him. He often emphasizes the firm's ability to find a stable point in all this and the possible suboptimality of regulation - I demure. When Alchian says "Vertical integration identifies for consumers a single point of accountability for service quality", I think - when I am in a polite mood - "Maybe ...".

Fortunately or not, Alchian never engaged in polemic over the conflict between his and Coase's points of view, probably because of their personal friendship and idealogical agreement. It's somewhat amusing to see Coase and Alchian talk as if their theories are in precise agreement when the generation of economists before them had raged as if the distinction was the most important thing in the world.

"A football coach knows that the condition of winning is making more points than his opponent. Does knowing this imply that the coach can know what his team must do in order to win? Does the coach know how this can be done?"

- Armen Alchian on the nature of profit maximization

"There are no implications to 'profit maximization'..."

- Armen Alchian on the positive implications of profit maximization

Alfred Marshall called biology the "Mecca of Economics". His vision was to integrate economics into the vast System of Herbert Spencer. As one of the few people who have read the entirety of Marshall's Principles, I have to admit that his biology was kooky Victorian BS. But the reasons he had this goal were sound, even if Marshall's biology was nonsense.

Alchian made a great stride into achieving a more realistic version of the evolutionary aim. He joined the fray in a rather kooky way - in response to a controversy over the meaning of "profit maximization".

Economists have long reduced the firm to a single equation "Profit equals price times quantity minus costs". This is no idle slogan that can be tossed aside by a sophisticate. This equation contains (in a rather mysterious way) the whole of the neo-classical theory of production. A firm - the neoclassical says - may not change the price because then another firm could spring up and undercut them. Cost - they tell us - are an increasing function of quantity. If the profit in an industry is greater than zero, then another firm will leap in. Therefore - the neo-classical economist says sagely - the quantity the firm may produce is exactly the quantity that sets the price to the rate of cost increase. Brilliant!

Or, perhaps, rather foolish. Nobody who studied the firm - from the unusual Thorstein Veblen to the very careful Richard Lester - could find any evidence that workers are paid "their" marginal product. This incomprehensible formulation is no straw man - it can be found in, for instance, Stigler.

Alchian throws away such nonsense - anyone would have to. Instead, Alchian rethought what it means for a firm to "maximize profit". Profit  maximization is meaningless in the presence of uncertainty. In this new interpretation, the competitive process is broken into two steps. In step one, a firm commits to producing a quantity and therefore paying a cost. In step two, consumers consume a quantity of produced goods. Firms are punished on the difference between the predicted demand for current production and the actual demand. Obviously, perfect prediction is an equilibrium of this system. Any dynamics that take you to this equilibrium will make economic analysis valid - no matter how stupid the dynamics are. Alchian gives the example of imitation dynamics (which are just about as 'dumb' as dynamics can get), Richard Day gave a more careful example which includes feedback.

Alchian's theory makes sense of Frank Knight's speculations on the nature of entrepreneurial rent - if a man has the local knowledge to better forecast than his sister, then that man may charge her for his time. But more importantly, they give a comprehensible connection between the Neo-Classical theory and the screaming torrent we call reality - it is the equilibrium theory.

However, Alchian's system isn't quite complete. He doesn't have a clear idea of what it is that punishes the badly predicting firms. Only later would Gary Becker explain correctly what Alchian meant to say - resource constraints on the firm were the whip punishing firms that missed their production quota. In other words - marginalism of the firm was just marginalism, scarce resources = scarce resources. Nothing else is involved.

Though very intellectually satisfying, this is harmful to the idea of a positive theory of the firm. Alchian's ideas apply equally to a monopolistically competitive firm, a monopolist and a firm in a highly competitive market. Luckily, we now understand the mathematics of evolution much better than Alchian and have done much work in the area. For a tour I recommend Bowles & Gintis's fascinating speculations, John Sutton's bounds approach to monopolistic competition and any paper with Tit For Tat in the title.

Incidentally, Alchian was one of the first people to play the iterated prisoner's dilemma. Alchian, unsurprisingly, played ungenerously.

"The year before the H-bomb was successfully created ... we in the economics division at RAND were curious as to what the essential metal was—lithium, beryllium, thorium, or some other. The engineers and physicists wouldn’t tell us economists, quite properly, given the security restrictions. So I told them I would find out. I read the U.S. Department of Commerce Year Book to see which firms made which of the possible ingredients. For the last six months of the year prior to the successful test of the bomb, I traced the stock prices of those firms. I used no inside information. Lo and behold! One firm’s stock prices rose, as best I can recall, from about $2 or$3 per share in August to about \$13 per share in December. It was the Lithium Corp. of America. In January, I wrote and circulated within RAND a memorandum titled 'The Stock Market Speaks'. Two days later I was told to withdraw it. The bomb was tested successfully in February, and thereafter the stock price stabilized."

- Armen Alchian inventing the Event Study

"We assert: 'All prices are Martingales.' And we conjecture a second proposition: 'No quantity variables are Martingales.'"

- Armen Alchian on flexible prices

"Nor do I find it warranted to call the stock market 'efficient' any in pertinent sense just because the present price is an unbiased estimate of the forthcoming price. I would rather call it unbiased."

-Armen Alchian on the efficient markets hypothesis

Because of his excellent writing, Alchian is often considered a 'literary' economist. But Alchian was capable of doing math. At RAND, Alchian wrote a paper (already linked) making sense of the notion of cost in a production process - cost is denoted in units of equity - this is the secret key between Alchian's analysis and the Becker analysis. Alchian kept a long interest in how the stock market moved information - one of his students, William Sharpe, got a Nobel Prize for stock market stuff.

But if I had to choose an Alchian paper with math, I would point to his more philosophical 1974 paper on the notion of a martingale as a sign of the depth of his thinking on these issues. This paper is not just about the idea of an unbiased random variable, but on the notion of what it means for relative prices to be flexible. He gives a picture of some imaginary time series that have clear patterns, but whose relative prices give no information.  An enormous amount of economics is dedicated to thinking through these kinds of models - all Tom Sargent's work for one.

The strict separation between price martingales and quantity martingales explains why neo-classicals placed such an emphasis on "price flexibility" - why the theorists like Stigler have such a horror of a price floor. If prices are a martingale, then their shifts don't disturb the underlying balance of goods and services - the scarce resources stay where they are. There is no unemployment, not really. The only "unemployment" comes from information frictions that cause the prices to not quite be martingales- they'd be "sticky".

Now, obviously, there is more unemployment that can be found from information friction. If it was just information friction, then unemployed workers could just reduce their so-called "reservation wage" (the lowest amount they'd work for, more or less) and always find a job. But recessions are real, and in a recession reducing the reservation wages raises unemployment. The price signals are confused.

Even when not in recession, unemployment is not just due to information frictions. Because unemployment is unpleasant, firms  can use it as a punishment for shirking workers. If workers are willing to work for less, this tool becomes less effective and unemployment must rise. Firms and workers can work out labour/"leisure" trade-off through the quantity channel too.

So Alchian misses important pieces of the market in his assumption that prices are martingales and quantities are not. With all due respect to Hayek, information is sent through the quantity channel, even if that channel is noisier than the price channel. But Alchian (and Hayek) deserve attention for putting the problem so clearly. Alchian deserves more respect, in my opinion, because he was less dogmatic on this issue - he cited and spread papers such as this one that featured "involuntary unemployment".

"Before condemning violence (physical force) as a means of social control, note that its threatened or actual use is widely practiced and respected—at least when applied successfully on a national scale. Julius Caesar conquered Gaul and was honored by the Romans; had he simply roughed up the local residents, he would have been damned as a gangster. Alexander the Great, who conquered the Near East, was not regarded by the Greeks as a ruffian, nor was Charlemagne after he conquered Europe. Europeans acquired and divided—and redivided—America by force. Lenin is not regarded in Russia as a subversive. Nor is Spain’s Franco, Cuba’s Castro, Nigeria’s Gowon, Uganda’s Amin, China’s Mao, our George Washington."

- Armen Alchian on alternative means of governance

"Incentives are the prizes in the game of life-the goals individuals seek - the carrots. Through the ages of Tutankhamen, Alexander, Caesar, Louis XIV, and the Atom, they have remained the same. Men want, and have always wanted, exorbitant wealth, tyrannical power, idolatrous prestige, lavish consumption, and undisciplined leisure. ... What does explain the disparities? Differences in the relations between costs and goals."

- Armen Alchian on the utility function

There is a mysticism to the work of many economists. An economist speaks of optimality more often than Dr Pangloss and stability in a world in constant torrent. One reason that I like the works of Alchain is that there is a sort of reality to them. They are philosophical, even when they are technical. "Cost is a choice" - dull. "Cost is a choice in different forms of equity" - interesting! "Firms, of course, evolve" - hogshew. "Selection on firms is on their ability to forecast" - neat!

Another example: Alchian's work Why Money? is a economic philosophy gem, deriving the existence of money from the network of trade. A good will become money if the costs of identifying the quality of a good are low for everyone - that's why kings put their stamp on it! This is obviously an important work and one that opened the door to fundamental new research.

I said I wouldn't quote Alchian on macro, but I can't resist one:

"Why would a cut in money wages provoke a different response than if the price level rose relative to wages – when both would amount to the same change in relative prices, but differ only in the money price level? Almost  everyone thought Keynes presumed a money wage illusion. However, an answer more respectful of Keynes is available. The price level rise conveys different information."

This is one of those things that is so obvious after it is pointed out to you. In many economists there is the magical thought - it all depends on the relative price level. But Alchian is more careful - prices of complex items adjust more slowly than prices of simple items and people are complex. In a crude Keyensian system,  a cut in money wages ahead of the business cycle would signal bad times and depress spending - a raise of the price level (with wages lagging behind) would signal good times and encourage spending. In the Classical & Neo-Classical system, this signalling does not occur, only the relative price levels matter. Yes, there really was a Keynesian Revolution.

This observation makes the difference between neoclassical and Keynesian economics at least partly testable! I know that Alchian has attempted these tests and generally hasn't found the desired lags. And I know many economists - such as Andolfatto - have pursued this avenue more deeply. But I don't know much beyond that and will therefore be silent.