I've been consistent in my criticisms on this blog of attempts by Paul Krugman to revive the IS-LM framework. That's not because I'm opposed to IS-LM as it appeared in its earliest incarnations; the papers by John Hicks and Alvin Hansen. It's because of the bastardization of the Keynesian agenda by what my friend and teacher David Laidler referred to as North American Keynesianism. In my view, articulated in Prosperity for All, macroeconomics went off the rails in 1955 when Samuelson introduced the neoclassical synthesis in the third edition of his textbook, Economics: An Introductory Analysis. (See Pearce and Hoover for a great discussion of the influence of Samuelson's text and my book How the Economy Works).
The IS-LM model is an effective way of capturing some empirical regularities in a graphical apparatus. But it is not a complete theory of macroeconomics in the sense that we mean by that term today. When Hicks wrote the famous paper that became IS-LM, he had already written a pathbreaking book, Value and Capital, that is, in many ways, a much much better book than Keynes’ General Theory. Value and Capital used a technique, temporary equilibrium theory, that presents a complete dynamic model of the macroeconomy.
In Value and Capital, time proceeds in a series of “weeks”. Each week, people meet in a market. An auctioneer calls out prices and only stops when the demands and supplies of all commodities are equal. At that point trade takes place and people trot off home to the family farm to produce goods for the next week’s meeting. At each market meeting, the demands and supplies that people announce are functions of their beliefs about future prices. And those beliefs may or may not turn out to be correct.
When Hicks read the first draft of the General Theory, he had a crisis of confidence and figured that everything he'd spent his life working on was wrong (see Michel De Vroey's piece on this point). He seized on Keynes’ idea that, at each market meeting, people trade with each other before the auctioneer has finished his job. The result is involuntary unemployment and he formalized that idea by developing what we later came to call the IS-LM model.
The Keynesian economics of the General Theory is static. It purports to explain how employment, GDP and the interest rate are determined at one weekly meeting, taking the price level as fixed. Modern macroeconomics is dynamic. It purports to explain how employment, GDP, the interest rate and the price level are determined in a sequence of weekly meetings. To knit together the temporary one-week Keynesian equilibria, Samuelson, in the new-classical synthesis, used the Phillips curve, which he saw as a price adjustment mechanism in which the wage adjusts in response to an excess demand or supply of labor. This was the biggest mistake in the history of macroeconomic thought and we are still suffering the consequences as central banks work with false ideas and broken models.
In Prosperity for All I articulate the evolution of an alternative research agenda. I argue that it is beliefs that are sticky: not prices. At each weekly meeting, the auctioneer finishes his job. The demands and supplies of all goods are equal and all markets clear; including the labor market. But the labor market is a search market, not an auction market, and there are many different ways in which it can clear (see my EJ piece on how this works). Labor market equilibrium is pinned down by beliefs about future prices and, for every belief, there is a different Pareto inefficient market-clearing unemployment rate.
The differences of this theory from all of modern macro, both classical and New-Keynesian, are profound. In my view, high involuntary unemployment is an equilibrium phenomenon. A market economy can get stuck in a Pareto inefficient equilibrium with high unemployment forever. It is the job of government to design political institutions that provide the equilibrating mechanisms that are missing from laissez-faire market economies.
You might think that the above paragraphs would make me an uber-Keynesian. Surely I should be riding in on my white horse unfurling the banner of fiscal intervention to save capitalism from itself. Not so fast. Although my work provides a foundation to the Keynesian theory of aggregate supply: I am skeptical of Keynes' views on aggregate demand. Much more on this in a future post.
I looked at your EJ piece, which I liked a lot.
ReplyDeleteHowever, you make the comment in section 8: "If a firm were to offer less than the competitive wage, it would receive no applications." I find that quite difficult to square with the idea of a persistent state of involuntary unemployment. Presumably what is going on here is that job-seekers are only applying for jobs at the full competitive wage even though they anticipate that they may not get those jobs. Do you not need some more story here as to why they would not rationally also apply for lower paying jobs, just in case?
Glad you liked the piece
ReplyDeleteDo I need a story? Perhaps. Stories are, after all, much of what our profession does best. But the equilibrium concept that I have proposed is, in that respect, no different from the competitive equilibrium concept. In competitive equilibrium, people take prices as given, they submit demand and supply schedules, and the market finds a consistent solution. In my model, households and firms take wages, prices and hiring probabilities as given, they submit demand and supply schedules, and the market fins a consistent solution. Neither approach has much to say about out of equilibrium behavior.
So do I need a story. Yes, but I would prefer not to be held to a standard that is higher than existing competitive theories.