The title is not a typo. Brad DeLong reproduces a list of papers that ten years ago he thought were on the frontier of macroeconomics. Pointer from Mark Thoma. My take on the list is that there is a strong negative correlation between the probability that the ideas are correct and the probability that they are relevant.
Some random thoughts about the state of macroeconomics:
1. Any post that contains a sentence “This chart proves that. . .” doesn’t.
2. Macroeconomists should be much more daunted by measurement issues than they are. I think that we have reasonably good ways of counting the number of people who do paid market work. But that is a far cry from knowing “labor input,” because (a) skills are very heterogeneous and (b) as Garett Jones famously tweeted, most workers are not making widgets but instead are building organizational capital. Output has become increasingly difficult to measure. In the case of goods, quality change has sped up, putting more pressure on statisticians to rely on imputations. And in the case of important services, including education, finance, and medical services, we have almost no conceptual idea for measuring output. We do not know how to factor into our statistics the increased diversity in consumption baskets across individuals. All this casts doubt on our measures of inflation, productivity, and real wages.
3. As of the early 1930s, many economists and commentators thought that the capitalist system had broken down. They saw the decentralized market process as no long working effectively to organize economic activity. See Katznelson, or even better read chapter two of Leuchtenburg’s The FDR years. The biggest intellectual influence at the time was nostalgia for the government planning that took over when the U.S. entered the first World War.
In contrast, Keynes blamed the Depression on what he called a drop in aggregate demand, and Milton Friedman blamed it on a contraction of the money supply. In fact, Keynes and Friedman led the profession down a false path. The pre-Keynesian diagnosis was more apt. I just disagree that central planning was the best solution, although I think it is fair to concede that when you have unemployment rates over 15 percent you are giving central planners a decent shot at doing something right.
4. The Monetary Walrasians (Patinkin and everything that followed) wasted our time. Money does not determine nominal aggregates. It does not determine transactions. The causality runs from the desire to undertake transactions to financial institutions/technology to what people use as money.
5. Money does not determine the price level. Habits determine the price level. Consider the amnesia experiment. If everyone developed amnesia about what money prices were yesterday, then Monetary Walrasianism predicts that today prices would soon find an equilibrium determined by the quantity of money. In fact, my best guess is that people would find money to be worthless in such a setting, and they would resort to barter until a new standard of value emerged.
6. The theory of rational expectations is a waste of time. The simple model of employment fluctuations as arising from errors in aggregate expectations is wrong. And the underlying principle of rational expectations, that everyone as the identical information, is anti-Hayekian, and not in a good way.
In short, almost nothing that gets taught in undergraduate macro is correct. And graduate macro is worse.