The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel 2018 was awarded October 8 to Paul Romer, Professor at New York University, “for integrating technological innovations into long-run macroeconomic analysis,” specifically for what is sometimes called “New Growth Theory.” He shared the Prize with Professor William Nordhaus, Professor at Yale University (and a Bonesman), who received it “for integrating climate change into long-run macroeconomic analysis.” The Romer Prize had been predicted for a long time.
Predictably, commentators, fellow economists, pundits etc. have been fawning over the “discoveries” of Romer and Nordhaus (including at least one major contemporary Austrian who apparently see no problems with treating “knowledge” as a homogenous factor of production). Nothing new here, and nothing wrong with celebrating major scientific advance. This is worthwhile, right, and of course, great fodder for the tweeters and the bloggers. But, this year’s Nobel is particularly illustrative of what gets rewarded in economics, and how this does not exactly conform with more intuitive notions of scientific progress.
If asked about what we mean by “scientific progress,” the proverbial (Wo)Man in the Street would probably respond something like, “Pointing to new phenomena that no one so far has been aware of,” “Explaining phenomena in a fundamentally novel manner” or similar.
Incidentally, many philosophers would agree. For example, to many scientists Karl Popper is a standard reference and many scientists will be able to not only invoke but also explain the “falsification criterion” (only theories that can be conceivably be rejected by facts can be deemed “scientific”). Popper is of course all about putting forward new predictions. His follower Imre Lakatos would agree: Scientific advance is about theoretically formulating and empirically corroborating “novel facts”—new insights in the world that previous generation of scientists had missed.
Economists are familiar with a bastardized version of Popperianism, namely Milton Friedman’s famous emphasis on economic theories as tools of prediction. So, these ideas most certainly have currency in economics. Do economists practice what (many of them) preach, in particular that scientific progress amounts to putting forward novel facts and/or novel explanatory mechanisms in the form of bold, new hypotheses? The late Mark Blaug repeatedly pointed out that they don’t. The recent Nobel may confirm this.
Consider how the Prize Committee justifies the Prize to Romer :
Romer demonstrates how knowledge can function as a driver of long-term economic growth. When annual economic growth of a few per cent accumulates over decades, it transforms people’s lives. Previous macroeconomic research had emphasised technological innovation as the primary driver of economic growth, but had not modelled how economic decisions and market conditions determine the creation of new technologies. Paul Romer solved this problem by demonstrating how economic forces govern the willingness of firms to produce new ideas and innovations.
There is absolutely no doubt that Romer is an extremely clever modeler and economist (which these days is often the same thing). But, notice the wording in the press statement: “previous research” was lacking because it had not “modeled” endogenous technical change. The press statement does not say that Romer discovered anything new per se.
Many years ago I ruminated on the success among research economists of the new growth theory that Romer is said to have created. Mymain point was that the essence of the explanatory mechanisms proposed and modeled by Romer and other new growth theorists has been known since at least the 1920s work of Allyn Young on increasing returns and externalities in the growth process. Kenneth Arrow, another Nobel laureate, highlighted the role of learning by doing in growth. Evolutionary economist Richard Nelson has argued that much of the content of the new growth theory can be found in the 1950s work of Moses Abramovitz. When economists talk about Romer’s (and others’) contributions as the “New Growth Theory,” the “old” theory implicitly referred to is very specifically the theory of Robert Solow from the mid-1950s, which not only had novel implications but also was a formal theory. However, the “old theory” does not include the many thoughts on the growth process that economists have offered (almost) since Adam Smith.
For example, Mises and Hayek, and even earlier, Carl Menger pointed to scientific progress and general advance of knowledge as strongly intertwined with the process of growth. Moreover, they, and many early writers on the growth process, stressed the importance of institutions—private property rights, the rule of law, and the price system—in promoting growth, and they emphasized that the advance of knowledge is largely determined by such institutions. Any reader of, e.g., Mises’ Socialism will recognize this theme. Nobel Laureate Douglass North thought these aspects were so largely absent in the New Growth Theory that it was begging all the important questions.
Of course, those who are familiar with modern economics know that “progress” mainly means what may be called “heuristic progress”—better, more elegant, and cleaner models that, for example, succeed in endogenizing what has hitherto been taken to be endogenous. As I wrote in 1998: “This, then, accounts for the success of the [New Growth Theory]: the advent of the NGT meant that a recalcitrant … problem—endogenizing technological change in the context of economic growth—could be treated in formal terms, and thus assimilated into neoclassical economics.”
To be sure, such exercises can have very significant value, in making some assumptions explicit, illustrating particular mechanisms, and the like. (Not always!) But, putting existing ideas in a different language does not, by itself, constitute a huge advance in scientific understanding. For this we need new, causal explanations for phenomena (such as Menger’s explanation of price formation, Mises’s critique of socialist planning, Hayek’s advances of business-cycle theory, and so on). For example, Mises’ famous critique involved the prediction that socialism cannot work for very specific reasons to do with the pricing and allocation of capital goods. There is a huge, and perhaps categorical, difference between such fundamental predictions and finding new ways to put existing, if informal insights into established models, such as neoclassical general equilibrium.
