“The Current Models Have Nothing to Say”
By Robert Higgs • Thursday March 22, 2012 3:23 PM PDT • 5 Comments
In the 2011 annual report of the Federal Reserve Bank of Dallas’s Globalization and Monetary Policy Institute, we find a report on the 10th annual Advances in Econometrics Conference, sponsored jointly by the institute and the department of economics at Southern Methodist University. This conference focused on dynamic stochastic general-equilibrium (DSGE) modeling.
As the report notes,
DSGE models have become an essential part of economists’ empirical toolkit in recent years. These models have their origins in the seminal contributions of Kydland and Prescott (1982) and Long and Plosser (1983), which revolutionized empirical econometrics. . . . Subsequent work by Christiano, Eichenbaum and Evans (2005) and Smets and Wouters (2007) laid the foundations for these models to become the workhorse frameworks for policy analysis in most central banks.
After describing the papers presented at the conference, the report concludes:
The conference confirmed that New Keynesian DSGE models are useful tools for understanding business fluctuations in closed and open economies and also for thinking about important monetary policy questions. However, the current models have nothing to say about how the financial system impacts the real economy; given the events of the past few years, that must now be a top priority for research. Also, to date, there have been relatively few attempts to develop open-economy versions of these models . . . . With globalization defining the environment in which monetary policy is now made, models that take seriously international trade and financial linkages will be crucial to the policy process. (emphasis added).
For mainstream economists, reports of this sort are commonplace. They nod in agreement: yes, we’ve got some fine models, but much work remains to be done to develop them and, perhaps someday, to make them applicable to economic reality. No rush, though. We need only keep the grant money flowing to the researchers and their institutions.
People outside this protected enclave of idiot savants, however, might well be astonished to learn that “the workhorse frameworks for policy analysis in most central banks . . . have nothing to say about how the financial system impacts the real economy” and in most versions are inapplicable to the analysis of events in a globalized economy.
Some readers may recall how, a few years ago, the recently announced Nobel laureates in economics, having been asked by journalists for their views on various pressing issues related to the current economic crisis, seemed to be struck dumb by the questions. They really had nothing to say, because their work as economists pertains to building models that make no genuine contact with economic reality. Such has been the ultimate result of the methodological revolution that began in economics between the world wars and swept the field in the 1950s, transforming mainstream economics, especially macroeconomics, from a serious effort to understand economic reality to an extended exercise in mathematical diddling with only a nominal connection to economic reality.
Small wonder that when the economic crisis hit in 2008, mainstream economists seemed dumbfounded by its occurrence and at a loss as to what might be done about it—except to retreat to the basic Keynesian nostrums that elementary textbooks had been serving up to beginning students since Paul Samuelson’s revolutionary text appeared in 1948. A trainload of advanced research in macroeconomics seemed, in the crunch, to avail these mathematical wizards nothing at all—which was to be expected, because macroeconomics had long since severed its connections with economic reality, its practitioners preferring to monkey with models, all the while telling themselves and their students that real scientific understanding and optimal economic policies lay just beyond the research horizon.
What a waste of time and talent.