A reason to reread Olivier Blanchard and Stanley Fischer’s Lectures in Macroeconomics (1989) is an early statement of what New Keynesians (NK) hoped to accomplish during the 30-year macro methodology war. We all know the short answer. NK economists were to be the keepers of the Keynesian tradition of stabilization-relevance as the academy’s mainstream rules of engagement increasingly, and properly, insisted on coherence between micro and macro theories. In particular, the NK mission was to microfound the causal link from adverse nominal demand disturbances to involuntary job loss. Over time, however, the interesting story became why and how most NK theorists abandoned that quest. While continuing to accept the need for model coherence, they now proactively ignore the forced job loss that most interested Keynes. There is much to be learned from the troubling NK descent into a Ptolemaic quagmire of stabilization irrelevance.
Early on in Lectures, B&F concede that Keynesian thinking, especially the Neoclassical Synthesis, was “in theoretical crisis.” They responded by organizing their text around a benchmark neoclassical model featuring “optimizing individuals and competitive markets.” However, while research on microfounding wage and price rigidities evident in the data was ongoing, B&F argued that Keynesian shortcut assumptions remained necessary to inform sound policy advice, the alternative being “a harmful utopia that leaves the real world to charlatans.” (pp.27-28) Lectures includes non-microfounded models that are “the workhorses of applied macroeconomics,” notably including Hicks’s IS-LM interpretation of The General Theory. B&F position their text to occupy a difficult middle ground in an increasingly contentious debate. But, given the criticality of stabilization relevance and the threat of abandoning policy advice to the charlatans, they judged their apologetic version of the Keynesian Neoclassical Synthesis to embody the best available analysis. “It would clearly be better for economists to have an all-purpose model, derived explicitly from microfoundations and embodying all relevant imperfections, to analyze all issues in macroeconomics (or perhaps all issues in economics). We’re not quite there yet.” (p.505)
The B&F text sets the table for a number of questions today. When did many NK theorists abandon the B&F middle ground? When did they give up on stabilization relevance? When did they decide to join the charlatans and pursue careers that fundamentally mislead policymakers? The answer, I think, is when prominent NK theorists quietly concluded that the sought-after super friction, i.e., the coherent market imperfection capable of suppressing wage recontracting, does not in fact exist.
Clearly, at some point the quest for the super friction no longer appeared to be the stuff of rewarding careers. The signal capitulation was the broad NK acceptance and heavy use of the now ubiquitous S/M/B model class. In order to work with some coherent model that has something to do with unemployment, NK theorists have broadly agreed to pretend that they can squeeze realistic-looking cyclicality out of search theory’s voluntary joblessness. Today nobody appears much bothered by the fact that job separation in recessions is, in fact, overwhelmingly involuntary. Everybody knows that, but macroeconomists do not care. What a mess. The huge investment of human capital in S/M/B modeling in all its variations appears to have helped trap many NK theorists in a conundrum where the primary objective of research is a Ptolemaic defense of the mainstream coherent market-centric DSGE model class that will never be stabilization-relevant.
The GEM Project provides, happily, an opportunity for redemption, another shot at “an all-purpose model” that is both coherent and stabilization relevant. The Project features meaningful wage rigidity, manifest in downward nominal inflexibility over the stationary business cycle as well as chronic, time-varying wage rents, that is derived from axiomatic model primitives. As a result, macroeconomics can be done the way we know, deep down, it should be done. Its only analytic cost is that NK theorists must give up restricting rational, price-mediated exchange to the marketplace – a constraint that has been nonintuitive since the advent of the Second Industrial Revolution. The opportunity presented by the GEM Project, in providing the sort of model that Lectures yearned for, is a test of character for NK theorists. I suspect that some remain sufficiently interested in the fundamental questions and focus of Keynesian economics to make the effort to investigate whether the coherent GEM model can indeed causally link nominal demand disturbances and involuntary job loss. Others, perhaps fearing for the human and reputational capital they have tied up in S/M/B modeling, will hope the Project goes away. What choices Blanchard and Fischer make will be interesting.
Blog Type: New Keynesians Saint Joseph, Michigan