This post looks the ambitions and evolution of Wendy Carlin and David Soskice’s Macroeconomics, the first edition of which was published in 2006 by Oxford University Press to immediate acclaim, and the second in 2015. It is an instructive story.
In the first edition, Carlin and Soskice (C-S) declare their “core audience is undergraduates”, whom they promise to introduce to “the major changes that have taken place in modern macroeconomics…. You will end up with a model that you can use yourself to understand macroeconomic behaviour and policy issues in the real world.” That objective, which is shared by the second edition, is exactly right for an undergraduate textbook. In pursuit of that goal, however, the two books fundamentally differ. In 2006, useful macro modeling refers to the New Keynesian (NK) market-centric dynamic stochastic general-equilibrium (DSGE) theory. C-S adopt the mainstream view that rigorous New Keynesian thinking had properly replaced Early Keynesian (EK) analysis. By 2015 and their apparently traumatic encounter with the intervening 2008-09 extreme instability, C-S make an about-face and throw the consensus NK model under the bus. Their new go-to theory that undergraduates can use “to understand macroeconomic behaviour and policy issues in the real world” reverts to a version of the EK 3-equation model.
First-edition Macroeconomics was written to align macroeconomics taught to undergraduates more closely with what has been taught, for a long time, to graduate students. In 2006, C-S objected to the former getting a steady diet of EK analysis with its keystone assumption of short-term downward wage rigidity, reversing neoclassical real-to-nominal causation, and its consequent dominant emphasis on nominal demand. The EK macro policy that needed to be deemphasized is focused on the discretionary management of total spending, used to ameliorate periodic fluctuations in total employment and output. Also requiring deemphasis in the modern stabilization narrative, contrary to Keynes’s intentions, is involuntary job loss. C-S argue that graduate students had long been provided a superior microfounded model. The NK DSGE approach directed advanced instruction to the derivation of its microfoundations, the specification/testing of real shocks, and the management of price inflation. NK theorists’ understanding of “macroeconomic behaviour and policy issues in the real world” centrally includes pushing rational nominal wage rigidity, forced job loss, and total spending from center stage.
It is no surprise that a founder of the GEM Project applauds the second edition’s rejection of market-centric DSGE modeling. C-S summarize their dissatisfaction with New Keynesian macro theory in the very last passage of their text (p. 611): “The NK DSGE model was widely adopted as the workhorse model of the research departments of central banks before the financial crisis. It is a common misconception that the NK DSGE model is simply a more complicated version of the basic [EK] 3-equation model. There are significant differences between the two models. The NK DSGE model is grounded in the RBC model, so the intertemporal substitution of labour is the mechanism driving business cycles and there is no involuntary unemployment. In addition, all agents are forward-looking and ‘solve the model’, which leads to jumps in inflation. In contrast, in the 3-equation model there are incomplete contracts and involuntary unemployment at equilibrium, and cycles are driven by the response of price-setting firms … to demand and supply shocks, wage and price inflation respond to the output gaps that arise. In this model, the central bank is forward looking and has to adjust interest rates to set the economy on a stable path to equilibrium. Introducing some forward-looking behavior of private sector agents to the 3-equation model dampens cyclical fluctuations, but does not change the nature of the propagation mechanisms or of the role of the central bank. The sharp differences between the NK model and the 3-equation model highlighted in this chapter are somewhat blurred in the NK DSGE models used in central banks where a large number of sources of persistence are introduced.”
The next two posts look more closely at the C-S case against the mainstream NK DSGE model and contrast their replacement (EK) 3-equation model with GEM analysis that generalize rational exchange from the marketplace to the information-challenged workplace. That two more posts continue to compare and assess the three macro frameworks, each purporting to be useful, allows this post to close with what I believe to be the critical flaw in the C-S revival of EK thinking. That defect is illustrated by C-S failure to make clear that their go-to 3-equation model, in its explanation of crucial macro facts such as involuntary layoffs, is constructed on deeply irrational labor-pricing behavior. They simply ignore the fact that no version of ubiquitous staggered wage setting is able to rationally suppress recontracting. Their macroeconomics is Early Keynesianism without the school’s saving grace of admitting its inherent irrationality. Nor do C-S adopt the NK top-priority emphasis on research seeking to microfound meaningful wage rigidity (MWR).
That’s it. The C-S critical flaw is the absence of a top-of-the-agenda priority assigned to microfounding an evidence-consistent model of MWR, defined by its capacity to rationally suppress labor-price recontracting. Their downplaying of that research is shared with New Keynesians. The necessity of rationality in wage modeling, its indispensable role in getting labor pricing and therefore stabilization theory right as well as providing powerful guidance on useful macro model-building, has long been lost in modern macro thinking.
Here’s what’s really disappointing. C-S apparently do not realize that their second edition could have been useful in understanding macro behavior as well as rooted in optimization and equilibrium (the fundamental tenets of economic theory) if they had exercised more discernment in summarizing the existing macro literature. More insight is especially needed with respect to the existing efficiency-wage research on rationally suppressing wage recontracting. Closer attention to genuinely significant advances in microfounding MWR could have led to a macro text that provides a stabilization-relevant EK model motivated by optimizing behavior that is organized by general decision-rule equilibrium and that explains the most significant evidence, including the Great Recession and its aftermath.
Blog Type: New Keynesians Chicago, Illinois