Sound Model-Building Guidance

Two weeks ago the GEM Blog took Olivier Blanchard to the woodshed for his assessment (in 2008) that the state of NK macroeconomics is “good”. A week ago, it went into some detail about the extraordinarily wrong-turn guidance from supposedly “good” mainstream macro theory. The short series ends today on a more upbeat note. This post offers an optimistic appraisal of the future of macro theory – enthusiasm that is rooted in the substantial, far-reaching improvements that have become available from the Project’s generalized-exchange modeling.

Early and New Keynesians both purport to provide valuable insight into modern highly specialized economies. Each school constructs its theory within the familiar neoclassical market-centric general-equilibrium framework. Seeking stabilization-relevancy, each must resort to irrational assumptions of nominal wage rigidity to suppress wage recontracting. Over time, one outcome of that shared practice stands out. The absence of rational-behavior, evidence-consistent labor pricing has deprived theorists of crucial guidance in their attempts to usefully model cyclical and  trend aggregate behavior. Macro analysis became more complex after the advent of complex bureaucratic firms in the Second Industrial Revolution. The new corporate forms enabled a jump in labor specialization and productivity that cannot be effectively captured by general market equilibrium. Inadequate market-centric roadmaps, especially with respect to optimizing exchange occurring inside large-scale establishments, have made macro theorists vulnerable to debilitating wrong turns. In an example cited last week, microfounded MWR would have steered macroeconomists clear of the extremely damaging wrong turn that resulted in the persistent, futile attempt to use labor-search theory to explain forced layoffs in the aftermath of adverse demand disturbances.

I believe that the crucial EK-NK difference is rooted in hubris. The Early Keynesians emphasized that their Neoclassical Synthesis is unfinished macroeconomics and, as a result, put microfounding MWR at the top of their research agenda. New Keynesians, by contrast, see their New Neoclassical Synthesis and its attendant dynamic stochastic general-market-equilibrium modeling as fundamentally finished theory. No core research remains to be done; none would be useful. The presumption of settled theory has effectively erased the quest for MWR microfoundations from the academy’s research agenda.

Erasure has turned out to be a huge mistake. It is, I believe, the biggest blunder produced by the academy’s collective hubris in simply ignoring the enormous mismatch between the NK market-centric, general-equilibrium theory and the range of evidence on aggregate behavior. The following list, summarizing examples of debilitating consequences resulting from the failure to inform macro model-building with microfounded MWR, provides substance to that claim. The inventory, limited to ten, is powerful testimony that the careful exploration of alternatives to market-centricity is a worthwhile direction of research:

• MWR, defined by its capacity to rationally motivate downward nominal labor-price rigidity over stationary business cycles as well as chronic wage rents, itself leads off. A central Keynesian problem has long been that the rational suppression of wage recontracting cannot exist in market-centric general-equilibrium modeling. There is no super market friction that does that. Rejecting market-centricity, generalized-exchange theory is able to derive MWR consistent with optimizing employer-employee behavior organized by continuous general decision-rule equilibrium.
• Absent the suppression of wage recontracting, the DSGME model class cannot accommodate  involuntary job loss and, as a result, can only produce the mildest of recessions. It is unsurprising that mainstream macro modeling was ignored by badly misled, and consequently grumpy, stabilization authorities in 2008-09.
• Chronic wage rents are now understood to be a powerful characteristic of information-challenged workplaces. Their absence in NK modeling has harmed efforts to explain a great deal of important evidence. Wage rents are, in one example, necessary to understand unemployment persistence in recessions and recoveries as well as the high frequency of laid-off workers who are recalled.
• Microfounded MWR demonstrates that, in highly specialized economies, textbook Wicksell-Wicksteed factor-income distribution is a special case that is unsuitable for general use. The rejection revives the existence of pure profit. Practitioners know that investment spending is largely motivated by the expectation of pure profit. Assigning the primary role instead to interest rates reflects a recklessly misleading wrong turn by NK theorists. It is even more misleading to pretend that interest rates are the primary driver of consumption. The GEM Project uses microfounded MWR to identify the dominant determinant to be household income, reviving the once-famous Barro-Grossman (1971, 1976) FWGE analysis.
• The GEM Project also demonstrates that Lucas’s rational-expectations Phillips curve is constructed on irrational employer behavior. His equation is now understood to be a special, largely irrelevant, case. In a related innovation, inflation catch-up in periodic wage determination – a practice that aligns with actual behavior – has been microfounded in generalized-exchange modeling. Macro theorists can now happily join practitioners in ignoring problematic inflation forecasts in labor-pricing.
• Microfounded MWR identifies the need to separate nominal-demand disturbances into their stationary and nonstationary components. The latter motivates the GEM Project’s continuous general-decision-rule-equilibrium modeling of the Great Recession. To conserve space, that important extreme-instability model with its emphasis on macro-shock propagation is not elaborated upon in this post. (See Annable and Schechter (2015).)
• Absent close analysis of information-challenged workplace exchange, macro theory cannot accommodate the obvious existence of profit-seeking labor practices in highly specialized, bureaucratic firms, thereby pushing aside how a substantial share of the labor force is compensated. Everybody knows that human-resource departments, always present in large bureaucratic firms, actually govern labor-management relations, including compensation. That fact goes hand-in-hand with the established message of neoclassical theory that, in the circumstances of workplace-information asymmetry, wages cannot be efficiently determined in the labor market. That message must not be, but almost always is, ignored.
• Along the same lines, rational market-centric theory can never adequately accommodate evidence-consistent labor-union modeling. Generalized-exchange theory can.
• Market-centric general equilibrium theory cannot integrate the powerful, Nobel-prize-winning, Lewis two-sector growth theory, which is recognized by specialists as our best explanation of the transition out of subsistence, into mainstream macro thinking. For the generalized-exchange theory, such integration is a piece of cake. Wouldn’t it be a good thing to be able to teach Sir Arthur’s important macro model to economics students?
• Microfounded MWR is central to the rigorous, evidence-consistent analysis of the second greatest instability crisis of the 20^th century: the stagflation decade that began in the early 1970s. My Price of Industrial Labor (1984), anticipating the subsequent work of the GEM Project, carefully explains that episode’s crucial price-wage-price spiral and provides an atypically accurate description of stagflation’s persisting market failure. Most NK theorists are unaware that stagflation subsequently fueled the 1980s “rust-belt” downsizing.

Pause and think about that list. Think about all the all the economic phenomena, essential to adequately understanding highly specialized economies, that mainstream New Keynesians are content to leave on the table. Do they really believe that involuntary unemployment is unimportant? Do they not understand that the derivation of rational wage rents greatly enhances useful explanation and policy conclusions? Don’t they understand that expectations of pure profit are crucial to investment decisions? That personal income is the primary driver of household consumption? That Human Resource Departments exist, managing systems of wage determination and work rules that respect employees’ preference for equitable treatment? That fulsome, evidence-consistent explanations of 1970s stagflation and 1980s downsizing would be good things? That a fulsome, evidence-consistent rational-behavior model of 2008-09 extreme instability, which perilously threatened depression, would be a really good thing?

Game Worth the Candle?

New Keynesian gatekeepers do not argue that generalization of exchange from the marketplace to information-challenged workplaces is incorrect. How can they? Instead, they contend that adding a second venue of rational exchange would be an upheaval in macroeconomics that would require a lot of work, especially with respect to the necessary retooling of the profession’s existing human capital. They are, in effect, arguing that the game is not worth the candle. This post is a thousand-word rebuttal to that objection.

Blog Type: New Keynesians San Miguel de Allende, Mexico