Single post

Our post-Keynesian model: Support a populist economic agenda

Discussion of macro models at a recent event occasions a somewhat self-critical look at the model that I have worked on since about 2011 and with the help of academic colleague Tai Young-Taft since late 2014—as well as a look at what to expect as the U.S. economy approaches full employment. I will warn readers here that some of the following discussion of economic models may be overly technical for some.

I reported in this recent post on my visit to Amherst, Massachusetts for an economics conference in honor of a new building for the economics department at the university in that small town. At an interesting session on heterodox macroeconomics and its future, professor and dynamic modeling pro Peter Skott expressed concern about models that leave out processes in labor markets that inherently limit growth in a capitalist economy. He argues that such economies are labor-constrained over the long run; pressures on labor markets that in Skott’s view have a tendency to raise wages—a phenomenon being waited for in the current US recovery by observers such as reporters at the New York Times—a development that squeezes profits, choking off growth in the model favored by the “UMass” school across many studies and eras, as noted by several speakers. In contrast, in a Keynesian and Kaleckian model like ours, the latter forces tend to spur growth with higher wages spurring needed product demand, even in a “growth” model that looks at long time windows. Limiting factors that bring cyclical economic contraction come in the form of imposed fiscal austerity, financial instability, threshold effects of short-term weakness, changes in the psychology of markets and entrepreneurs, and more.
Considered as a mechanism, how does the Hannsgen-Young-Taft model work? Our assumptions about aggregate demand imply that near equilibrium, self-reinforcing cycles of goods demand and income push the economy outward. The intensity of these forces decreases as the economy gets further away from equilibrium. Moreover, the countercyclical policy rule (as opposed to a rule with a deficit target) acts to stabilize output around equilibrium. Since public and private production targets are expressed relative to capital stock, the policy rule implies that public wage spending grows as capital accumulates, insuring rough balance between sectors and a steadily rising impetus to demand. The factors that keep the economy from growing faster than it does include the fiscal policy targets, which allow for resource slack (less-than-full employment of capital) and the appearance of a capacity utilization target in the investment function. Policymakers choose subpar output targets because policymakers believe inflation would increase, a belief that is to some extent true in versions of our model with nonfixed nominal wages and prices. Given these forces based on utilization of the capital stock, labor supply poses no additional impediment to growth in our basic model. In the largest version of the model, the labor force is added as an additional state variable and policymakers use an unemployment rate target rather than a capacity utilization target. Moreover, the forces potentially generating inflation in the model include markup inflation from changes in monopoly power, not just wage pressures. The growth-restraining effects of the capacity utilization target prevent the model from generating implausible pathways in which producers perpetually add new machines and equipment in the face of ever-rising excess productive capacity—a problem illuminated in the writings of early British Keynesian Roy Harrod and hammered in the literature by Skott, UMass Ph.D. graduate Soon Ryoo, and others.

These properties of the model should help it survive critiques that apply to the basic Kaleckian model based on allegations of unaccounted-for destabilizing forces. What advantages does it have relative to basic models that in some cases might share this strength in withstanding common critiques? Here is a partial list:

(1) It has the stock-flow-consistent property of accounting for all stocks and flows in a leak-proof set of accounting relations (though the most basic version has only government liabilities)
(2) Given that it models the markup and nominal wage rather than the wage and price level, it offers a possibility of a Keynesian closure in which fixed proportions between wages and prices prevail, barring changes in market power that cause the markup to change. In other words, in the usual situation in which unemployment exists, efforts to reduce the real wage fail because changes in the markup are determined by goods-market power and real-wage norms, rather than as the joint and incidental byproduct of demand pressures in labor and goods markets. Recent analytical advances based on nonlinear 2D Goodwin framework (capacity utilization and the wage share) are an example of the latter “wage-price module” approach.
(3) Two papers introduce shifting profitability, default, and financial market expectations as emphasized in G. L. S. Shackle’s “Keynesian kaleidics” and Minsky’s Financial Fragility Hypothesis. In our simplest “Shackle model,” the driving process allows for long-range dependence and infinite variance in the increments—properties common in financial data and some macro time series. In our simplest “Minsky model,” the probabilities of shifts are endogenous to economic variables that indicate financial fragility. Shifting expectational terms formalize procedures widely used in stock-flow consistent economics of observing impacts of exogenously imposed one-time shocks, a less satisfactory approach for generating a typical solution pathway for a growth-and-distribution model. Most basic models in the growth-and-distribution genre lack any role whatsoever for financial dynamics.
(4) In contrast to econometric “models that work” for nonstationary economic time series, ours is an analytical model that represents causal forces and allows one to imagine counterfactual situations, rather than to fit the data. Our simulation approach allows us to be more realistic about obstacles to estimation posed by the properties of the data, including nonlinearity and fat-tailedness. Moreover, in addition to illuminating actual forces moving the economy, we are able to think about counterfactual policy scenarios. Finally, in contrast to time series econometric approaches in general, our model has endogenous disequilibrating forces; hence it does not attribute instability strictly to bad luck.

Post-Keynesians who find our model amenable hope for wider coalitions in favor of pro-growth fiscal and monetary policies—an end to austerity, monetary suicide pacts, etc.—in spite of the fact that spending priorities are often warped. They expect wages to rise with strong demand, but are optimistic that demand-inflation does not set in until excess demand is very high. They support the processes that they believe are favorable for growth, confident that benefits will be widespread and popular. These pro-growth measures include antitrust enforcement and increases in the minimum wage that are known to ripple throughout the wage distribution and even increase the profit rate in all but the most export-oriented economies. Here, they find themselves in agreement with many mainstream (neoclassical) Keynesians. Environmental concerns bring policy imperatives of their own but do not change this message. We post-Keynesians and adherents to modern money theory (MMT)–another theoretical influence–worry about levels of private-sector debt but find current warnings about rising public debt as a growth-reducer to be crackpot, rather than merely alarmist or overblown.

It was a great chance to learn more about this department and its approach and meet more of its faculty, staff, and graduates.

For the curious:

working papers about the model: Hannsgen (2012); Hannsgen  and  Young-Taft (2015)

published version of Hannsgen (2012) (Metroeconomica 2014) (abstract; paper available on request from me; my address: mail at greghannsgen dot org)

Further thoughts: I  should note that some details of the models mentioned above  remain to be worked out; things are by no means set in stone about all model versions mentioned above, and Tai and I always welcome comments and ideas; also, the comments by Skott mentioned above came during an exciting conference session on the state of heterodox economics featuring fellow-Kaleckian and SFC partisan Michalis Nikiforos of the Levy Economics Institute and others.


theme by teslathemes