General Equilibrium: From Fixed Prices to Imperfect Competition

In 1977 the International Economic Association (IEA) published a well-known collec­tion of essays devoted to the Micro economic Foundations of Macro economics. In his con­tribution, Frank Hahn (1977: 25) summed up nicely the gist of the problem: after forty years, Keynesian economics was still “plainly in need of proper theoretical foundations”.

Models of such a kind were provided in the 1970s under the label of “disequilib­rium” analysis (an extensive appraisal is provided by Gale 1983), most remarkably by Jacques Dreze and Jean-Pascal Benassy. Though different in their formalization of rationing, they share one basic feature: assuming prices as given, the focus is on the general equilibrium consistency of exchange and production choices under rationing. While this fixed-price methodology left one key question unanswered, it did provide a clear microeconomic account of spillover effects - the most popular instance of which is the well-known Keynesian multiplier. That key question, of course, concerns prices. Different “regimes” and policy options could be studied, as the given price constellation allocates rationing across markets: indeed, a Keynesian situation could be characterized as one where the workers’ rationing on the labour market links up with the firms’ on the commodity markets; but price determination, and hence what pushes the economy in a Keynesian regime, was left as an unsolved problem.

The NK literature offers a solution to that problem by dropping the commonly made assumption that agents are perfect competitors (see Hart 1982: 109): an approach, pio­neered by Hahn’s work on conjectural equilibria, which came to the forefront in the early 1980s and was rich in important implications. It can be looked at as providing an answer to two related, though different, general questions: whether price making behaviour can shed any light on nominal rigidities; and whether the interaction among price making agents can account for equilibrium unemployment, beyond the microeconomic ineffi­ciency we standardly associate with imperfect competition. While historically both per­spectives can be questioned as being somehow beyond Keynes’s own theoretical outlook, many results do have a Keynesian ring to them: notably, the emphasis on equilibrium unemployment as the upshot of “low” aggregate demand.

In principle, imperfect competition with price-making agents cannot in itself lead to nominal price rigidity: agents are no less rational for being price makers, and their con­trolling their own price should not in itself prevent them from adjusting it to its optimal (utility or profit maximizing) level - a change in (say) money supply should lead to a rea­lignment of nominal variables, leaving relative prices unchanged irrespective of competi­tion being perfect or imperfect. This being so, the NK approach focuses on price making to ask how indeed prices are made: it is the process whereby a price is set, which becomes the issue - in fact, a long-standing issue in Keynesian economics, from mark-up pricing seen as the upshot of rule-of-thumb behaviour, to the analysis of monopoly pricing when the demand curve is kinked.

In general, NK models differ from this tradition by emphasizing the distinction between the individual cost of non-optimal choices, and the resulting aggregate effect. Any theory having agents not reacting optimally to an exogenous shock, has to confront itself with the problem that suboptimal behaviour involves by definition some loss - why then should it persist? The two main approaches to nominal sluggishness in this vein go under the names of near rationality (for example, Akerlof and Yellen 1985) and menu cost (for example, Blanchard and Kiyotaki 1987): both answer that question by arguing that the cost of maladjustment may be “negligible”, the former using first order approxi­mations, and the latter “small” adjustment cost, to define precisely what “negligible” means.

The standard framework to convey these ideas is that of monopolistic competition - a suitable setup for macro analysis, for each agent’s market power is by assumption con­sistent with his perceiving no aggregate effect of his own choices. In a general equilibrium model so conceived, an increase in money supply would result in each firm facing an outward shift in its own demand, and increasing its own price to the new profit maxi­mizing level - therefore, money would be “neutral”, as output and employment would settle at their initial values. However, menu costs or near rationality may lead firms to refrain from price changes, as by doing so they perceive just a negligible loss - a percep­tion owing to the single producer not taking into account how his own choice affects his own demand, via its indirect effect on aggregate demand. In this case the macroeconomic externality would work through the aggregate price level: in fact, the latter will not adjust to the new money stock, real money balances will be higher, and monetary policy effec­tive on output and employment.

This aggregate demand externality thus provides an important argument for nominal rigidity, as real, first-order aggregate effects of nominal shocks are supported by second- order (negligible) losses at the individual level. However, it also points to a more general interpretation of the working of the economy, which is broadly consistent with tradi­tional Keynesian concerns and, in fact, independent of the role of money.

Whenever firms are endowed with market power, equilibrium price will be above mar­ginal cost - firms would like to expand production at given prices, but are constrained from doing so by the quantity the market is willing to take at that price.

This being so, two points are worth stressing. First, equilibrium unemployment in this framework is not simply due to the standard (partial equilibrium) effect of imperfect com­petition: the macroeconomic externality is grounded in the coordination failure among agents, and cannot be reduced to the market power of the single agent vis-a-vis his own market.

As a final remark, it should be borne in mind that some might deny this approach to the Keynesian label: equilibrium “underemployment”, rather than “unemployment”, should be used to emphasize the inefficiency associated with imperfect competition. To this one can retort that in a general equilibrium of this sort firms and workers are rationed, in what is arguably a “Keynesian” situation where both would like to supply more at the going prices - the traditional definition of rationing makes sense only under perfect competition, and switching to imperfect competition does involve a different reference framework.