The Nobel Prize winning economist George Stigler wrote a famous intellectual history of the concept of perfect competition in the late 1950s (see Box 1, below). The article has subsequently attained canonical status among orthodox economists. It does not mention the contemporary notion of competitiveness once, but as Matthew Watson shows in the latest post for the Fools’ Gold history files, it is full of important implications for those who wish to question the intellectual authority of the modern-day competitiveness agenda.
In particular, Stigler is forced to admit that there is no purely economic basis for supporting the idea of a perfectly competitive economy. His argument that there are still good grounds for keeping up this pretence boil down to straightforwardly political beliefs for wanting to keep the government out of economic life. Despite what will be shown to be their important differences, this is one thing that the theory of perfect competition and the modern-day competitiveness agenda have in common. Both are typically dressed in apparently economic clothes, but each in its own way must be understood as a purely political intervention into public affairs.
George Stigler and the Noble Lie
By Matthew Watson
It was an argument for the connoisseurs, but the agitated response on the economics blogs showed that it was not to be overlooked. It is not often that the pages of the American Economic Review are used for a bit of academic trash talk, but this is what Paul Romer has recently done in calling out some of his colleagues for their ‘mathiness’. Romer’s role as the accuser is noteworthy in itself, because he is widely regarded as a Nobel Laureate-in-waiting, and this is the sort of attention you try to avoid attracting if you want to keep in the Nobel Committee’s good books. Even more noteworthy, though, is the fact that two existing Laureates, including one of Romer’s academic mentors, were identified by name as chief perpetrators of this new professional sin. And this really is not standard behaviour for anyone who has their eye on the big prize.
Box 1: The Mixed Messages of ‘Perfect Competition’
Perfect competition is a state of the world assumption used in economic theory in which prices adjust, automatically and without hindrance, to any economic disturbance. It therefore represents a realm of pure market relations, where any influence that government might have over economic outcomes is assumed away. The assumption of perfect competition is a simplifying device designed to make theoretical work in economics more straightforward. Yet all too often it spills over into the normative argument that the real economy will work better, the greater the success in sidelining active government policy.
Mathiness is Romer’s description of something rather more than the tendency for economists to dazzle with mathematical sleight of hand. If this was all there was to it, most non-economists would think they were being told what they already knew only too well. Romer’s more specific complaint is that, in pursuing mathiness, many of his peers now work with models whose mathematical and economic dimensions do not add up to a coherent whole. These models may or may not be based on good mathematics that works well in its own terms. Some do, and their authors are rightly praised for the skill they have displayed; but some don’t – and Romer notes that economists are becoming less proficient at spotting one another’s mathematical mistakes.
More importantly, however, even when the mathematics is unquestionably of high quality, Romer says that this does not mean that it makes for good economics if its content is unnecessary for, or even runs counter to, the economic argument being made. This is the mathiness trap: protecting a dubious economic argument by cushioning it with some highfalutin’ but superfluous mathematics.
What makes Romer’s intervention particularly interesting for us at Fools’ Gold is its detail.
The accusation of mathiness is directed at those who take a specific stance on the macroeconomics of growth: they espouse a perfectly competitive worldview. However, it is not always intuitively obvious where the line is to be drawn in their work between the purely functional use of the assumption of perfect competition to make the theoretical modelling process more straightforward, and the purely political use of the assumption of perfect competition to allow the economic theory to say, ‘let the market rip’. Obviously it makes a big difference if the same assumption is being used analytically or normatively – or a mixture of the two. But the tracks left by mathiness serve to conceal the point at which one drifts into another.
Romer, meanwhile, is in any case positioned on the other side of the theoretical debate. He believes that new technologies feed into new growth trajectories in complex and unpredictable ways that simply cannot be captured by the neat symmetries of economists’ orthodox models of perfect competition. He also thinks that the weight of evidence on growth dynamics is increasingly on his side, but that the facts are being obscured by the mathiness of those whose professional reputation is tied to future generations of economists continuing to believe in the perfectly competitive world view (see box) . Mathiness, then, is a means of protecting this worldview when a sober judgement based on the accumulation of data points in a different direction.
Enter George Stigler: perfect competition as the Noble Lie
Economics bloggers have responded by turning to the crucial question of why – or, indeed, whether – the perfectly competitive worldview is important enough to justify defensive manoeuvres of the ‘mathiness’ kind.
Here we need to inaugurate into the Fools’ Gold Hall of Fame another Nobel Prize winner, the late Chicago economist George Stigler. Stigler’s view, according to Brad DeLong, was that the assumption of a perfectly competitive market structure represents a ‘Noble Lie’. That is, he knew that models cast in this framework act, at best, as partial descriptors of the economic system they are supposed to capture in its entirety. Indeed, he thought that the charge that they had a purely tenuous grip on reality was often justified.
Yet this did not mean that Stigler could see no merit in them. He argued that they could be defended in their own terms, because they showed what might be possible if an autonomous economic sphere could be created beyond the reach of government. They could also be defended in terms of what they were not, because as soon as you begin to model the economy through imperfectly competitive markets you turn economic theory into a planners’ paradise, and this was the last thing Stigler wanted economics to be. It is pretty obvious in his case, then, that the assumption of perfect competition is used normatively in the first instance and only analytically after that. If the practical political objective was to create an economy where each participant would be guided solely by competitive instincts, what better than the Noble Lie to serve that end in an educative capacity?
The most significant piece that Stigler wrote along these lines was his 1957 Journal of Political Economy article, ‘Perfect Competition, Historically Contemplated’. It closes with an extended eulogy to the ability of the perfectly competitive model to outperform all alternatives as the basis for rigorous economic theory.
“Since the 1930s,” he writes, “when the rival doctrines of imperfect and monopolistic competition were in their heyday [bringing with them, it must be remembered, arguments for an expanded role for economic planning], economists have increasingly reverted to the use of the concept of perfect competition as their standard model for analysis.”
It takes little to imagine an ear-to-ear grin breaking out on Stigler’s face when committing to paper a report of that free market triumph. “The vitality of the competitive concept in its normative role has been remarkable,” he gushes. In this way he feels he can record the unconditional defeat of its “newer rivals”.
Yet this is where things starts to get intriguing.
Box 2: Neoclassical economics and perfect competition: Stigler’s History
Stigler produces a typical economist’s Whig history of the evolution of his field. This is an approach to writing intellectual history in which bad theory is always driven out by good theory, and the modern day relates to its past through a series of theoretical advances. From this perspective, the theory of perfect competition is something which itself has had to be perfected. It starts off, Stigler says, from humble beginnings in the work of Adam Smith, who set in train a process through which the concept of competition “was long treated with the kindly casualness with which one treats of the intuitively obvious”. Indeed, he thinks that it was impossible before the Marginalist Revolution of the 1870s and the subsequent onset of the neoclassical era for the concept of competition to begin to take anything approaching the form of the modern notion of perfect competition. Without the changes to the whole basis of economic thinking enacted at that time, there could not have been the advances in clarity, precision and rigour brought to the concept initially by Francis Ysidro Edgeworth in the 1880s and latterly by Frank Knight in the 1920s. As argued in an earlier Fools’ Gold post, it was Knight who first introduced into economics an agent whose identity mirrored the competitiveness imperative in the form of the legendary ‘slot machine man’. True to the principles of Whig history, Stigler considers this third generation marginalist theory to be the high water mark of the exact theory of perfect competition within neoclassical economics: in fact, he calls it its “complete formulation”. Yet it also helped to propel neoclassical economics headlong into the position in which the pure market economy is conceptually indistinguishable from the pure planned economy.
Stigler insists that “the competitive concept can be no better than the economic theory with which it is used”. This is slippery. It allows Stigler to say: ‘my normative preference for the freest of free market economies will not let me subscribe to any economic theory other than that which enshrines the model of perfect competition, but I do not have to justify my normative preference because these models will always produce the best economic theory anyway’. Or, as he actually put it: “The normative role of the competitive concept arises from the fact that the equality of rate of return on each resource in all uses which defines competition is also the condition for maximum output from given resources”.
This makes it clear that, for Stigler, the competitive concept works best – and perhaps only works – within the theoretical framework of neoclassical economics (see Box 2). Unfortunately for him, however, it has been well documented from the 1920s onwards that neoclassical theory cannot differentiate on efficiency grounds between a pure market economy and a pure planned economy. Neoclassical theory produces equally strong solutions in both contexts. Sigler’s prized competitive concept therefore works equally well for institutional arrangements dominated by government planners as by price signals. In other words, there is nothing to distinguish analytically between the political solution that Stigler is desperate to avoid and the one that he wants to impose.
This makes it a purely political decision as to how resources are to be organised before neoclassical principles are put to work in the service of economic theory. And this in turn provides some very interesting insights for those of us at Fools’ Gold who are interested in learning more about how the modern-day competitiveness agenda works to enforce the image of whole countries somehow ‘competing’.
From perfect competition to competitiveness: two very different things
The modern-day competitiveness agenda typically involves special pleading on behalf of one economic sector or another in the name of the ‘competitiveness’ of the whole economy. This discourse is often wrapped in the rhetoric of ‘market realities’ by businesspeople and politicians alike, so that claims can be made about the advantages of a less intrusive state bureaucracy. The government is thus required to fall on its sword, in a manner that Stigler celebrates as a bonfire of planning regulations.
But what emerges in its place is not the freest of free markets that forms the basis of Stigler’s Noble Lie. In practice, the use of contemporary competitiveness discourse is much more likely to lead to the asymmetric application of market rules – as some people and firms lobby to have market rules bent in their favour, while others lack the resources to lobby successfully for their interests.
So-called tax ‘competitiveness’, being an issue exclusively for the well-heeled, is a perfect case in point. If your lobbying power is backed by clever accountants who can show how existing loopholes can be exploited, it is not hard to avoid paying for the public goods through which market rules are institutionalised in the interests of a level playing field. Finding that you possess a veto on contributing to the upkeep of the market’s regulatory institutions is equivalent to finding that you are no longer bound to respect market rules. They apply only to the rest of us, who have to foot the whole bill because we do not have the resources to negotiate our release from these obligations.
So despite what at first glance might appear to be their close similarities, modern-day talk about competitiveness imperatives is therefore a very different entity from the economic theory of perfect competition. Stigler’s Noble Lie of everywhere assuming perfect competition is designed to serve the general interest of enhanced efficiency, but his argument seems to point in two different directions at the same time. The economics of his argument cherishes the abstract purity of the perfectly competitive model, and it is consistent with imagining into being a realm of pure market relations. By contrast, its politics endorses the market-distorting effects of competitiveness discourse.
Stigler’s Noble Lie only becomes effective by superimposing onto the economic theory of perfect competition something resembling the politics of the modern-day competitiveness agenda. It bears repeating that the economic theory on its own cannot show that an economic system embedded in pure market institutions is any more efficient than any other type of economic system. Stigler, however, is prepared to try to pull the wool over our eyes by making precisely such a claim on behalf of the “rigorously defined concept” of perfect competition as bequeathed by the cumulative efforts of three generations of neoclassical economists. The claim works through the ungrounded assertion that allowing private owners to dispose of their capital in the way they see fit necessarily buys for the entire economy a step increase in the efficient allocation of resources. A very similar claim forms the basis of contemporary competitiveness discourse. Yet we know that political submission to that discourse leads not to altruistic investments of privately-owned capital to the benefit of society as a whole. Instead, it rationalises the political flexing of corporate muscles as a means of negotiating self-serving opt-outs from existing market rules.
Re-set in this way, how noble does Stigler’s lie look now?