Yesterday (here) I linked to a post by Matias Vernengo on Keynes’ theoretical contribution in light of the capital debates. I thought it might be worthwhile to elaborate on a central aspect of Vernengo’s post, particularly as it concerns the fundamental differences between Keynes’ ideas and the interpretation of neoclassical synthesizers of his work. Vernengo’s perspective on the significance of Keynes’ theoretical insights and the deep flaws in the marginalist interpretation is one that is probably held by most economists working in Sraffian and Post Keynesian traditions. I would think MMT economists also tend to share this perspective.
Briefly, the perspective is as follows. Keynes argued that unemployment is caused by a deficiency of aggregate demand. The neoclassical followers of Keynes – including the Neoclassical Synthesis, Old Keynesians and New Keynesians (but not Keynes’ immediate circle at Cambridge, Post Keynesians or the Sraffians) – tried to maintain the neoclassical notion that unemployment is caused by ‘market imperfections’ that prevent real wages and the real rate of interest from adjusting to their ‘natural’ levels.
The neoclassical position was simply a reversion to pre-Keynesian, marginalist modes of thought, in which it was considered legitimate to conceive of an aggregate labor market determining the real wage and an aggregate capital market determining the real rate of interest. It was already well recognized within neoclassical economics that imperfections could cause wage and price stickiness. If Keynes had been trying to say only that wages and prices could be sticky, he would not have been saying anything different to the orthodoxy of his day.
But Keynes clearly rejected the orthodox position on employment and interest. In chapter 19 of the General Theory, he argued that even if wages and prices were flexible, this would not eradicate unemployment. To the contrary, if anything it would be likely to make matters worse during economic downturns, because of the impact on demand of a redistribution of real income from workers to capitalists and rentiers. Keynes argued that, in general, nothing very decisive could be said about the impact of falling wages on aggregate demand and employment.
Keynes also rejected the notion of an aggregate labor market because the demand for labor is a derived demand. It is derived from the demand for the final products that labor is being employed to produce. Similarly, demand for capital goods is a derived demand.
An implication of Keynes’ reasoning is that alterations in wages and interest rates can impact on the demand for the neoclassical factors of production – labor and capital – in unpredictable and “strange” ways, because in impacting on product prices and the purchasing power of various income groups, these alterations in factor prices produce a complex of substitution and income effects that can operate in different directions. It suggests that there is no level of employment that the economy automatically gravitates toward (no natural rate of unemployment) and no tendency for the real rate of interest to move to a level that induces just the right amount of private investment to sustain full employment (Keynes explicitly rejected Wicksell’s notion of a natural rate of interest).
Vernengo, in the linked post, then discusses how the two opposing views – Keynes’ position that unemployment is the result of insufficient aggregate demand and the neoclassical position that unemployment is due to market imperfections – fare in light of the capital debates.
The debates showed that the marginalist claim of an automatic tendency to the full utilization of the factors of production could not be sustained. If there was unemployment, neoclassical reasoning suggested that a fall in the real wage would cause a substitution of labor for capital goods and a rise in employment. But this reasoning was shown to be incorrect. A wage reduction will cheapen not only labor but also capital goods that are produced through the employment of labor, complicating the substitution effects. Even if the substitution effects operate in the assumed direction, income effects associated with workers receiving lower real wages, capitalists receiving lower prices (though higher relative to wages) and rentiers receiving a higher real interest rate due to falling product prices (which redistributes income from capitalists to rentiers) could have an overall impact on demand that was negative, depending on propensities to spend of the various groups, and so could cause less of both labor and capital goods to be employed.
In general, the impact on employment of a fall in real wages is ambiguous and will depend on circumstances. At the moment, with demand depressed globally and domestic private sector demand also weak, falling wages would seem, if anything, more likely to exacerbate the employment situation than improve it.
The same difficulties afflict the neoclassical notion that investment responds inversely to the real rate of interest. It cannot be supposed, in general, that a lower rate of interest will be associated with a greater employment of the factor capital, because the change in interest impacts on the price of capital goods and product prices, therefore impacting on real wages, the distribution of income between workers, capitalists and rentiers, and therefore aggregate demand and the employment of labor and capital goods. This means that strange things can occur – referred to as ‘reswitching’ and ‘capital reversing’ – as Samuelson conceded in “A Summing Up” of the debates in the Quarterly Journal of Economics:
The phenomenon of switching back at a very low interest rate to a set of techniques that had seemed viable only at a very high interest rate involves more than esoteric difficulties. It shows that the simple tale told by Jevons, Böhm-Bawerk, Wicksell and other neoclassical writers — alleging that, as the interest rate falls in consequence of abstention from present consumption in favour of future, technology must become in some sense more ‘roundabout,’ more ‘mechanized’ and ‘more productive’ — cannot be universally valid.
The ramifications of the capital debates for pre-Keynesian neoclassical theory are therefore severe. The notion of a well-behaved demand for capital function and, with it, the supposed real determination of the rate of interest, is invalid. The supposed determination of real wages through the interaction of supply and demand in an aggregate labor market is similarly deprived of legitimacy. The neoclassical marginal productivity theory of distribution is also a casualty. As Vernengo concludes, there is no automatic tendency to full employment, and the natural rates of unemployment and interest do not exist.
In contrast, Keynes’ position – that unemployment is due to a deficiency of aggregate demand and that interest is a monetary phenomenon – is left unscathed.
In the previous post, I mentioned in closing that it is partly this understanding that informs Sraffian and Post Keynesian (and MMT) approaches to the determination of employment and interest. Specifically, output and employment are considered demand determined and the rate of interest to have a monetary determination.
For example, MMT suggests that unemployment will occur whenever government net expenditure is insufficient to sustain the full-employment net saving desires of non-government. The pre-Keynesian neoclassical response to that contention would have been that falling real wages would result in higher demand for labor and at the same time the real rate of interest would adjust in such a way as to induce non-government to alter its aggregate investment and saving behavior so that it was rendered consistent with full employment, irrespective of the government’s fiscal position. Keynes’ argument suggested that wage and interest-rate adjustments could not be relied upon to render aggregate investment and saving behavior consistent with full employment. The capital debates vindicated his position.
Addendum March 3, 2012
In a recent post, Vernengo has provided another in-depth explanation of the capital debates and its implications. It is an excellent and accessible introduction to the topic: