The ‘Lowbrow Theory’: Macroeconomics and Growth Theory
General economic equilibrium and expected utility theories were taken up by mainstream US economic culture as reference as far as pure theory was concerned. However, various problems remained open, and in particular questions regarding the analytical background for the formulation of policy strategies: macroeconomics, monetary theory, public finance.
In these fields, the Great Depression of the 1930s and the diffusion in that context of Keynesian-type policies of income and employment support gave rise to a dilemma: on the one hand, the pure theory of competitive equilibrium implies full employment of productive resources, labour included; on the other hand, actual experience brought forcibly to attention the issue of unemployment as a possible or, better, all too probable prospect. How could mainstream economists reconcile these two things?
As Sraffa (1930, p. 93) had suggested in a different context, when theory and reality clash, we can forego neither realism nor internal consistency in the theory: we should, rather, abandon the (neoclassical or marginalist) approach that had led to such an irresolvable conflict.
However, the US mainstream followed different paths, such as a dichotomy between long and short periods. In the long period and under competitive conditions the pure theory would retain full validity, including the full employment thesis. In the short period, on the other hand, the adjustment processes leading to equilibrium may be assumed not to be able to exercise their full effects, so that a situation of unemployment would be possible; in this context Keynesian policies may help insofar as they accelerate convergence to the equilibrium position. Alternatively, assumptions other than free competition were adopted, in particular concerning the labour market, where the presence of trade unions may impede reduction in real wages towards the equilibrium value notwithstanding the presence of unemployment.
Both roads allowed for separation between micro- and macroeconomics: the realm of pure theory from a dusty but more realistic field of research, or in other words the optimal realm of perfect competition from the actual world of market imperfections. In any case, in the field of macroeconomics, too, theorising was to rely on the requirement of equilibrium between supply and demand in simplified general equilibrium models describing the working of interrelated markets.
Thus, in a 1937 article John Hicks (1904-1989, Nobel Laureate 1972) proposed the so-called IS-LL scheme, which translated Keynes’s theory into the more traditional terms of a simplified general economic equilibrium model, with the presence of three markets: for goods, money and bonds (though the latter plays only a purely passive role). The goods market is in equilibrium when supply, i.e. production, is equal to aggregate demand (which, under the simplifying assumption of a closed economy with no government expenditure and no taxes, corresponds to demand for consumption and investment goods). This happens when savings, which are an increasing function of income, are equal to investments, which are considered a decreasing function of the rate of interest. The money market is in equilibrium when the supply of money (determined by the monetary authorities) is equal to the demand of money (for transactions, which is an increasing function of income, and speculative demand, considered a decreasing function of the rate of interest). In this model, commonly utilised for illustrating fiscal and monetary policies in support of employment, the traditional adjustment mechanism leading to full employment does not come into play because, quite simply, the labour market is not considered.
In two articles dated 1944 and 1963, Franco Modigliani (1918-2003, born in Italy, eventually emigrating to the United States to escape racial persecution, Nobel Laureate 1985) extended the IS-LL scheme to consider the labour market, too. In it, changes in the wage rate bring labour demand and supply into equilibrium, thus ensuring full employment. The ‘Keynesian’ (persistent unemployment) result is then arrived at by introducing some obstacles hindering the free operation of the labour market, connected to the trade unions’ bargaining power, which determines the downward rigidity ofwages. In this way, Keynesian theory is presented as a particular case of marginalist theory: the case in which full employment equilibrium cannot be reached, because the labour market is not a competitive market. We thus have the neoclassical synthesis, a synthesis between the neoclassical theory of value and Keynes’s theory of employment, which for decades dominated macroeconomics teaching all over the world.
Whenever the trade unions are able to exert some market power, public intervention aiming at reducing unemployment can lead to an increase in the rate of growth in money wage rates, which in turn generates an increase of inflation. The trade-off between unemployment and rate of inflation, known as the Phillips curve (Phillips 1958), represents, according to neoclassical synthesis economists, the set of possible economic policy choices.
There are some variants of the neoclassical synthesis. The first was originated by Robert Clower (1926-2011) and Axel Leijonhufvud (b. 1933), who interpreted Keynes’s as a disequilibrium theory, whose microfoundations are to be found not in the Walrasian approach but rather in the Marshallian and Wicksellian ones, taking into account the problems of information diffusion and intertemporal coordination of real
The second line of research is the so-called new Keynesian economics. Joseph Stiglitz (b. 1943, Nobel Laureate 2001) and others tried to locate the origin of unemployment in different kinds of market failures, related to the lack of certain elements that should characterise a perfectly competitive state. We thus have models based on menu costs (costs of adjusting prices that induce firms to adjust to changes in demand through changes in levels of production and hence of employment rather than through prices), insider-outsider models (in which those already employed enjoy a margin of market power that they use to get higher wages at the expense of higher employment levels), efficiency wages models (in which firms prefer to avoid reductions in money wages in order to retain experienced workers, presumably more efficient than potential new employees) - and the list might go on. Unemployment is thus explained through ad hoc assumptions of quite dubious generality, on the sandy theoretical foundations of one- commodity, one-representative agent models and/or partial equilibrium models with their inverse relationship between real wages and unemployment (a relationship that, as recalled previously, cannot be deduced from a general equilibrium model and that was the object of destructive criticism on the part of Sraffa and others).
The third line of research concerned extension of the neoclassical synthesis to the field of monetary theory. James Tobin (1918-2002,
17 Cf. Clower 1965, Leijonhufvud 1968. However, the models by Barro and Grossman (1971) and Malinvaud (1977) were formulated in Walrasian terms, with prices and money wages fixed and transactions taking place at disequilibrium prices, bringing about rationing of either demand or supply, and hence a ‘classical’ unemployment provoked by downward wage rigidity or a ‘Keynesian’ unemployment provoked by insufficient effective demand.
Nobel Laureate 1981) explained demand for money as a portfolio choice on the part of a rational economic agent in the presence of risk. Modigliani and Miller (1958) showed that, under conditions of perfect competition and absence of uncertainty, the different sources of finance (bank loans, issue of new shares or bonds, profits not distributed to the shareholders) are equivalent for the firms.
Among those who showed faith in the equilibrating powers of the market and hostility to state intervention in the economy, and thus to Keynesian theories and policies even in the watered-down version of the neoclassical synthesis, the Chicago or monetarist school was prominent. Milton Friedman (1912-2006, Nobel Laureate 1976), the recognised leader of this school, took on and developed the theses of the old quantity theory: in the long if not in the short run, the equilibrium level of income depends on ‘real’ factors such as resource endowments, technology and preferences of economic agents; the velocity of circulation of money is considered a stable function of the rates of return of various kinds of assets (money, bonds, goods, human capital). Friedman maintained that the money supply can influence income and employment only in the short run; in the long run it influences the general price level: the Phillips curve, negatively sloped in the short period, becomes vertical in the long period.20
Moreover, Friedman criticised monetary and fiscal policy measures, not only because their efficacy is limited to the short period, but also because the short period effects are uncertain and may well be negative. Indeed, economic policy measures are subject to three kinds of lags and uncertainties, arising over: evaluation of the situation in which to intervene; transition from such evaluation to choice of policy measures and their application; and, finally, the very impact of the policy adopted.
A still more extreme thesis was proposed by rational expectations theoreticians. In a 1972 article, Robert Lucas (b. 1937, Nobel Laureate 1995) combined the assumption of markets in continuous equilibrium with that of rational expectations, originally formulated by Muth (1961, p. 194), according to which ‘expectations ... are essentially the same as the predictions of the relevant economic theory’, so that economic agents learn to take account of public intervention in the economy, discounting its effects beforehand. Thus, for instance, deficit public expenditure, which is not financed by a simultaneous increase in taxation, adopted to stimulate aggregate demand, is counterbalanced by a reduction in private consumption resulting from the decision to put aside savings to pay the taxes that will sooner or later have to be introduced to pay the public debt with which public expenditure is financed. In this context, the Phillips curve proves vertical also in the short run: expansionary monetary and fiscal policy interventions can only produce an increase in the rate of inflation, not in the level of employment. Only policy measures unforeseen by economic agents (policy shocks) may have an impact, albeit temporary, on the real variables.
The only kind of economic policy admitted by rational expectations theoreticians is policy designed to reduce the natural rate of unemployment, i.e. the rate of unemployment corresponding to equilibrium in the presence of frictions in the working of the market that cause some unemployment. We thus have the so-called supply-side policies, consisting, for instance, in facilitating worker mobility from one job to another or ensuring that the qualifications with which the labour force is endowed correspond to the economic system’s requirements or reducing fiscal pressure, since increase in income net of taxes is accompanied, in equilibrium, by an increase in the amount of sacrifice (namely, productive effort) that economic agents are ready to make, and hence by an increase in accumulation and growth.
The neoclassical synthesis also embraces a theory of growth. The history of these developments began with a 1939 article in which Roy Harrod (1900-1978) used Keynes’s approach to define the warranted rate of growth, which corresponds to continuous equality between growth rate of productive capacity and growth rate of aggregate demand. Harrod’s model is based on three equations: the first defines savings as a function of income, the second follows accelerator theory in setting investments equal to the product between change in income and capital-output ratio, and the third expresses the Keynesian condition of equilibrium between aggregate supply and demand as equality between savings and investments. Substitution in the third equation of the expressions for savings and investments defined by the first two equations makes the ‘warranted’ rate of growth equal to the ratio between propensity to save and capital-output ratio.
Harrod, moreover, stressed the instability of the actual growth rate as soon as it diverges from the warranted rate: the knife-edge problem, as it came to be called. Whenever actual growth is higher than warranted growth, productive capacity lags behind. This implies an increase in investments, hence in aggregate demand, in the following period, which generates new increases in the growth rate. Conversely, if the actual growth is lower than that corresponding to the warranted rate, investments will be reduced and the consequent decrease in aggregate demand will bring about a further slowing down of growth.
An increase in unemployment may take place when the actual growth rate corresponds to the warranted one, but the latter is lower than the natural rate of growth, equal to the rate of growth of productivity plus the rate of population growth. Different mechanisms were then proposed in support of a tendency of the two growth rates to converge. According to the Malthusian approach, adjustment takes place through the growth rate of the population, which falls when increasing unemployment brings down the wage rate. According to Kaldor (1956), when unemployment grows, the wage falls, and since the workers’ propensity to save is lower than the capitalists’, the average propensity to save increases, corresponding to an increase in the warranted growth rate. Finally, according to the neoclassical approach developed by Robert Solow (b. 1924, Nobel Laureate 1987), the fall in wages brought about by increasing unemployment leads firms to adopt production techniques that use relatively more labour; thus the capital- output ratio falls; once again, this corresponds to an increase in the warranted growth rate.
These equilibrating mechanisms are not, however, without defects. For instance, it is dubious whether in present-day conditions population growth depends on the wage level, according to an inverse relation, as required by the Malthusian approach. The Kaldorian theory requires that increase in unemployment provoke a change in distributive shares in favour of profits, while in general during a crisis or a depression profits may well decrease more than wages. Finally, Sraffa’s 1960 critique and the ensuing debate showed that the capital-labour ratio cannot be considered as an increasing function of the wage. We thus return to Harrod’s original thesis, a typically Keynesian one: growth in a capitalistic economy is intrinsically unstable, and full employment is far from being guaranteed by automatic market adjustment mechanisms.
Solow’s theory of growth mentioned previously, despite its basic weakness, stimulated various streams of research. Solow (1956) introduced exogenous technical progress into the original model. A rich stream of empirical research followed, seeking to determine the relative contribution of capital, labour and technical progress to economic growth in different countries. Such works identified technical progress with the residuum, that is, with that part of income growth that is not justified by increase in labour and capital inputs; this means failing to explain the major component of economic growth. After some attempts at reducing the size of the residuum by including accumulation in human capital alongside accumulation in fixed capital, a new stream of research was opened by Romer (1986), connecting technical progress to income growth by introducing increasing returns or learning-by-doing mechanisms. This stream of research has unstable foundations: increasing returns are incompatible with competitive equilibrium of individual productive units, except for the case of economies of scale external to individual firms but internal to the industry (that is, to the economic system as a whole, in the ‘one- commodity world’ formalised in endogenous growth models); as Sraffa had already remarked in his 1925 and 1926 articles, this is a very specific case.
In opposition to the reinterpretation of Keynes’s theory proposed by the neoclassical synthesis and to the monetarist critiques, there was a decided reaction on the part of some post-Keynesians, exponents of the ‘new Cambridge school’ such as Richard Kahn, Nicholas Kaldor and Joan Robinson (see Chapter 14), together with American economists like Sidney Weintraub (1914-1983), Hyman Minsky (1919-1996) and Jan Kregel (b. 1944), who stress the importance of uncertainty, expectations and their volatility.
Instead of the simultaneous equilibrium of the various markets, post-Keynesian economists proposed a sequence of cause and effect relations: speculative demand for money affects the interest rate; this in turn, together with expectations, affects the level of investments; and in turn investments, through the multiplier, determine income and employment. Thus the influence exercised by monetary and financial markets on income and employment was stressed, in opposition to the thesis of the neutrality of money accepted in the classical and marginalist traditions. Moreover, various post-Keynesian economists maintained that the supply of money is endogenous: that is, the quantity of money (in particular bank money) in circulation is not rigidly controlled by the monetary authorities but depends at least in part on the decisions of other agents.
-  As a matter offact, the abstract theory to which reference is made is not the pure theory ofgeneral economic equilibrium, but a simplified version of it with only one commodityand one representative agent, utilised in order to maintain the thesis of existence andstability of a full employment equilibrium.
-  Among the instances of this line of research we may include instances of mismatchbetween the qualities of labour demanded and supplied (too many philosophers, toofew plumbers) that cannot be overcome in the short period.
-  This second road coincides with the first one if we assume that the trade unions are ableto impede the reduction in real wages only for a limited time span.
-  This line of research also includes Fama’s (1970) thesis holding that with efficient capitalmarkets the prices of financial assets correspond to the equilibrium values determined bythe ‘real’ factors of the economy and the CAPM (capital asset pricing models), which hasdominated the theory of finance over the past few decades. This line of research yieldeda few Nobel prizes: apart from Modigliani and Tobin, the other Nobel laureates wereHarry Markowitz (b. 1927), Merton Miller (1923-2000) and William Sharpe (b. 1934)in 1990 and Robert Merton (b. 1944) in 1997. However, the outcomes of the 2007-8financial crisis should suggest greater caution in this respect.
-  Cf. in particular Friedman 1956. 20 Cf. Friedman 1968; Phelps 1967.
-  This theory presupposes that all economic agents share the same model of the way theeconomy functions: the neoclassical one-commodity model, in which an inverserelationship between real wage rate and employment holds, so that, under competitiveconditions, a stable full employment equilibrium is obtained. In a multi-commoditymodel, the uniqueness and stability of such an equilibrium cannot be proved.The rational expectations assumption applied to a model of this kind would thusgive quite different results - in fact, everything becomes possible, and Keynesianuncertainty once again becomes relevant.
-  Learning-by-doing phenomena, discussed in Arrow (1962), appear when unit cost ofproduction decreases as experience is acquired, which means in proportion to thecumulated amount ofproduct. These effects should not be confused with the connectionbetween growth of production and technical progress (a dynamic form of increasingreturns to scale), which goes under the name of Verdoorn’s law (cf. Verdoorn 1949); itmay be associated with investments in new machinery, generally more efficient than themachinery already in use.
-  Let us recall here two other lines of research: one, at the boundaries with economicstatistics, originated by Simon Kuznetz (1901-1985, Nobel Laureate in 1971), lookingfor empirical regularities in the process of economic growth; and another, at the boundaries with economic history, originated by Walt Rostow (1916-2003), with his theory of‘stages of economic development’ (cf. Rostow 1960). We then have various otherresearch lines, quantifying qualitative variables such as democracy, corruption or thegood functioning of public administration or the administration of justice in order tostudy empirically their relationship with the rate of growth of the different nationaleconomies; obviously, the results of these studies depend on the way the qualitativevariables under consideration are translated into one-dimensional quantitative variables.Cf. for instance Acemoglu and Robinson 2006, combining empirical analysis of democracy with a theoretical analysis conducted on the lines of the public choice schooltradition and utilising aggregate growth theory.
-  Cf. Harcourt 2006 for an attempt to delineate a unifying theoretical structure for the newCambridge school, and Marcuzzo 2012 on some of the leading figures of the school.
-  Minsky (1982) developed on this basis an ‘endogenous’ theory of financial crises,based on the notion of ‘financial fragility’. This theory was utilised by Kindleberger(1978) in his historical investigation; Minsky’s theory has also been continuouslyreferred to in interpreting the most recent financial upheavals. Other lines ofresearch within the post-Keynesian framework are the stock-flow analysis developedby Godley and Lavoie (2007) and the monetary circuit proposed by Graziani(2003) and others.