Reviewed by Jelle Versieren
In this book Weeks has turned the logic of neoclassical (full employment) and synthesis models (less-than-full employment) against itself. Whereas (post-)Keynesian models explain growth of income as a multiplier process of investment and consumption – the analysis of the effects of excess aggregate demand on income and output – the initial neoclassical real model does not rely on money income as a dynamic variable (109). In a post-Keynesian dynamic system, relative prices and commodity quantities interact in an asymmetrical and complex way. Neoclassical economists cannot explain this interactive process of relative prices and quantity output, and falsely presume the independence of demand from the absolute price level (40-9). The neoclassical model gives a formalised coherent explanation of changes of the real price and the real value of a single commodity, but it fails to relate the equal relationship of nominal price change to the real price change. Knowing the real value of a commodity with a nominal price is an ex post facto scientific activity. The neoclassical economists instead think aprioristically. The process of calculating constant prices contains a reduction of concrete commodities to an empty abstract single commodity without giving a ground of reasoning. To have constant prices, an index number has to be assigned, which is not an arbitrary inference (107). Because the one-stage neoclassical models rest on market clearing price flexibility, this index number problem is not an arbitrary measure in the monetary model.
Neoclassical economists have presumed that the concrete world of nominal prices only passively reflect real values of a homogenous single commodity. But Keynes had proven that nominal money values influence business and household decisions. Neoclassical economists sought new solutions to render their false dichotomy model between nominal values and real values coherent. Nominal values still needed to reflect the explanation of a single commodity model. The Neo-Walrasian Patinkin designed a formal model connecting the two sets of values by what is being called the real balance effect: the effect of a change in the general nominal price level on the real value model in order to have conceptual and empirical neutrality of the nominal money-form. There is still only one kind of commodity in both worlds. But it also implied that the price adjustments to money supply do not happen at once, and that the fluctuations of the price-form cannot be automatically equalled to the demand for money. Thus, this real balance effect implies that with excess demand for money a change of the purchasing power of money occurs, which negates the homogeneity postulate between the price level and the amount of money (59-61). Patinkin did nothing more than establish a formal procedure without tackling the important logical inconsistencies of the real/nominal model, especially when the price level is a variable in a money model, when engaging with disequilibrium processes. Money demand still relies on the labour market-income relation and a unique pricing mechanism. Therefor money cannot be reduced to a restricted amount of saving goods in the single commodity expression.
The neoclassical supply-side equals the value of final commodities to value added in production (22). This means that the income consisting of profits, rents and wages can be interchanged with the output of the aggregate supply of final commodities. Output or income equals to the factor payments of the marginal products of labour and capital. Output/income will not be analysed according to distinct productive and distributional properties, which makes the output/income of the single commodity unaffected by the change of absolute or relative price changes. This single commodity erases the differences between distinct consumption and investment commodities. The production function of income relies on a fixed amount of capital and the marginal productivity of labour with a constant return to scale. Hypothetically, both capital and labour change proportionally, the single commodity has to be produced by the same proportions, and labour obtains its equilibrium value at full employment. The production process does not influence nominal or real values. Therefore, in the real and monetary model only the labour market can disturb the general equilibrium. Nonetheless, only a formal link exists between the labour market and the commodity/money market, and the established determination of the former goes without explanatory devices (39-40; 68-9). The equilibrated real wage determines the interest rate. In order to avoid any surplus approach of the profit rate, the interest rate equals profit as a reward for capital lending. Thus investments are being determined by the interest rate. In the end, factor income in real or monetary terms is nothing more than tautological, because the value added consists of the sum of wages and capital (26-7).
The demand and supply of labour is always set in an equilibrium position, therefore real wages only have one optimal unique price, and the level of output/income via the interest rate is determined by the optimal amount of capital. In this situation, via the consumption function the level of savings equals the level of investment determined by income/output. In this neoclassical paradigm, outcome can never be analysed according to independent investment decisions, and income cannot be restricted or accelerated by the excess demand for investments or by a coordination problem between savings and investments. On the one hand the equilibrated supply of labour to a certain demand is independent of the interest rate, thus wages determine income/output, on the other hand the very same wages do not play a specific role in demand schedules. Different kinds of ‘endowments’ in the production function do not have an impact on savings or investment behaviour. According to Walras’ Law, investment automatically adjusts to savings and excess demand or excess savings will vanish when transactions occur at the unique equilibrium price (Morishima 1977: 95). At the same time, the demand for labour services has been calculated as a cost for the capital owners in order to put their capital in a productive motion. Furthermore, external restrictions on the supply of labour do not exist, because neoclassical economists presume a trade-off between leisure and income at the full employment wage rate. In the schizophrenic world of the neoclassical economist a wage labourer is at the same time completely free to decide whether to work or not for an equilibrium price set by the demand of capital owners, and the wage rate does not have any repercussions on the level of savings and investment in the commodity market. In addition, the neoclassical economist does not distinguish the different effects of excess supply of labour for the labourer or the excess supply of commodities for the capitalist (34). The result is a ‘system of simultaneous equations that yield a unique solution in which the components of aggregate demand are derivative from the determination of output/income in the labour market’ (29).
The Keynesian consumption function is still interest-elastic. Keynes’ innovative contributions on aggregate demand need modifications regarding the dynamics of qualitative changes of different kinds of income between two distinctive time periods (Domar 1957: 30). But even in the synthesis model Weeks can prove that:
1. The simple transposition of the neoclassical assumption into a monetary model creates multiple logical contradictions and therefore no dichotomy can exist.
2. In the monetary model Walras’ Law of the price adjustment process and market clearing cannot be linked to the quantity theory of money and the general equilibrium condition (43; 53-59).
3. Money is non-neutral and endogenous when a barter model has been set aside. Patinkin’s real balance effect was intended to shift the saving function in the commodity market through a price adjustment. But in this revised monetary model, the distinction between outside money and inside money, together with bonds and money, already negates the neoclassical identification of money with transaction demand (and the Pigou-effect correlation between prices and demand). As with the former dichotomy model, the labour market formally determines all other variables (81-2). Also, at a level of less-than-full employment, changes in the money supply will alter real values (71). Money supply is not independent from the demand for money.
4. The less-than-full employment models with wage rigidity demolishes the idea of a unique pricing system of the neoclassical general equilibrium according to the supply and demand of labour
5. Initial changes in the commodity and money markets have a direct effect on the labour market.
6. Lags and contradictions occur between circulation and distribution of money and commodities in which consumption and investment will play a determining role.
7. The Walrasian model is only suited to explaining the circulation process of commodities without a labour market or production conditions (126-7).
Aggregate demand, according to the synthesis model, is the derivative result of an arbitrary determined money supply at a fixed income and interest rate (30). From a rigid wage perspective employment drops faster than income, because labour produces according to diminishing returns. Within an equilibration of the money supply and demand prices will rise less. Unemployment occurs when wages are too high and output/income will fall because of higher prices (75-6; 126). Excess demand for the single commodity is tied to low wages/labour shortage and high income/commodity shortage:
The less than full employment solution is associated with equilibrium in the commodity market and the money market. Saving equalled investment and the demand for money equalled the supply … Therefore, rigid money wages yielded a solution in which the labour market had excess supply, but the excess supplies and demands in all other markets were zero … The result has the appearance of being forced upon the theory by necessity rather than arising from its logic … No systematic tendency to full employment is logically possible with the (Walras’) Law … Should the commodity market be nominated to balance the excess supply in the labour market with an excess demand, then the situation is logically inconsistent. An excess demand for the single commodity, provoking a rise in price and output/income, implies that the money wage is too low, a labour shortage, contradicting the initial situation of excess supply of labour, and rendering the downward inflexibility of the money wage irrelevant (125).
In the real world, scarcity of labour is related to high wages and high prices to excess demand. The mentioned inversed relations are only possible under the conditions of autonomous/exogenous money supply.
Also the neo-Walrasian interest rate in the commodity market is flawed: it is perfectly possible that wages stay rigid because the interest rate has to stay positive, which contradicts Walras’ law (77). In the synthesis model, the interest rate equilibrates both saving and money supply, which already implies an equilibration of investment/consumption and money demand. The interaction within a certain time-frame between investment and the propensity to consume has not been problematized (Domar 1957: 246-54). The relation between a shift of cash from idle to active balances and the amount of commodity transactions are without any possibility of contradictions, because Keynes’ interest-elasticity of the demand for money is still congruent with money neutrality and quantity theory regardless of the purpose of the demand (86-92). Prices and money value rise, while income and real values drop (IS-LM curve). However, extrapolating from the liquidity trap example, at the very low rate of interest the demand for idle balances becomes infinitely elastic, which makes the future preference for money indeterminate. The rigid wage model with a conflict between saving and investment returns through the back door. The neoclassical solution of price and wage deflation is unable to relieve this paradox at the cost of a complete collapse of the commodity market (92-4) and a domino-effect of bankruptcies because of debt deflation. Thus, the rigid wage synthesis model with money neutrality cannot support Walras’ Law. The original purpose of the synthesis model of obtaining full employment equilibrium by the means of interest-elasticity signalled the paradigm incommensurability of money neutrality, quantity theory and market clearance. The synthesis model with Pigou’s wealth effect has the opposite result. In order to have real wealth optimisation, to possess units in equal real value for market clearance, money neutrality needs to be abandoned and needs to be replaced by a negative demand relation between money and bonds. There is a direct impact of money supply on real demand. The resulting price indeterminacy makes several equilibria between the commodity and money market possible (97-100).
Neoclassical theory ignores the heterogeneity of capital, joint-production and capital reswitching. In the real world individual capitals cannot consist of units of the single commodity. Production is not a homogenous process. This capital debate emphasised the necessity of a system of relative prices, and promoted a more fundamental understanding of the indeterminateness of income distribution which is closely tied to growth models and the sphere of production itself. Samuelson’s synthesis model failed in attempting to save the homogeneity postulate of the aggregate production function with a surrogate production function. Weeks dismantles the neoclassical idea of the devastating effect of rigid wages and capital substitution. Neoclassical economists presume that only flexible wages are compatible with high income, in which new techniques require more labour (147). The mechanisms of unemployment, underutilisation and productivity of capital are absent. A fortiori, Weeks reiterates the analysis of factor intensity and profitability in a simple two-commodity model with two available techniques. The two-commodity world contradicts the neoclassical negative related wage-capital factor price model: the capital-intensive technique is more profitable with high and low wages, whilst at a certain medium wage the labour-intensive technique accrues more profit. It also can be concluded that:
it is not the case that more capital-intensive techniques will always be chosen as the real wage rise … the measured factor intensity of a technique is not determined by technology alone … With no unchanged technical coefficients of production, changes in the wage and the profit rate alter the factor intensity of a technique … the capital stock is not unique with respect to the distribution between wages and profits except in the case of a one-commodity world … capital stock varies with the wage and the profit rate … Because the factor intensities of the input and the output are different … relative prices varies with the wage-profit ratio (153-4).
The neoclassical inability to prove its factor price model becomes even more problematic when in a simple input-output model technical coefficients have dynamic qualities in order to preserve a certain rate of profit. It did not pay attention to effective demand, income constraints, investment and expectations, the relation between asset prices and wages and the connection between income distribution and class structure of monetary exchange (159-66).
John Weeks’s book delivers what it has promised: a thorough critique of neoclassical theory and the synthesis models by undermining the contradictory propositions and implicit assumptions and turning these against themselves.
31 January 2014
- 1957 Essays in the Theory of Economic Growth (New York: Oxford University Press)
- 1977 Walras’ Economics. A Pure Theory of Capital and Money (Cambridge: Cambridge University Press)