Does “New Keynesian” macroeconomics suffer from schizophrenia?
from Lars Syll
Taking part of the debate on microfoundations that is going on so intensely among economists these days, I wonder if Heinz-Peter Spahn isn’t more on the right track than those guys who desperately offer more or less contrived defenses of the microfoundationalist programme:
The crucial point however is: market conditions, which are presupposed in the model of intertemporal choice, are not given in reality. Distributing consumption optimally over time depends on the possibility of individuals to lend money on their permanent income, if temporary periods of low market income are to be bridged. Because this perfect financial market does not exist, consumption behaviour necessarily depends strongly on current income. Consumers know that their future expected income is distorted by spells of unemployment, the occurrence of which is hard to predict though; these quantity constraints are important also for firms …
Professional modern economics appear to suffer from schizophrenia as in the field of financial-market economics all these deviations from the Utopian ideal market are well known (information asymmetries etc.), which are stubbornly ignored when it comes to talk about macroeconomics in NKM [New Keynesian Macroeconomics]. The assumption of complete markets means that all agents’ intertemporal budget constraints always are satisfied, bankruptcies and insolvencies are impossible. The NKM world is populated by agents who never default … Basically, NKM designs a non-monetary economy … Questions regarding financial instability cannot be answered within this models, they cannot even be asked …
NKM faces an uncomfortable trade-off. On the one hand, General Equilibrium Theory has shown that preferences and behaviour of heterogeneous agents cannot simply be aggregated. Variances between individuals matter! The Sonnenschein-Mantel-Debreu problem states that choices may not be transitive; the representative agent’s ranking differs from individual rankings; reactions to shock may be different … On the other hand, if people are assumed to be identical, NKM may keep the representative agent, but as a consequence the model has no interaction of agents, no distribution problems, no asymmetric information and no meaningful stock market.
The critique so far may appear as unfair as it neglects the various refinements that were proposed in order to develop and improve the basic model set-up … But these extensions of NKM – due to the Walrasian method – yield many precisely-looking results … but do not grasp the impact of bank credit on goods demand, market income and employment in a typical monetary economy.