Vanishing income redistributions: Keynesian clues about model surprises in the short run

A compact, one-sector version of recent large-scale income distribution models is presented. Different specifications of endogenous and exogenous variables in the model (or assumptions about how it is closed) change its qualitative behavior directly. A ‘Keynesian’ closure in which investment is exogenous and nominal changes in prime cost are passed along into price increases demonstrates relative insensitivity of the functional income distribution to factor prices. An alternative ‘neoclassical’ closure in which investment is endogenous (determined by savings) and output prices are fixed is much more responsive. The large-scale models are closed on Keynesian lines, and also have relatively insensitive functional distributions. If these models adequately describe real economies, causes besides general equilibrium price response must be invoked to explain distributional change.