In an earlier paper (Benhabib, Farmer (1992)), Jess Benhabib and Research Fellow Roger Farmer showed that if increasing returns to scale (IRTS) are as large as 1.5 or 1.6, theoretical models can be constructed in which the principal source of business fluctuations are the ‘animal spirits’ of investors, rather than of real shocks to the economy. Unlike Real Business Cycle (RBC) models driven by technology shocks, fluctuations in this model are inefficient and may warrant government intervention. Most estimates of scale economies place these only in the range of 1.03-1.15. In Discussion Paper No. 1403, Benhabib and Farmer show that even when the scale parameter is revised down to this lower level, it is still possible to construct models in which the principal source of economic fluctuations arises from inefficient swings in ‘animal spirits’.

Technically, the paper shows that the equilibrium of the economy may be indeterminate if consumption and investment goods are produced in separate sectors, which is the condition required to reduce the magnitude of IRTS needed to generate belief-driven equilibria. The authors are unsatisfied with existing Keynesian macroeconomic models because they are unable to explain why individuals behave in the way they do. The RCB approach is unsatisfactory because it leaves no role for economic policy and does not account for business fluctuations in a believable way. Therefore the authors adopt a compromise which explains market failures as a consequence of inadequate assumptions of the RBC model. In particular they explore the implications of the failure of these assumptions for dynamic general equilibrium models, notably perfect competition and constant returns to scale. The expected pay-off from this approach is a coherent theory of economic policy that explains when and why fiscal and monetary policies should be used as a means of stabilizing economic fluctuations.

Indeterminacy and Sector-specific Externalities
Jess Benhabib and Roger E A Farmer