|
Attempts to model exchange rate behaviour have produced a wide variety of models with an equally wide variety of assumptions. Microeconomic theory often has ambiguous implications for the structure of such aggregate models. The effects of macroeconomic policy may well, however, turn on these uncertain aspects of the model's structure. In an earlier Discussion Paper No. 152, Research Fellow Neil Rankin studied a model of a two-country world with flexible exchange rates in which disequilibrium arose because the wage was inflexible in the first period (the short run), producing unemployment. In Discussion Paper No. 185, he explores the behaviour of a similar two-period model but introduces an additional source of disequilibrium through the short-run rigidity of goods prices. This gives rise to an excess supply of goods and thus the Keynesian multiplier process familiar in the disequilibrium literature. Rankin notes that whether the goods price is better assumed to be flexible (as in Discussion Paper No. 152) or fixed is an open question among supporters of 'Keynesian' models. He argues that both specifications are plausible and worth exploring. Rankin's model is based on explicit microeconomic foundations yet closely resembles the 'standard' two-country model of Mundell, based on the IS-LM paradigm (in which there is implicitly a fixed price of goods giving rise to excess supply). It thus enables us to see what effect the introduction of microeconomic foundations may have on Mundell's results. Rankin analyses the impact of policy changes such as a permanent increase in the money supply or in government spending, balanced in each case by adjustments in lump-sum taxation. For a 'small' open economy he finds that, as in the conventional analysis, a monetary expansion leads to an increase in current output and an exchange rate depreciation. The intertemporal nature of Rankin's model introduces a degree of freedom into the behaviour of the domestic real interest rate relative to the world rate, however: with a monetary expansion the real (and possibly the nominal) interest rate falls. Rankin's model suggests that the effects of a fiscal expansion differ from those suggested by conventional models: the exchange rate and the nominal interest rate are unchanged, while the real interest rate falls and the current account is unaffected. The multiplier effect on output is unity, reminiscent of the 'balanced-budget' multiplier for a closed economy: under Rankin's assumption of perfect foresight, government borrowing is equivalent to (future) taxation so that all fiscal changes are effectively balanced-budget changes. Rankin extends his analysis to a two-country model and obtains a surprising result. Cross-country multipliers are zero: each country is isolated from the other's policy changes, and there is no 'policy interdependence'. This recalls the 'insulating' properties which a flexible exchange rate regime possesses in the simplest open-economy model, but which Mundell argued would break down in the presence of capital flows. Rankin cautions that the results, though striking, should not be interpreted as an unqualified dismissal of the importance of policy interdependence. In particular, they show that the question of whether economies have the structure assumed by Mundell, as opposed to that assumed by Dixit (studied in Discussion Paper No. 152), is of vital importance. Both specifications are inherently plausible and common in the literature, and both have considerable claims to the label 'Keynesian', yet they yield very different policy conclusions. An Intertemporal Version of Mundell's Two-Country Flexible Exchange Rates Model with Disequilibrium Micro-Foundations: Is Policy Interdependence Inevitable? Neil Rankin Discussion Paper No. 185, June 1987 (IM) |
|