Developing countries committed to reform that are unable to abandon `temporary' price controls must decide whether to end them `cold turkey' or gradually, and this issue is now also of great practical significance for Eastern Europe and the USSR. In Discussion Paper No. 510, Research Fellow Sweder van Wijnbergen develops a two-period model of a small country with traded and non-traded goods sectors seeking to decontrol its domestic prices. `Cold turkey' requires an immediate transition to market prices, while `gradualism' entails increasing (controlled) prices today while announcing market prices for tomorrow. If a gradualist reform programme is abandoned tomorrow, prices remain at today's levels, but if cold turkey fails they revert to their lower, pre-decontrol levels. Producers of non-traded goods maximize income for a given real interest rate and expected second-period prices, while consumers maximize utility subject to rationing and an intertemporal budget constraint. No individual agent believes that its own actions will affect the `collapse probability', which first-period aggregate shortages influence nevertheless, since voters choose after the first period whether to vote an `anti-reform' opposition into power.

Under gradualism, a greater given `collapse probability' implies a smaller likelihood that prices will rise between periods, less incentive to hoard, and a greater observed supply response. As first-period prices and output increase, however, producers increase inventories, shortages develop, net supply falls, voters become more pessimistic, and the `collapse probability' increases. Under `cold turkey', in contrast, market prices are equal in both periods, and there is no incentive to hoard, so observed supply elasticity remains high and the `collapse probability' low. Further, even with no hoarding, gradualism implies a supply response no greater than under cold turkey; and since there will always be hoarding under gradualism, cold turkey is unambiguously more credible. Van Wijnbergen concludes that modelling a reform's collapse probability as endogenous when there is intertemporal speculation strongly supports the case against gradualism, so immediate decontrol is preferable on not only microeconomic but also macroeconomic grounds.

Intertemporal Speculation, Shortages and the Political Economy of Price Reform: A Case Against Gradualism
Sweder van Wijnbergen

Discussion Paper No. 510, May 1991 (IT)