Early experience of transition in Central and Eastern Europe appears to confirm that `big bang' restructuring is not feasible ex ante: recent election results in Lithuania, Poland and Russia indicate a public backlash against reform, while many hidden subsidies remain in the form of easy bank loans, inter-enterprise credit, tolerance of tax arrears and poor financial discipline. In Discussion Paper No. 942, Research Fellow Gérard Roland argues that the presence of political constraints implies that privatizing as quickly as possible and leaving restructuring to the new private owners increases the risk of partial renationalization and delay. An alternative policy of gradual privatization may establish a screening mechanism to separate good from bad firms, allow a sound private financial system to emerge, and impose more credible, harder budget constraints within the state sector.

Gradual privatization provides screening that facilitates the flow of private savings to good firms yielding high returns: while a high-quality loan portfolio enhances banks' incentives to apply hard budget constraints and extract greater effort, a low-quality portfolio increases creditor passivity, since a bank that initiates bankruptcy proceedings risks revealing its own financial weakness. Such screening also enhances the bargaining power of good firms vis-à-vis the government by improving their access to private finance and also that of the government vis-à-vis bad firms: the ex post incentive to bail them out falls as the cost of taxing good firms rises. The increased size of the private sector also implies that the constituency for reform is larger, so the political constraints on restructuring are correspondingly reduced. This enhances the government's credibility in threatening to close loss-making enterprises and allows it to harden the budget constraints of the remaining state-owned enterprises.

In Discussion Paper No. 943, Roland outlines the different strategies required to cope with ex ante and ex post constraints. While the former block decision-making and require relaxation or postponement of reforms, the latter concern the risk of backlash once the outcomes of such reforms have been observed and should be met by measures to make them irreversible. Advocates of the big bang maintain that a rapid transition can achieve a fait accompli that will constrain successor governments by raising the costs of reversing the policies adopted today. Lifting price controls and disbanding administration raises the costs of their reinstatement, while free distribution of state assets runs down government wealth and raises the costs of renationalization or the expansion of public expenditure. The success of this form of `scorched earth' politics depends critically on the assumption that the probability of losing power is exogenous.

Roland notes that gradualist programmes not only address ex ante constraints directly by reducing the costs of reversing reforms in their early stages but also alleviate ex post constraints by making the reform process irreversible: building pro-reformist political constituencies today enhances the likely support for the difficult decisions that will be required tomorrow. Since this requires careful sequencing, the reform process should start with the measures yielding the higher expected outcomes, those that will benefit a majority under a democratic system, and ceteris paribus those entailing a greater level of risk. Such careful sequencing enables reformers both to take advantage of the partial resolution of uncertainty to build constituencies and also to use `divide-and-rule' tactics to secure the adoption of reform.

On the Speed and Sequencing of Privatization and Restructuring
The Role of Political Constraints in Transition Strategies
Gérard Roland

Discussion Papers Nos. 942-3, April 1994 (IM)