The transition to a market economy in Eastern Europe is proving much more complex and painful than initially expected. Throughout the region, dramatic falls in output and sharp declines in living standards have fuelled discontent. The result? An electoral backlash against reformers and a dangerous rise in nationalism. Neoclassical economists advocate the `big bang' approach to transition. In Poland, where this approach was tried first, growth has indeed resumed and economic prospects are encouraging, but the political backlash continues. Elsewhere, reformers have suffered electoral defeats and the heirs of the old regimes have been returned to power. Are further reforms doomed?

We believe that a more gradual approach, with sensible sequencing, will help put reform back on track. Many forms of `gradualism' have, it is true, proven to be nothing more than excuses or alibis for maintaining the status quo. But the experience since 1989 warns against underestimating the political opposition to reforms. The only durable reforms are those which are politically sustainable. In order to be sustainable, reforms must be carried out in the right sequence, so that the early reforms help build constituencies which will support further changes.

The big bang strategies were motivated by the urgent need for macroeconomic stabilization. Hyperinflation does indeed require radical shock therapy: it cannot be curtailed through a gradual approach. A new government can take tough measures immediately and hope that stabilization-induced recession will be over by the time new elections are due.

The transition in Eastern Europe, however, involves much more than macroeconomic stabilization. Fundamental social and institutional changes must take place if economic agents are to adapt their behaviour to the market environment. Such fundamental changes cannot be achieved through new legislation alone. A new entrepreneurial class must be created and managers must develop the skills to cope, not with shortages, but with market competition. Painful restructuring is needed in loss-making sectors and enterprises, leading to enormous redundancies and shifts in the structure of employment. Changes of this magnitude are painful and create a political backlash. They cannot take place overnight: they must be sustained over a longer period and so must be designed to attract support from sufficiently large coalitions. Reforms which hurt all categories of the population simultaneously are unwise: what is needed instead are measures that create the constituencies needed to support future reforms.

We propose three general principles of sequencing which take into account these political constraints and so are designed to build constituencies for further reform:

• Implement first those reforms with the highest expected benefits and delay those reforms that are expected to hurt most.
• Implement first reforms that deliver even small benefits to a majority, not reforms that deliver large benefits to a minority.
  
• Implement painful reforms such as industrial restructuring gradually. Political support for these reforms requires compensating at least some of the losers, which may endanger budgetary equilibrium and macroeconomic stability. Gradual restructuring is less costly, because the government can `divide-and-conquer' its industrial sectors, by restructuring them one at a time.

China has successfully applied these three principles in sequencing its own reforms. There was no restructuring of state industry initially: instead, reform started with agricultural decollectivisation.

Agricultural output doubled between 1978 and 1987. This ignited a process of economic growth, which in turn led to a boom in the non-state industrial sector, especially in rural regions where output of township and village enterprises grew at almost 30% a year between 1981 and 1990. This has built constituencies which will enthusiastically support further reform. Prospects for the transition to a market economy in China are now very bright.

Industrialized Eastern Europe cannot, alas, simply copy the Chinese strategy. The importance of decollectivisation is unique to the Chinese experience, and no miracle can be expected from agricultural reform in Eastern Europe. Yet the same three principles of sequencing can be applied in the East European context:

• Encouraging development of small private enterprises should take top priority. This can deliver large welfare gains and allow an entrepreneurial class to thrive by providing new goods and (especially) services to consumers. Given the lack of existing private wealth, government policy should promote the allocation of credit to small- and medium-size private enterprises through measures such as partial credit guarantees or tax cuts.
• Privatization should start with the strongest state enterprises. Privatizing inefficient enterprises which need restructuring can wait.
• Restructuring should be tackled gradually, with selective closing down of loss-making enterprises. This requires continued subsidization of lame duck industries, but in the meantime their managements should be subject to tighter government monitoring. Such subsidies, however, make it more difficult to balance the budget, a key condition for macroeconomic stability. Such stability is important: governments should pursue it, but by concentrating on revenue, not public expenditure. This can be achieved by enforcing greater discipline in tax collection and by extracting more revenue from privatization. Deep cuts in subsidies and government expenditures are likely to derail reform: they are politically unsustainable and therefore ultimately self-defeating.

Mathias Dewatripont and Gérard Roland

Mathias Dewatripont is Professor of Economics at the Université Libre de Bruxelles, co-director of the European Centre for Advanced Research in Economics (ECARE) and a Research Fellow in the Industrial Organization programme at CEPR.