lifted.¹ Paradoxically, some ardent market reformers (domestic and foreign) seem to have underestimated transformation costs for the opposite reason: they were so thoroughly convinced of the massive inefficiencies of the old systems that they hoped changes in key incentives would trigger rapid improvement after only brief disruption. The repercussions from the collapse of the Council for Mutual Economic Assistance (CMEA), the disintegration of the Soviet Union, and (more briefly) the Middle East War were also unexpected.

Although the extent of social costs was a surprise, both the public and reformist officials certainly did expect costs; many were deeply concerned. Broadly, policy responses took two main forms: (1) ''selective gradualism," or measures to delay (and therefore, it was hoped, soften) those reforms that would most directly harm large numbers of ordinary people, and (2) compensatory programs, mainly through direct cash transfers.

Selective gradualism with respect to consumer price decontrol and state enterprise reforms was designed to prevent or delay social costs from occurring. Many countries retained full or partial price controls for medicines, rents, utilities, and sometimes selected food staples for months or years after most other price controls and consumer subsidies had been removed. Most governments also moved quite slowly on privatizing or otherwise reforming large state economic enterprises, in part for technical reasons, but also because of concern about the social costs and political risks of rapidly rising unemployment. For instance, in the Czech Republic, despite far-reaching and determined market-oriented reforms in general, bankruptcy legislation that would have forced action on inefficient state enterprises was thrice delayed; when it was finally passed, its implementation was again postponed.

In contrast to selective gradualism, compensatory programs were designed to mitigate the impact of costs after they occurred. Concern focused mainly on the unemployed and the elderly, but also on families with young children. Between 1987-1988 and 1993-1994, most post-communist countries increased cash social transfers, usually by 2 to 4 percentage points of gross domestic product (GDP). Hungary and especially Poland did much more; by 19931994 both were spending almost a fifth of GDP on these programs. The Czech Republic, Moldova, Romania, and Russia maintained but did not increase cash transfers as a share of GDP.² However, even in those countries that substantially increased money transfers and partly indexed some benefits, inflation eroded their real value. In countries where increases were smaller, the real value of social transfers shrank to a fraction of earlier levels. More-