As a symptom of fundamental institutional weaknesses, corruption needs to be viewed within a broader governance framework. It thrives where the state is unable to reign over its bureaucracy, to protect property and contractual rights, or to provide institutions that support the rule off law. Furthermore, governance failures at the national level cannot be isolated from the interface between the corporate and state sectors, in particular from the heretofore under-emphasized influence that firms may exert on the state. Under certain conditions, corporate strategies may exacerbate mis-governance at the national level. An in-depth empirical assessment of the links between corporate behavior and national governance can thus provide particular insights. The 1999 Business Environment and Enterprise Performance Survey (BEEPS) - the transition economies component of the ongoing World Business Environment Survey - assesses in detail the various dimensions of governance from the perspective of about 3,000 firms in 20 countries. After introducing the survey framework and measurement approach, the authors present the survey results, focusing on governance, corruption, and state capture. By unbundling governance into its many dimensions, BEEPS permits an in-depth empirical assessment. The authors pay special attention to certain forms of grand corruption, notably state capture by parts of the corporate sector - that is, the propensity of firms to shape the underlying rules of the game by"purchasing"decrees, legislation, and influence at the central bank, which is found to be prevalent in a number of transition economies. The survey also measures other dimensions of grand corruption, including those associated with public procurement, and quantifies the more traditional ("prettier") forms of corruption. Cross-country surveys may suffer from bias if firms tend to systematically over- or underestimate the extent of problems within their country. The authors provide a new test of this potential bias, finding little evidence of country perception bias in BEEPS.
