Measuring the Institutional Environment
| Peter Klein |
An central problem in empirical research on institutions is the difficulty of measuring key attributes of the institutional environment. Secure property rights, respect for the rule of law, transparency, free markets, norms of fairness and reciprocity, and similar characteristics are held to be critical for economic development. But how do you know them when you see them?
Most of the literature has used indicators derived from secondary data, such as the economic freedom measures produced by the Fraser Institute and Heritage Foundation, Witold Henisz’s polcon database, the World Bank’s database of political institutions, various indexes of shareholder and creditor rights compiled by the La Porta gang, and the like. These measures have much to recommend them, but may proxy only indirectly for the real institutional constraints of interest.
This paper by Jan Svejnar and Simon John Commander takes a different approach: it asks. The authors use primary data from the Business Environment and Enterprise Performance Survey (BEEPS), in which managers of several thousand firms in 26 transition countries are asked for their subjective perceptions about tax and regulatory policy, uncertainty about regulatory change, macroeconomic stability, the effectiveness of the judiciary, corruption, crime, infrastructure, and other institutional characteristics. (The authors conclude that most of the variation in firm performance is explained by time-invariant country fixed effects, and that individual institutional variables have little explanatory power, suggesting that existing studies using secondary measures of institutional characteristics may overstate their effects.)