Productivity bias, institutional factors, and real exchange rates: An empirical analysis

Productivity differential between two countries has been identified as the major real factor contributing to the deviation of the purchasing power parity based exchange rate from the equilibrium rate. According to a proposition by Balassa (1964) and Samuelson (1964), a country with higher productivity experiences appreciation in her real exchange rate. This notion usually comes under the heading of productivity bias hypothesis. Previous researchers who tested the hypothesis mostly employed cross-sectional data and provided mixed support for the hypothesis. A few time-series and panel studies, mostly employing developed country-samples, have extended limited support for the hypothesis. This study, however, contribute to the existing literature by embarking on the most comprehensive study which employs a sample of 69 cross-sectional units (countries) over the 1960–90 period. The empirical results strongly support the hypothesis and are not sensitive to model specification and estimation procedure. We also test the relevance of other variables such as a measure of resource abundance and the black market premium, as other determinants of real exchange rate, confirming previous research.; In the second part of the dissertation, we analyze the impact of institutional rigidities on real exchange rates. No study has attempted to establish the empirical relation between measures of institutional rigidities and the real exchange rate. We argue that countries with higher degree of institutional rigidities in the forms of high corruption, bureaucratic inefficiency, deteriorating law and order condition and political instability do experience higher rate of inflation and decline in productivity, thus, currency depreciation in nominal as well as in real terms. To establish the empirical validity of our conjecture, we incorporate three proxies, which include three subjective indices measuring corruption, law and order condition, and efficiency of bureaucratic system, and for rigidities in the government institutions and a proxy for political instability for rigidities in the political institution into the productivity differential model proposed by Balassa. By assembling data from 65 countries over 1982–1990 period, we estimate cross-sectional regressions for each of the nine years as well as pooled regressions over the entire period. Most cases, the empirical analysis provides strong support to our claim that high corruption, inefficient bureaucracy, or declining law and order condition leads to depreciation in the real exchange rates. After controlling for the proxies of institutional variables, the findings also confirm the Balassa-Samuelson proposition that more productive countries experience real appreciation in their currencies.