ABSTRACT
Recent resurgence of interest in economic growth has led to a blossoming
literature.1 Among the many approaches proposed, cross-country growth
regressions have been widely applied to examine the role of various structural
and policy indicators in economic growth. The findings in this large and growing
literature are often conflicting and yet to be sorted out. For example, Levine and
Renelt (1992) carried out a sensitivity analysis of the determinants of the average
annual growth rate of GDP per capita for a sample of 101 countries during the
period 1960-89. According to their extreme bounds analysis, among many only
three variables (i.e., investment, international trade and initial income) are found
to possess fairly robust predictive power. Recently, Sala-i-Martin (1997)
presented a more optimistic study showing that a substantial number of variables
are strongly related to growth. This finding is supported by Ley and Steel (1999).
