The Productivity, Trade, & Institutional Quality Nexus: A Panel Analysis

We estimate the relationship between productivity and trade for a panel of countries over the period 1980 to 2000 using instrumental-variables estimation of a productivity equation. We note that some estimates of productivity gains attributed to trade capture instead the roles of institutions and geography. The endogeneity of trade and institutional quality is accounted for by using instruments. We extend the Frankel and Romer (1999) specification, using real openness to measure trade (following Alcala and Ciccone, 2004), which allows for identification of channels through which trade and production scale affect productivity. The trade instrument is based on a ‘theoretically motivated’ gravity equation. The instruments for institutional quality come from Gwartney, Holcombe and Lawson (2004). Contrary to Alcala and Ciccone, our results suggest no robust relationship between real openness and labour productivity in the 1980s. Conversely, the relationship between productivity and real openness appears to be robust from 1990 onwards and similarly in the case of institutional quality. We also find evidence implying that countries with low-quality institutions are also able to benefit from openness to trade.