The link between governance and real effective exchange rates : a Balassa-Samuelson model

Abstract

This paper examines the relationship between Real Effective Exchange Rates and measures of governance provided by the World Governance Index database through the Balassa-Samuelson model. An additional control variable of terms of trade was added to partly capture country-specific effects within the dataset. Through both static and dynamic panel estimation techniques, the model was applied for a sample of 151 countries across the period starting from 1996 till 2017. The inclusion of such models was made to address any potential biases associated with panel data estimation. Static model results suggest strong supporting evidence toward the presence of the B-S effects, revealing general wage increases factored in higher prices of tradable goods that follow relative increases productivity of the tradables sector. Dynamic panel models find little supporting evidence toward the effect for the entire sample, mainly reporting a negative yet significant coefficient for the first lag of the relative productivity proxy, contradicting the theory, and an insignificant coefficient for the lag of the dependent variable. However, following segmentation of data into developing and developed countries, the study strongly substantiates the B-S hypothesis for all categories of developing countries. Furthermore, all six governance indicators were found to be statistically significant for different model specifications, having a mixture of effects on the dependent variable, suggesting the importance given to the determination of exchange rates through high quality management of governance.