The Balassa-Samuelson effect reversed: new evidence from OECD countries

Matthias Gubler,Christoph Sax

doi:10.1186/s41937-019-0029-3

Matthias Gubler, Christoph Sax

Open Access

https://doi.org/10.1186/s41937-019-0029-3

Copy DOI

Abstract

This paper reconsiders the Balassa-Samuelson (BS) hypothesis. We analyze an OECD country panel from 1970 to 2008 and compare three data sets on sectoral productivity, including newly constructed data on total factor productivity. Overall, our within- and between-dimension estimation results do not support the BS hypothesis. For the time since the mid-1980s, we find a robust negative relationship between productivity in the tradable sector and the real exchange rate, even after including the terms of trade to control for the effects of the home bias. Earlier, supportive findings may depend on the choice of the data set and the model specification.

Highlights

The Balassa-Samuelson (BS) hypothesis—stated by both Balassa (1964) and Samuelson (1964), with a research precedent in the work of Harrod (1933)—is one of the most widespread explanations for structural deviations from purchasing power parity (PPP)1.According to the BS hypothesis, differences in the productivity differential between the non-tradable and the tradable sector lead to differences in price levels between countries when converted to the same currency
We analyze a panel of Organization for Economic Co-operation and Development (OECD) countries from 1970 to 2008 and compare three different data sets on sectoral productivity provided by the OECD, including a newly constructed data set on total factor productivity (TFP)
Our within-dimension dynamic ordinary least squares (DOLS) and betweendimension fully modified OLS (FMOLS) estimations point to a robust negative equilibrium relationship between productivity in the tradable sector and the real exchange rate for the time since the mid-1980s

Summary

Introduction

The Balassa-Samuelson (BS) hypothesis—stated by both Balassa (1964) and Samuelson (1964), with a research precedent in the work of Harrod (1933)—is one of the most widespread explanations for structural deviations from purchasing power parity (PPP). Differs from these studies in a number of dimensions, including the estimation methods, the data set, the sample period, and the set of OECD countries analyzed While these authors find a statistically significant negative relationship between the labor productivity of tradables and the real exchange rate, our analysis relies on sector-specific TFP, which is the preferred measure for productivity as noted by De Gregorio and Wolf (1994). Capital stock estimates may have been calculated differently and in a non-standardized way in the ISDB13 We use these three OECD data sets for the following reasons: first, sectoral productivity data from PDBi has, to our best knowledge, not yet been used in testing structural deviations from purchasing power parity (PPP). All results are in line with the results found in similar empirical studies (see, e.g., Calderón (2004); MacDonald and Ricci (2007) or Ricci et al (2013))

Methodology: cointegration tests and panel DOLS

Findings

The Balassa-Samuelson effect from the 1970s to the 1990s