The Role of Nontraded and Traded Wages in the Productivity Differential Model

Abstract

Listen

Translate article

Purchasing power parity (PPP) remains the benchmark model for long-run nominal exchange rate determination since Cassel's pioneering work [3] despite weak empirical support [6; 7; 12; 14]. Seminal work by Balassa [1] and Samuelson [15] explains that movements in real exchange rates are due to changing real factors in the economy. Balassa and Samuelson show that differing productivity rates between traded and nontraded sectors can explain deviations in PPP; Hsieh [10] has empirically supported the importance of productivity differentials in explaining real exchange rate movements. Although the Balassa-Samuelson productivity differential model is commonly cited by international textbooks as an explanation of PPP failure [4; 11], several assumptions have not been explicitly tested. This paper examines four critical assumptions of the Balassa model for fourteen OECD economies and demonstrates that some are weak and deserve additional study. Real exchange rate movements affect the competitiveness of a nation's tradable goods sector and represent deviations from purchasing power parity. Seminal work by Balassa on real exchange rates and PPP violations relies on several underlying premises: (1) the traded sector possesses higher productivity growth and larger intercountry differences than the nontraded sector; (2) intracountry wage rates are equalized between traded and nontraded sectors due to labor mobility; (3) wages in the traded good sector are linked to productivity in that sector; (4) innovations in unit labor cost (wage compensation adjusted for productivity) cause corresponding movements in the relative price of nontraded goods. This paper examines general patterns and interrelationships among traded and nontraded productivity, real wages and relative prices for 14 OECD economies and extends the work of Strauss and Ferris [16]. We construct GDP price indices, productivity measures and real wage compensation rates for the traded and nontraded sectors for the years 1970-1990. The key finding is that differences in intersectoral wage growth appear to influence the relative price of nontradables more than differences in productivity growth. The results support the proposition that produc-