Reallocation and restructuring: A generalization of the Balassa–Samuelson effect

Despite its strong theoretical position when it comes to explaining inflation in transition economies, the empirical findings of the Balassa–Samuelson (B–S) effect assign only a minor role to structural inflation – to the disappointment of analysts and policymakers. This article points to 3 theory-based contributing factors to these ‘weak’ findings and offers an alternative methodological approach. First, a short-term focus makes B–S prone to underestimating the magnitude of the productivity growth differential. Second, the conventional demand side CPI based definition of sectoral value added reduces the extent to which the productivity growth differential is passed through to inflation. Third, by ignoring the dependence between the 2 main B–S components, a further downward bias to the productivity growth pass through comes about. The key to our proposed alternative methodology centres on an endogenous relation between the productivity growth differential and sector sizes. Together with the long-run supply-side approach this allows us to capture inflation drivers that conventional B–S fails to incorporate. In our extension to the conventional B–S model a reduced productivity growth differential can be compensated by an increased productivity growth pass-through, or vice versa – with the effect of augmenting inflation pressure. Hence, the link between productivity growth differentials and the dynamics of structural inflation is shown to be more complex than previously assumed.

► We critique the Balassa–Samuelson Effect (B–S) and provide a generalization. ► The generalization allows to distinguish between 2 types of structural shifts: reallocation and restructuring – both driven by productivity growth differences. ► Introducing inter-temporal associations between productivity growth differentials and sector sizes allow for leads and lags in these relationships. ► We endogenize relative sector sizes – defining them as functions of sectoral productivity growth differences. ► A reduction in productivity growth differentials can now be compensated by an increase in productivity growth pass-through and vice-versa. ► The generalized model shows that the link between productivity growth and inflation is more complex than previously assumed.