Disaggregated Spending and the Productivity Bias Hypothesis

What is it about?

According to the productivity bias hypothesis countries have a tendency for real appreciation in their domestic currency as a result of a productivity shock and the tendency is more pronounced in the non-tradable sectors. Balassa (1964) examines this thesis for 12 OECD countries in a cross section framework. This paper reexamines this empirically using a disaggregate data of consumption, investment, and government expenditures in a panel regression set-up. Using five different panel specifications and controlling for country specific heterogeneity, time specific heterogeneity, and openness this paper explores that the bias is more prominent in government sector compared to other component of aggregate spending.

Featured Image

Why is it important?

Productivity bias hypothesis is the most widely used theory for examining the long run behavior of real exchange rates. Since its inception many researchers have used different econometric models or methods to examine the validity of this hypothesis. The existing literature is this line is exhaustive but there is room for empirical reinvestigation of this hypothesis by using disaggregated price across different category of aggregate spending. This kind of alternative framework adds to the existing literature by revealing the differences in productivity bias across tradable and non-tradable sectors directly by examining it at four different categories such as overall price level, price level for consumption, investment, and government spending in a panel regression set-up. With this end in view, in this paper, we use a standard econometric model that has been used by researchers in this area and examine the productivity bias hypothesis in five different panel specifications and found that no matter which case we consider in most of the time productivity bias is mostly observed in government spending which has a larger fraction of non-tradable goods compared to other component of aggregate spending. To make the result comparable to other studies and to get additional insight we perform the same estimation for different country grouping in the light of Heston, Summers, and Aten (2001) and conduct the same analysis with and without openness as the control factor. No matter which case of panel we consider in most of the cases government spending displays the productivity bias mostly. This kind of result suggests policy makers to be aware of the situation that as the country becomes richer or more productive the public sector becomes less competitive compared to other sectors and special care must the given to this sector to retain its competitiveness.