Dual Inflation Under the Currency Board: The Challenges of Bulgarian EU Accession

William Davidson Institute Working Paper 487

ones²⁵. In this study we are not making a separation between those groups but rather put
them in a group, which we consider as tradable. The non-tradable group comprises
primarily of services.

2. Measuring the BS effect

As we have already pointed out (in section 2) before the direct econometric
estimation of BS effect existence, we make a preliminary verification of its basic
analytical components.

In principle when we speak of sectoral productivity, we need to look at total
factor productivity, but this requires estimation of the production function, which is
limited to some extent given the short period of time available. We follow the literature²⁶
in measuring instead labor productivity, the ratio of output to employment. We use a
Hodrick-Prescott (HP) filter, econometric technique, to take out the cyclical and seasonal
fluctuations from the productivity values in both sectors and thus obtaining their trends
(figure 4). As expected, the productivity of the tradable sector (industry) is higher than
the one in the non-tradable sector, which is a prerequisite of the presence of the BS
effect.

²⁵ See AEAF (1997).

²⁶ Most of the studies dedicated on the BS effect involve labor productivity and it should be clearly
stated that this “partial” productivity could produce different results. For more details, see also Backe
and al., 2002.

18

<<   17   18   19   20   21   22   23   >>

More intriguing information