engleski

Testing Gibrat's Law in Croatia: Comparative Analysis for Eastern Croatia

According to Gibrat’s law, growth rate of the firm is independent of firm’s size. This paper empirically tests Gibrat’s law on the panel data of Croatian firms during the period from 2006 to 2015. Moreover, we test for any structural differences between full sample and subsample of firms, the subsample comprising the firms that are registered in at least one of the five Slavonian counties. The aim of this paper is to show the importance of taking into consideration regions within countries when testing the Law. Our data comes from Bureau Van Dijk’s Amadeus database and originally included more than 80, 000 firms. We use panel data analysis to disentangle effects of various firm-level variables on the growth rate of the firms. We use sales of the firm and the number of employees as proxies for the size of the firm. Moreover, we include variables such as the age of the firm, growth opportunities (proxied by ratio of intangible fixed asset to total fixed asset) and solvency ratio, in order to control for variables other than the size that affect the growth rate of the firm. Simple statistics showed that average growth rate of the small firms is almost 2 p.p. higher than the one for medium and large firms, econometric model (fixed effects (FE) model) showed that there exists statistically significant negative connection between size of the firms and their respective growth rates. That is, we reject Gibrat’s law for the case of Croatian firms in general. Econometric analysis on subsample of firms from Eastern Croatia showed similar results (size of coefficients and significance) for all variables except the age of the firm. We find that there exist significant structural difference in “age of the firm-growth rate” link between firms in Eastern Croatia as opposed to of the rest of the firms in Croatia. While age of the firm is not significant at all when we estimate the FE model on the whole sample, it is highly significant and positive for the firms in Eastern Croatia. We offer several explanations for that puzzling finding.

Gibrat's law, Croatia, panel data analysis, fixed effects

nije evidentirano