Naslov
Growth and convergence across Croatia: evidence from county-level data

Autori
Šergo, Zdravko

Vrsta, podvrsta
Ostale vrste radova, rukopis

Godina
2005

Ključne riječi
unconditional; Solow and augmented Solow regression;

Sažetak
The neoclassical Solow-Swan growth model predicts convergence. The idea of convergence can be understood in two different ways. The first is in terms of the level of per capita income, absolute convergence. Poor counties will grow faster than rich counties so that output per capita across all county-split regions will eventually be the same, even though some may start out way behind. The second is convergence to the steady-state level of income, conditional convergence. Per capita income in a given region converges to the steady-state value determined by that region's characteristics. Regions need not all reach the same level of per capita output. We use Croatia county-level data, to study growth determination and measure the speed of GDP per adult convergence. County-level data are particularly valuable for studying productivity convergence because they allow us to study a sample with substantial homogeneity and exceptional mobility of capital, labour and technology. The main aim of this paper is to try to find evidence of convergence in GDP per adult across the 21 Croatia counties, 1990-2003, to test convergence we running unconditional, Solow and augmented Solow regression. Using data from County Statistical Yearbook 1990 & 2003 we try to examine whether the Solow growth model is consistent with domestic variation in the productivity. Nowadays, the economic convergence is an issue that is occupying a lot of the economic research community, mainly in Europe. One of the most important questions around the European enlargement (Croatia is the country which has strong aspirations for accession) is the economic convergence. However, in order to have a sustainable growth in the European countries, it is necessary that one can assist to economic convergence between countries, and also within countries. In this paper we address the problem of convergence in Croatia because we consider that the convergence of the Croatian counties is a precondition to sine qua non in the strategy of Croatian growth. In the socialist self-management phase of development of the former republic of Yugoslavia the proclaimed strategy of balanced development of all the republics and autonomous regions in Yugoslavia was in fact, by different financial transfers, a subsidized policy of convergence income per capita. With the break up of Yugoslavia and the creation of sovereign Croatian state the difference in relative productivity between particular regions in Croatia is increased thus the transition towards capitalist economy marks a drastic disparity in the productiveness of labour. The relative disparity in the productiveness of labour was 3.45 in 1991 - in fact the realised GDP per adult in the city of Zagreb (always maximal in the given structure) was 3.45 times higher than in Ličko - Senjska County. In 2003, the GDP per adult of the city of Zagreb was even 9.48 times higher than that of Vukovarsko-Srijemska County. As a consequence of the focus on output decline of post-socialist countries in the early transition period, there has been very little work done on the relevance of the standard growth factors framed in the neoclassical growth model (Solow, 1956 ; Mankiw, Romer and Weil, 1992) in explaining differences in growth performance in transition countries. As far as we know there is not a methodologically similar study on the basis of cross-sectional county data, which elaborated the problem of convergence in Croatian counties during the transitional period. This paper seeks to reconcile the growth empiric's technique of Mankiw, Romer, and Weil with the empirical technique of Barro and Sala-í-Martin through the application of an econometric analysis on the regional database of Croatia covering the 1991-2003 period. We test if there has been σ convergence across Croatia's counties. Because GDP per adult person dispersion increased from 0.14 in 1991 to 0.23 in 2003 ; we find that the cross sectional variance has fluctuated a sustainable in the 13-year period from 1991 to 2003. This rise reflects the war shock to regional economy during the nineties. For the vast majority of counties, and as a result as well as for the full Croatia, σ -divergence is present. A likely explanation for this result is that war affected counties tended to be poor counties and richest counties much more richer (some counties such as Vukovarsko-Srijemska suffered the most from the war destruction even though it was relatively developed before the war on the other side we have an example of accelerated development of the city of Zagreb). Since the divergence is indicated by positive sign β in OLS it is not surprising that the speed of divergence given by parameter  in NLLS is significant and positive number. The rate of divergence in encompassing transition time is approximately 0, 54% per year among counties of the Croatia. We didn't find hard evidence of conditional growth in our econometric model. The conditional augmented Solow growth model result are theoretically excepted and these results give us confidence that our data can be used to address other issues concerning regional economic development in Croatia. The significant value of human capital results suggest that 10% rise in secondary school enrolment to adult person ratio will raise the output-to-adult person ratio by almost 1, 34 % (1, 45% with dummy) in the long run. The implied rate of convergence is 1, 27% per year in augmented conditional Solow growth model (about 4% with war dummy) is consistent with the theoretical expectations.

Izvorni jezik
Engleski

POVEZANOST RADA