The Impact of Public Capital Stock on Regional Convergence in Turkey

Abstract

The effect of public capital on regional output and private sector productivity has been the subject of considerable research in the field of regional development literature. However, there have been only a few studies that develop linkages between public capital and regional economic convergence. This study examines the dynamic effects of public capital and transportation capital stocks on output per capita in terms of regional convergence in Turkey at NUTS 2 level. A conditional convergence model based on per capita gross domestic product and per capita public capital and transportation capital stocks is estimated using the panel data set for the Turkish regions for the time period of 1980–2001. The results reveal that public capital has a positive and significant effect on output per capita and thus on regional convergence in some of the models in the Turkish regions,. However, the transportation component of public capital stock has a negative effect on regional convergence in all models employed in the study. This implies that transportation capital stock leads to larger regional disparities between the Turkish regions.

Acknowledgements

We thank the Scientific and Technological Research Council of Turkey (TUBITAK) for the financial support of this project under grant SOBAG-104K112. We also thank the anonymous referees for comments and critiques. However, the authors are responsible for all remaining errors. Previous versions of this paper were presented in Turkish Economic Association International Conference on Economics, Ankara, 11–13 September 2006 and in 47th Congress of ERSA, Paris, 29 August–2 September 2007.

Notes

We need to give a clear delimitation of spillover effects, distributive effects and convergence effects. Spillover effects of public capital basically mean the effects of public capital installed in one region or the other regions (see Deliktaş et al. (2009) for details), while distributive effects (often referred as equity or cohesion effects) are one of the a wider range of impacts of public capital (Lopez et al., 2008). Convergence effects of the public capital mean the effects of public capital on the development of certain economic variables such as income per capita, growth rate and total factor productivity over time between the regions.

See, for example, Boldrin and Canova (2001) and Vickerman et al. (1999) for the theoretical background regarding public transport infrastructure in detail.

Therefore, like in Shioji (2001), regional growth and regional convergence are used synonymously in this study.

The share of transportation expenditures is around 25% on average of total public investment expenditures between 1998 and 2004 (see public investment reports, www.dpt.gov.tr).

The construction of sectoral variables are as mentioned in Shioji (2001). The data include the sectoral variables for agriculture and manufacturing. The share of sector in output of that region in 1990 and the average output per worker across all sectors at the national level between 1980 and 2001 are obtained from the TURKSTAT.

We have conducted the same analysis with another dispersion statistic, i.e. standard deviation of per capita income for sigma convergence, and obtained similar results.

The sectoral variables are found to be insignificant and their inclusion had a very minor effect on the other regression coefficients. Therefore, we have omitted that variables from the estimations.

λ convergence rate and φ steady-state effect of public capital are derived from equation (8).

The test statistic is F _m,n−k  = RSS_r − RSS_ur)/m)/(RSS_ur/n − k), where RSS_r and RSS_ur stand for restricted and unrestricted residual sum of squares. m, n and k denotes number of restrictions, number of observations and number of estimated parameters, respectively.