Regional development and regional planning subjects are very closely linked with the development projects in Turkey. The main applications are usually based on the elimination of disparities among regions. Also, regionalization movement has brought a new structuring process for candidate countries to the European Union. The main applications in this process can be simply based on the smaller region groups to collect more detailed and reliable data-information. This movement, of course, has to be considered together with developmental issues to solve regional inequalities among more and less developed regions by applying better regional development projects between specific region groups.
Firstly, this study explains Turkey's regional components which are regionalization, administrative structure, regional planning system, regional integration, and regional interactions. It has been also aimed to set out the regional development objectives in Turkey by focusing on the effect of new regional approaches in the administrative structure, sustainable regional development in the regions, interactions between project implementation processes and regional disparities (or parities), and integration of Turkey with the EU and the world regional policies.
As the beginning step of regional analyses, regional GDP per capita developments have been examined and discussed by Neoclassic Growth Model. As the findings, GDP per capita income convergence in geographic and NUTS (Nomenclature of Territorial Units for Statistics) level-1 and level-2 regions of Turkey was not clearly observed. These findings of regional divergence provide a remarkable explanation for the level of steady state per capita across regions. However, they are not capable to seek multi dimensional decision making for better steady state regions.
As the second step of regional analyses, to eliminate points, which convergence analyses can not determine, and to reach more detailed information about the sectoral conditions in the regions, some regional programs in NUTS level-2 regions have been examined by regional GDP per capita by sectors. This data, which is a more comprehensive than single income distributions, has been examined by the Gini coefficient in terms of sectoral disparities. Only in XG (government services) and XT (trade) decreasing tendencies are found.
As the third step of regional analyses, CDA (Canonical Discriminant Analysis) from different aspects (intraregional and interregional) is tested to show the effect of sectors in GDP per capita in the cause of convergence or divergence. It is observed that the intraregional divergence is increasing across underdeveloped NUTS level-2 regions. In developing NUTS level-2 regions, although some increases are observed in standard deviation in 1989, 1992, and 2000, there is a remarkable sigma-convergence tendency across developing regions after 1987. Lastly, a strong intraregional sigma-convergence is observed across developed regions from 1975 to 1995. After 1995, the sigma-convergence tendency is decreasing towards a divergence trend. Also, sectoral distances of the regions from the centre of axes (in CDA figures according to the years) and negative relations of the regions with the sectors are becoming dominant factors of divergence in developed regions. However, developing and underdeveloped regions have adverse tendencies. The reason of this is confirmed by findings of discriminant sectors in interregional convergence analyses. In a word, when the standard deviation decreases, the regions are getting centralized under the coverage area of discriminant sectors.
As one of the key purpose of this study to find a solution mechanism for regional disparities, the income distributions by sectors in 1991 (the lowest level of standard deviation in the regional divergence) are very meaningfu1 in terms of implementing the proper sector-oriented project(s). In this regard, XIC (less-imputed bank service charges) is primarily dominant field for developing regions and secondarily for underdeveloped regions. However, these regions need more than this because XIC is not a real sector and it refers to the interest income as 'the interests earned by financial organizations' - 'the interests paid by these organizations'. Naturally, developed regions based on the size of their economies have negative relation with this item. Finally, XG seems the most optimistic sector in short run for underdeveloped regions and XA (agriculture) is for developing regions. XTC (Transportation and Communication) can be also used for both regions as a tool in decreasing regional disparities in longer term.
To discuss today and evaluate the future of regional disparities, Turkey's solution efforts are also examined in terms of regional projects, regional programs, and a sustainable regional development.
First, in terms of major regional development projects, NUTS level-3 regions are combined in two groups; first group is the regions which are under the development project implementations, the other is the regions which are not. This is a restructuring process for project implementation that gives us an opportunity to test regional convergence with/without projects. At the same time, we would test the regions as regional convergence across western regions and eastern regions because the regions which are under the project implementations are mainly underdeveloped regions of Turkey located in eastern part and the others are more developed regions located in western part. As the findings, the regions with projects have convergence tendency after 1996. Even though it seems an optimistic point, the difference of standard deviation between 1976 and 2001 still very high compared to developed regions. The regions without projects do not show any convergence findings, but they do not have strong divergence trend, too. In this regard, regional development projects did not help underdeveloped regions to catch up developed regions, but merely started establishing a steady state growth across themselves after 1995. As GAP (Southeastern Anatolia Project), which is the biggest regional development effort ever undertaken by Turkish Government, started in the 1980s very earlier than the other regional development projects (mainly started after the mid-'90s) in Turkey, it could not provide any positive effects for underdeveloped regions and neighboring regions because the high divergence trend among 1980 and 1995 signifies that the effect of GAP project was very low during this period.
Second, in terms of major regional development programs, related sector parameters of GDP per capita have been calculated by the regression analysis through (balanced) panel data. It is observed that there is a remarkable sectoral disparity among program regions as well, and the sector priorities of the programs are not sufficient to bring any long-term solution for the regional disparities as similarly observed in GAP which is very far from expectations.
Finally, even though the scope of regional programs has been currently transformed
to reach the basic regional socioeconomic standards of the EU by applying new
cooperative regional programs, Turkey needs to reconsider its regional efforts from
regional and/or multiregional perspectives for decreasing the disparities as much as the
development of the regions, severally.