Introduction
This approach emphasizes the importance of socio-economic indicators in determining the level of regional development. Welfare level is observed and measured by reasonably chosen a set of indicators which consist of the Gini Coefficient, Poverty Rate (Regional Basis), Unemployment Rate, GDP per Capita, Budget Revenue / Expenditure Rate. These indicators are chosen to be able to compare regions in Turkey by the main socio-economic indicators, which represent welfare.
For Multidimensional Scaling Analysis, the indicators are defined in terms of welfare. The selected individual indicators are most representative indicators for the regions of the country’ economy. The Gini Coefficient, whose main advantage is that the measure of inequality is determined by the ratio, not a variable that represents the majority of the population, such as GDP per capita. The Gini Coefficient is also used to measure income inequality. Poor countries (countries with low GDP per Capita) have Gini coefficients that varies from low (0.25) to high (0,71) while rich countries have generally low Gini coefficient (up to 0.40). In Turkey Gini Coefficient measure varies from 0,35 (Southeastern) to 0,51 (Marmara).
Poverty level is the level of the population with family income below the officially defined threshold, which varies depending on the size and composition of the family. The standard definition of long-term unemployment is determined by the number of unemployed with a continuous period of unemployment within a year or longer (52 weeks or more) and this figure is expressed as a percentage of the total workforce in the region.
The cost of unemployment is defined as the social and economic parameters, such as increased crime, alienation, the cost of unemployment, loss of output and tax revenues. Gross domestic product (GDP) measures the economic aggregate of the country (and regions), while GDP per capita assesses the degree of prosperity (in the country or particular region).
In the economic circle, in the most cases is taken GDP per capita as an important indicator of economic development stage of the country (and the region). Other important economic variables are the unemployment and inflation. Inflation is a component of growth in the general price level of goods and services in the region. Optimal inflation rate depends on many factors, and is likely be different for each region in the country. If inflation is too high, real incomes and economic activity is declining. Budget Revenue /Expenditure Rate also vary significantly from one region of the country to another.
Material and Method
In our analysis, we used data from the State Planning Organization (Turkey) as the main economic indicators. We have taken the following indicators of "GDP per capita ", "long-term Unemployment Rate", "Budget Revenue / Expenditure Rate", "Inflation Rate", "Poverty Rate" and the "Gini Coefficient". These indicators in our view demonstrate the relationship between the regions in Turkey, as well as determine possible similarities and differences between them.
As a result, the perception of the map indicates the relative position of all elements. The researcher interprets the basic dimensions in a way that best explains the layout of the map, in particular as regards the basic theoretical foundation. In addition, MDS method allows standardized (transformation) analysis of data collected at different scales. This study is based on the data standardized using Z score transformation. In this analysis we used multidimensional scaling procedure in SPSS 19,0 and created a matrix of raw data using the Euclidean distance measure.
Results
Results of Multidimensional Scaling and Distance Between Variables. It is advisable to MDS analysis was used to determine the Stress Value of statistics at a level close to zero. Compatibility of the distance to the original configuration based on stress values is expressed as follows:
Euclidean distance model
When we evaluate the data in two dimensions, from the standpoint of the welfare of the region, the most significant variable is the inflation rate. Followed, in order by "Poverty Rate", "Gini Coefficient", "Unemployment Rate", "GDP Per Capita" and "Budget Revenue / Expenditure Rate". When we looked at it on the second dimension, it can be argued that as a secondary choice group, are important the following variables: Inflation Rate, GDP Per Capita, Poverty and Gini Coefficient. But these variables can be evaluated as secondary choice variables. In the analysis of well-being and development indicators of regions, we concluded that the "Poverty Rate", "Gini Coefficient" and "Inflation Rate" are important variables in both dimensions.
Figure 2. Results of Multidimensional Scaling Analysis According to Distances Between Regions in Turkey (Case:1 Marmara,Case:2 Aegean, Case:3 Mediterranean, Case:4 Central, Case:5: Black Sea, Case:6 Eastern, Case:7 Southeastern).
Euclidean Distance Model in terms of cases (Regions)
Derived Stimulus Configuration
Euclidean distance model
Conclusion
The analysis revealed the discrepancy of socio-economic indicators of Turkey's regions welfare level. But all indicators of the regions are in the same dimension. We can only assume that discrepancy in the indicators could be more if compare variables between countries (for example, European Union and Turkey) as well as cross-country analysis will be seen more clearly the differences with respect to inflation, the Gini coefficient and poverty levels.
References
Akai, N. and Sakata, M., 2002. Fiscal Decentralization Contributes to Economic Growth: Evidence from State-level Cross-section Data for the United States, Journal of Urban Economics, 52 (1), pp. 93-108.
Hair, J., Anderson, R., Tatham, R., Black, W., 1995. Multivariate Data Analysis with Readings. Englewood Cliffs, NJ: Prentice-Hall.
Howard, McCain, “Visualizing a Discipline”; Moya-Anegón, Jimé- Figure 2. SOM map nez-Contreras and Moneda-Corrochano, “Research Fronts,1998. http://europa.eu.int /comm/eurostat/Young, F. W., &
Kruskal J., Wish M., 1978. Multidimensional Scaling. Sage Publications.
Kinnucan, Nelson, Allen, 1989. Statistical Methods in Information Science Research. Leung, K.& Bond,M.
Leung, K., Bond, M., 1989. On the empirical identification of dimensions for crosscultural comparison. Journal of Cross-Cultural Psychology, 20 , 133_/151.
