Application of modified Klein regional model to cohesion policy analysis

The methodology of the regional models from the first papers published in the 60-ties has significantly progressed from the basic macroeconomic models through the models of growth and production to the regional microanalysis and spatial econometrics. On the other hand, practical application worldwide still stands far behind methodological backgrounds. The used methodological formulation of the regional model was originally formulated by Klein (1969). Empirical experience has shown that the formulation of the Klein model was too optimistic and significant statistical limitations led to the comprehensive modifications. Nevertheless, it was possible to design methodology of empirical model that estimates the components of the GDP by the expenditure at the regional level. This approach requires the utilization of a reference model on the aggregate national level (like Klein used the results of the Wharton model). The absence of any indicators (by expenditure approach) does not allow us to compare the achieved results with reality. Nevertheless this part of the paper shows that the balancing data on the regional top-down approach is feasible and provide additional information about regions of the country. One of the key limitations of the used models is the absence of statistical data in a sufficiently detailed structure. Recognition of regional data has a typical time delay t+2 due to processing aggregated and regional data, which are available only on an annual basis. It is therefore necessary to substitute the regional data (except for data on the labour market) with the semi artificial data based on regression estimates (partial information). One of the biggest problems in the terms of regional indicators represents the lack of information about the internal trade, transfers and public spending as well as the lack of information on regional prices and labour migration. Due to these problems, we had to apply a simplified version of model based on the regional Klein design. The next part of paper describes the creation of the primary and alternative cohesion scenarios using modified regional model. As a driving variable for scenario analysis was used indicator for regional investment. On the basis of the available statistical data, we created a regional model for Slovakia, together with an updated and revised econometric model framework for Slovakia producing a long-term forecast up to 2020 (together with scenario analysis). The part dedicated to the estimation of the components of the regional GDP by an expenditure approach has not yet been applied in Slovakia and should be considered as unique. Empirical results To verify the analytic ability of the model in addition to the basic method of top-down forecast, three alternative scenarios with different approach to regional policy were developed, that consist of the selected control variable resulting from the structure and composition of the model. The modified model can be estimated through bottom-up approach. Variant scenario number 1 (divergent scenario) was oriented to support the development of strong regions and regional centres. Variant scenario number 2 (uniform scenario) was oriented on equal opportunities and costs, i.e. each of the regions will be re-allocated the same share of capital formation. Compared to this basic scenario this one is slightly cohesive, since the volume of the capital from the region was based on a different basis. Variant scenario 3 (cohesive scenario) was oriented on a strong cohesion policy directed towards a rapid reduction of regional disparities, especially through redistribution the public finances. The assumptions about the behaviour of the model were confirmed in the application part, although surprisingly the most positive impact on the economic development and employment was in the cohesive variant. In the case of the cohesive approach and a greater investment in other regions, the production in the developed scenarios declined compared to the baseline scenario, but the decline in employment was negligible. The positive effect of a higher employment to the cost of a lower productivity (extensive economic growth) appears to be more suitable in order to achieve an additional economic growth in Slovak conditions. The secondary impact of a higher employment on the socio-economic indicators is an equally important factor, especially during the reduction of social tensions and social exclusion. Looking at the results they should be taken with some sound sceptics. Especially there are some assumptions about the perfect regional absorption capacity of an additional capital. Potential problems can be traced, for example in the process of implementation of the National Strategic Reference Framework (NSRF), and some of the original results are in contradiction to the presented Strategy. From an analytical point of view, it is clear that the projected development in various scenarios is strongly influenced by several factors. One of the greatest is the possible efficiency of subventions from the NSRF and the Structural funds (Radvansky, Frank – Ecomod 2010). Slovakia significantly lags behind in the use of these funds in the recent period. Given the need to maintain comparability of scenarios, the total investment volume remained the same in all considered scenarios. Considering the short length of the time series and the small number of regions, the standard statistical methods do not always achieve required characteristics and therefore the qualitative methods were used to measure the convergence and they have confirmed expected convergence trends of the selected scenarios. In the final section of the paper author addresses the possible development in the regional wages. The problem of assuming the same regional prices has been repeatedly mentioned throughout the work. In (Radvansky, Fuchs – Ecomod 2012) has been empirically confirmed that at least in terms of comparisons of income (wages), there are different price levels for each region. "Realistic" comparison of nominal and real incomes can be achieved when we take into account the information about different prices leading to a lower degree in the regional disparities. Statistically, there are no regional wages reported, while we consider them necessary to determine regional inflation given the results of the analysis. In the discussion we summarize the main problems of the chosen approaches, possible solutions and recommendations for further research. In terms of the selected approaches it is not necessary to describe and analyse the estimated macroeconomic forecasts in detail, but to verify the accuracy and the real possibility of the use of regional models. A major weakness of the chosen approach is the absence of spatial dependence, which enables to analyse the impact of the investment in the selected region, as well as the lack of feedback to the national model, which allows a recursive-dynamic approach to the regional model. See above See above