Volume 16 (1998)

The Regional Problem and the Break-Up of the State:
The Case of Yugoslavia
Časlav Ocić

The Nature and the Scope of the Regional Problem
Regional Policies and Changes in the Institutional Framework
Regional Development Levels: Grouping of Regions
Structural Change: Shift-Share Analysis
Efficiency: Shift-Share Analysis
Interregional Relations: Autarky
Some Other Results of the Regions' Development
Regional Development Costs: Ratios of Investment
Interregional Income (Re)distribution
Regional Convergence or Divergence?
Equality: The Failure of the Positive Discrimination Model
"The National Question" and Nationalism
Separatism: Economic and Political
A Long Journey from Utopia to Dystopia
Selected Bibliography
Data Sources & Documents
Appendix  (1)  (2)

Structural Change: Shift-Share Analysis
The study of the regional growth/sectoral structure relation involved various techniques of shift-share analysis. In the standard shift-share analysis, regional (economic) growth (in terms of various indicators such as: GNP, employment, fixed assets) was broken down into three parts: proportional hypothetical growth, structural shift and differential (regional) shift.
The results of the shift-share analysis regarding employment, classified according to a modified Boudeville typology,8 were interpreted from a purely economic point of view, i.e. on the basis of an assumption of an economic logic at work, which labour as a variable factor that accurately reflects both business trends and qualitative and quantitative changes in economic efficiency. According to this assumption employment can be considered as a general indicator of growth, structural changes, success or failure of the economy (whether national, regional or sectoral). Employment, however, is not an economic indicator only: it also reflects social, historical and political aspects of growth. Therefore, the results of an analysis of the components of regional changes in employment cannot be interpreted purely in classical economic terms. Underdevelopment and a relatively abundant supply of labour exerted a strong pressure on employment. Because of the rising expectations of the latently unemployed rural population, growth of employment is often accompanied by an increasing rate of (registered) unemployment. The number of people employed was constantly rising (with the exception of Vojvodina in the 1965-1970 subperiod) thanks to formal and informal channels of job procurement (nepotism, corruption). A high correlation between non-productive employment and development levels suggests that a considerable number of workers were not employed for production purposes. The political idea of creating a working class (by means of industrialization and urbanization) as the social base for new ruling elites undoubtedly affected the magnitude and the sectoral and regional dynamics of employment in the social sector. Under soft budget constraints, which characterized the business environment, the social function of employment prevailed over the function of an efficient economy.
Thus, for example, according to the modified Boudeville typology of regions, Montenegro, Kosovo-Metohia and Macedonia, respectively, were the most successful. The least successful were Slovenia and Croatia, with above average growth of employment in only one subperiod. However, this does not mean that Montenegro was economically more successful than Slovenia, but only that employment in the former grew more rapidly than in the latter. If, by chance, both of these regions had applied exclusively or predominantly economic criteria of employment, such a result could have indicated that Montenegro grew at a higher rate than Slovenia. Then it would have followed that one of the basic goals of Yugoslav regional policy (rapid development of all accompanied by faster development of underdeveloped regions) had been achieved. By formal standards, it was achieved in terms of employment, the growth of which was indeed more rapid in underdeveloped regions than in the developed ones. However, since employment was strongly affected by non-economic factors, it does not mean that the development of these regions was in fact more rapid.
By pointing to non-economic determinants of employment we by no means devalue the results of shift-share analysis: they do provide accurate information about actual changes in employment. These other, non-economic factors undoubtedly produced economic effects. The analysis identifies the components of regional changes in employment and the interpretation of its results should take into account both the non-economic and the economic context of change.
Similarly, the results obtained by shift-share analysis of fixed assets have to be interpreted in economic terms but without loosing sight of the social and political contexts. In terms of economic theory the change in fixed assets value is equivalent to the gross investment during the defined subperiods. Increased investment, if efficient, makes an economy successful. Under the conditions that prevailed in Yugoslavia, however, the very problem lay in the efficiency of fixed assets. First, the Yugoslav economy displayed all the characteristics of a relatively underdeveloped economy (e.g. a relative abundance of labour and a relative shortage of capital) and, second, it was a socialist economy: labour was intended as the pivot around which the system revolves, just as capitalism revolves around capital. In the Yugoslav case, the price of capital was below the price suggested by its relative availability, which under soft budget constraints inevitably resulted in inefficient investment. Thus, more investment did not mean a more successful economy.
When the results are reviewed in this specifically Yugoslav context, it becomes clear why the relatively least developed regions were by Boudeville's typology classified as the most successful ones: the value of their fixed assets grew at the highest rate. Thus, just as with employment, Montenegro, Kosovo-Metohia, and Macedonia were the most successful regions, while the least successful were Croatia and Slovenia. It should be stressed here as well that, despite the apparent paradox, the results of the shift-share analysis precisely describe the actual changes. They only show the effects of a regional policy reduced to mere transfers of money to underdeveloped regions: such a policy may (and did) secure an increase in the book value of fixed capital. Since a status of underdevelopment automatically guaranteed a steady and abundant inflow of cheap capital (through the Federal Fund for Financing the Accelerated Development of the Underdeveloped Republics and the Autonomous Province of Kosovo), there was a negative correlation between the size of inflow and the efficiency of capital use. Inefficient investment does not support economic development, but prevents it.
Assuming an "organic growth,"i.e. the domination of the market as the main factor of economic activity coordination, GNP can be considered as the general indicator of growth, of structural changes, the success or failure of an economy (whether national, regional, or sectoral). When market forces are suppressed by various forms of nonmarket coordination, and free enterprise by normative dirigisme and by standardized agreement among "economic agents,"there is no organic growth. Consequently, the growth rate of the GNP cannot be taken as a definite indicator of the economic success of Yugoslav regions.
In general, results of the shift-share analysis of employment, fixed assets and GNP, and, in particular, the results of a modified Boudeville typology of regions clearly suggest the following conclusions: (a) there is a negative correlation between the degree of development of a region and its success (performance); (b) crucial to a region's success is a differential shift, i.e. regional particularities are the key to the differences in their success; (c) the structure of regions is not a significant factor of the difference in their success, from which it may be concluded that regional structures do not significantly differ, i.e. that these differences are not so great as to significantly influence the differences in regional success.
In order to make these conclusions more distinct, the regions were ranked according to their success measured by the modified Boudeville typology of regions with respect to all three indicators: employment, fixed assets and GNP. The criterion for ranking was the number of successful or unsuccessful subperiods. The results of the ranking show that the observed interdependence is the most striking in employment, a bit less marked in fixed assets, and least in the case of GNP. Additionally, the differences between the most successful region and the least successful region are the most striking in regard to employment (the top regions have no unsuccessful subperiods, whereas the lowest ranking regions are successful in only one subperiod). The ranking of regions according to their performance in terms of employment growth resulted in the largest number of groups -six. Regional differences are narrower both in terms of fixed assets (there are four groups) and success (top regions have only one unsuccessful subperiod each, whereas the lowest ranking regions have two unsuccessful subperiods each). The smallest interregional differences were observed in regard to the GNP: there are three groups only, the top group consisting of two regions with two unsuccessful subperiods each and the lowest group consisting of four regions with three unsuccessful subperiods each.
A rather strong connection between the success of a region and its level of development in the case of employment and fixed assets (the less developed a region, the greater the increase of the two indicators) suggests that regional policy had a strong impact on the growth of production factors in underdeveloped regions, but also that it was primarily directed toward them. If we consider how important employment is for keeping the social peace, which is one of the major objectives of regional elites, it is obvious why this connection is the most striking in the case of employment. With respect to the growth of the GNP as an indicator of success, this connection is less noticeable. On the one hand, Kosovo-Metohia and Macedonia, the least developed regions, rank among the most successful ones, and Slovenia and Croatia, the most developed regions, among the least successful ones, still, on the other hand, the least successful regions also include Bosnia-Herzegovina and Montenegro, while central Serbia ranks among the more successful regions. This only shows that GNP growth is not merely dependent on the factors of production growth but that it is determined to a large degree by their usage upon which, in turn, the federal regional policy had no influence whatsoever.