摘要：Transition from the planned to the market economy involved profound changes in Ukrainian economy. Before the liberalization of prices in 1992, milk processing in Ukraine was concentrated in regionally distributed large state milk processing factories. In the meantime, the dairy sector has been privatized. However, the emerging privatization forms were heterogeneous. That is why different kinds of market conduct and, consequently, market performance can be expected, depending on ownership and incentive structures involved. Particular macroeconomic conditions have led to agricultural farms and milk processing factories dragging in conformation to the new situation. The necessary economic reorganization has caused a considerable decrease in milk production. Therefore, the capacity utilization of milk processing industry in the 1990s has strongly diminished. Due to this reason, the increase in average production costs has lowered the sectoral profitability to a significant extent. Although at the beginning of the market-oriented transformation changes and uncertainties of this nature were to a large extent similar for agricultural and food sectors in different Central and Eastern European countries, so far only little is known about market conduct of milk processing enterprises. The descriptive analysis provides evidence of limited market transparency and an influence of regional administrative authorities on regional trade with raw milk via trade restrictions. In such conditions milk processing enterprises were able to gain a regional monopsony or oligopsony position on the raw milk market, and, hence, there is reason to presume that they exercise market power on that market. In addition, in 2002 the Antimonopoly Committee of Ukraine detected price cartels among milk processing enterprises in some administrative regions of Ukraine. This may be interpreted as a prerequisite for the exercise of market power on the part of milk processing industry. On the other hand, because of low concentration of milk processing industry at the national level, the assumption about competition cannot completely be dismissed. Moreover, the low capacity utilization of milk processing industry and, consequently, the competition for raw milk stand against the hypothesis about existence of market power. Since the 1980s, in studies analyzing imperfect competition on agricultural and food markets, quantitative approaches have increasingly been based on microeconomic theory and on concepts of New Empirical Industrial Organization (NEIO). The NEIO approach was applied in numerous studies on market structure and pricing on agricultural and food markets. The applications, however, have more or less been limited to developed market economies such as the USA, Canada and others. On the other hand, due to the transition from a planned to a market economy, in countries such as Ukraine, there is also reason to expect market conditions on agricultural and food markets which considerably deviate from the model of perfect competition. For the analysis of the market structure and pricing on the Ukrainian raw milk market, a comparative static structural model has been built. The model consists of a short-run supply function for raw milk on one side and of the first order condition for profit maximization for the raw milk demand of milk processing enterprises on the other side. An important advantage of the market structural model is that the existence of market power can be explicitly tested. For econometric analyses, aggregate industrial (market) data at both, national and regional level as well as data at the level of the individual firm can be used. Before the specification and estimation of the market structural model, as a first step the production technology of the milk processing industry and the raw milk supply in Ukraine have been econometrically investigated. The results of the econometric analysis of production and supply functions were used to analyze the market structural model. Different special cases of the production function for the Ukrainian milk processing industry were econometrically estimated based on the transcendental logarithmic (translog) functional form showing differences both, in the scale elasticity and in the rate of technological progress. The estimated scale elasticity in the special case of a Cobb-Douglas production function is statistically significant and amounts to 1,4. Due to this result, there is evidence of increasing returns to scale in Ukrainian milk processing industry. Results considerably diverging from the Cobb-Douglas case are obtained from the estimation of production elasticities based on a non-restricted translog production function. Negative production elasticities are obtained for the production factors labor and capital. This can be associated with under-utilization of capacities of Ukrainian milk processing plants. Furthermore, the results have to be interpreted against the background of the change of social and institutional basic conditions in Ukraine. Therefore, the results obtained from estimating the neoclassical Cobb- Douglas production function, have to be interpreted carefully. Consequently, with regard to the market structural model for the Ukrainian raw milk market, it seems to be problematic to specify the production technology of the milk processing industry as a neoclassical production function. This can be attributed to the specific situation in the industry during the transformation process. Hence, the non-restricted translog production function has been used within the scope of the market structural model. In numerous studies, the estimation results for agricultural supply functions were presented. The functions have been estimated on the basis of either single equation models or multiple equation models as a part of a complete system of supply and factor demand functions for the agricultural sector. However, in most cases, the modeling was carried out for both, single equation models and multiple equation models, assuming perfect competition. With regard to the modeling of the market structural model, some aspects and problems about derivation of supply functions from a variable profit function are discussed, and an overview about empirical supply and factor demand functions in the NEIO-studies is given. Afterwards estimation results for several versions of the supply function for raw milk of the Ukrainian agricultural sector are discussed. The main objective of the econometric analysis, however is to measure the market power of milk processing enterprises on the Ukrainian raw milk market, within the analytical framework of a market structure model which is based on the theoretical concepts of NEIO. The model serves to estimate a market power parameter and to test econometrically for the existence of market power. It is specified taking into account the estimation results for the production function of the milk processing industry and the supply function for raw milk. The results of the econometric estimation of the model do not show evidence of the exercise of market power by Ukrainian milk processing industry over the sample period from January 1996 to December 2003. Since aggregate data on the national level is used in the estimation, the results relate only to the overall raw milk market in Ukraine. In terms of market structure of Ukrainian milk processing industry at the national level the results are somewhat plausible because the Herfindahl-Hirschman index (for 2001) is 7, which is relatively small compared with other countries. Moreover, the results of the descriptive part of the work show that the market structure in the Ukrainian milk processing industry has not changed over the sample period. However, it is necessary to note, that the available industry data at the national level are too highly aggregated for a final decision about the existence of market power. Considering the structure of the milk processing industry on the regional markets and also the basic conditions on the market for raw milk, the existence of market power cannot be excluded. The descriptive analysis shows that in 8 of 25 regions the Herfindahl-Hirschman index is larger than 200. For a determination of the relevant market boundaries the regional market for raw milk seems to be of prime importance for the analysis. Therefore, in order to make final statements about the existence of market power in future studies disaggregated industry data should be used for the estimation of market structural models at the regional level.
