Joint Research Project
Working Paper Series

Work Package 5, Working Paper 2/8, September 2001

Variations in Farm Performance in Transition Economies: a case study of the Czech Republic

Sophia Davidova, Matthew Gorton, Belen Iraizoz[1] and Tomas Ratinger

Research Group of Agricultural Economics

and Business Management, Imperial College at Wye, University of London

This Research Project is Financed by the EU Commission's 5th Framework Programme (QLRT-1526)

1. Introduction

In the initial literature on the transition of the CEECs, debates concerning the restructuring of farming systems and whether certain structures have inherent productivity advantages over others featured prominently (Mathijs and Swinnen, Nelson 1993; Schmitt, 1993). This literature focused on the debate over whether collective farms and their successor forms would survive, and on the economics of farm size, which was linked to arguments about land reform and desirability of land restitution.

On the basis of theoretical arguments concerning the superior efficiency of family farming (Schmitt, 1991), many predicted the disappearance of co-operatives and that variations in productivity would lead to a wholesale transfer to individual farming. Empirical evidence on changing farm structures in the region indicates that the co-operative sector's share of total agricultural area (TAA) has shrank, but the complete collapse of co-operative farming, predicted by some, has not occurred. Moreover, others have argued that the superiority of individual family farming has been overstated and that the capacity of other structures of production to be efficient has been underestimated (Gorton and Davidova, 2001).

This paper revisits this literature by looking at the nature of variations in productivity and profitability between farms in the Czech Republic using data from 1999 and 1998. This allows the examination of variations in total factor productivity (TFP) and farm profitability for different structures, sizes and regions. The managerial and operational characteristics of clusters of farms identified for the Czech Republic are investigated in greater depth to draw out a clearer categorisation of farms, rather than just relying on size and structural type. These clusters form the basis of a discussion of the overall survivability of different groups of farms and, thus, the likelihood of future restructuring.

This paper is divided into seven sections. The next section discusses the literature on farm performance in the Czech Republic in the context of wider debates on de-collectivisation. Section 3 briefly discusses the results from previous studies on farm efficiency in the Czech Republic. The data used for the empirical analysis is discussed in Section 4. Section 5 details the TFP and profitability indexes produced for 1998 and 1999, compared with previous studies, considering variations by farm type, size and region. Section 6 looks at the variations between farms in greater depth by identifying clusters of farms, considering their managerial and operational characteristics. Relevant conclusions about the overall survivability of different groups of farms and the likelihood of future restructuring are presented in Section 7.

2. Debates on Farm Performance and De-collectivisation in the Czech Republic

At the outset of transition, Czech agriculture was widely perceived to be inefficient (Csaki and Lerman, 1996). In 1989, collective farms accounted for more than 60 per cent of Total Land Area (TAL) in the Czech Republic and state farms one quarter of the TAL. This preponderance of large farm collective structures sparked in the early 1990s a debate on two issues. The first was whether there was a clear superiority of one organisational type, namely family farms, over corporate structures (production co-operatives and various types of farming companies) and if so, whether the structures will align to those prevailing in the EU, namely family farms. The second aspect concerned the relationship between size and productivity in agriculture.

The question about “optimal farm structure” and “optimal farm size” has a long history in agricultural economics, in general, and in transition economies, in particular. When land reform strategies were being formulated at the outset of transition some argued that it was desirable to preserve large farm structures and pursue attempts to administratively impede farm fragmentation on the basis that smaller farms are less efficient. These authors tended to see restitution strategies, where they would lead to farm structures returning to the pre-war pattern of small-scale peasant units, as highly undesirable (Kanchev, 2000). Large farms have been seen to have advantages stemming from economies in using lumpy inputs, better administrative organisation, better marketing, access to credit and research and development (Hill and Brookes, 1993). In contrast, others argued that the large farms in Eastern Europe suffered from diseconomies of scale, so that land reform strategies must include proposals to reduce the mean size of farms (Koester and Striewe, 1999).

The superiority of family farms over other organisational types in agricultural production has been justified on the basis of the need to minimise both production and transaction costs. As the costs of supervision and monitoring of hired labour in agriculture can be high, it has been claimed that family farms appear to be the best suited organisational form as they minimise transaction costs (Schmitt, 1993). Focusing on collectivised agriculture in the centrally planned economies, Schmitt (1993) adds to this argument the principal-agent problem in producer co-operatives where the members can vote out the manager and, therefore, there can be disincentives for the manager to monitor workers. The formulated hypothesis (Schmitt, 1990; Hagedorn, 1994) is that “if the freedom of self-organisation is guaranteed, mainly family farms develop and survive, except for explicable exceptions, because they have low transaction costs” (Hagedorn, 1994: 5).

Since 1989, the Czech Republic has witnessed a growth in individual farming but not a rapid transformation to a predominance of family farms envisaged by some (Beckmann, 1996). The most dramatic changes were observed in 1992 and 1993, after farming entities were forced to adopt new legal forms. By 1999, individual farms, involving part-time farming, managed just over 20 per cent of TAL; the rest is managed in a corporate way. Corporate farming is organised in three different legal and management forms. Producer co-operatives are mainly successors of the previous collective farms, however, some of them are transformed state farms. Most of them still have outstanding liabilities to owners of co-operative assets. Limited liability companies have their origin in the privatisation of the state farms. At the beginning of the process, state farms’ assets were leased to small groups of people, normally involving the former farm managers (Ratinger and Rabinowicz, 1997). Gradually the non-land assets were sold to the lessees at favourable conditions with rescheduled payments. The joint stock companies have a large number of shareholders (a few hundreds). A portion of them has their roots either as state farms or as inter co-operative enterprises. However, some of the companies were created post-reform. In 1990’s several producer co-operatives were transformed in joint stock companies, as this allowed easier transactions with company’s shares. Thus, the Czech Republic has a set of legal and management forms that is still significantly different from the West European 'family farm' model.

3. Previous Studies on Farm Efficiency in the Czech Republic

Three previous studies on variations in farm efficiency in the Czech Republic have found mixed support for initial propositions drawn from the literature on the economies of size and structure debate (Table 1).

The expected pattern of sharply rising average productivity, reflecting scale efficiencies followed by diseconomies of scale at higher farm sizes, is present with economies of scale for arable farming up to 750 ha. Arable farms under 150 ha are significantly less efficient in all three studies. These studies on data for the mid-1990s point to small farms in the Czech Republic being relatively less efficient than in several other CEECs. Hughes (1998) argues that Czechoslovakia had a much less conducive environment for small-scale private farming than, for example Poland and Hungary, and this accounts for why small-scale farming appeared relatively less efficient in the early and mid-1990s in the Czech and Slovak Republics. The availability of external services for crop production, such as harvesting services and inputs for small farms, have been historically more developed in Hungary and Poland and the availability of such services are an important means of overcoming some of the sources of diseconomies of size.

For the Czech Republic, both Hughes (1998) and Mathijs and Swinnen (2000) found that individual private farms were significantly more productive than corporate farms for livestock farming but not crop production. Curtiss (2000) analysed crop production in the Czech Republic. She found that co-operatives performed better for wheat and rapeseed cultivation compared to individual farms but the latter were superior with regard to sugar beet production. It, therefore, appears that arguments that co-operatives or other forms of corporate farming are inherently less efficient, for all types of farming, compared to family farms is misplaced. Even for produce or types of farming where the average corporate farm is less productive than the average family farm, one still sees some co-operatives and companies which are on the frontier or registering high TFP scores (Hughes, 2000; Mathijs and Vranken, 2000). It appears that at least some corporate farms can solve the governance problems alluded to in the literature or that there are some types of farming for which such problems are less severe.

These productivity studies present some interesting findings but it is argued that further work is required on three counts. First, it is important to see if the trends identified for the early and mid-1990s reflect short-term characteristics of restructuring or are more long-lasting phenomena. For example, is the comparative inefficiency of small farms in the Czech Republic still present or has the switch to a more market based economy created a more conducive environment for small-farms allowing them to overcome their initial size disadvantages? Second, from the efficiency studies it is possible to identify farms which are relatively more efficient (e.g. on the production frontier or with a higher TFP index score) in a particular sample. However, this says nothing about profitability and return on assets in agriculture which will guide further restructuring in the sector. Finally, previous studies have focused principally on the farm size, structure and efficiency debate. The performance of farms will be shaped by many other factors than just size and ownership type, such as agri-environmental region, inherited debts and managerial characteristics. There is a need to consider a fuller range of factors in guiding our assessment of farm performance in the CEECs, identifying groups of farms with common characteristics. These points guide the profitability, productivity and cluster analysis presented in this paper.

4. Data employed in the productivity, profitability and cluster analysis

The paper utilises data from the Czech Republic's Farm Accountancy Data Network (FADN) which is administered by the Institute of Agricultural Economics (VÚZE). FADN, which is implemented in all existing EU states, aims to provide a detailed breakdown of revenues and costs incurred in agricultural enterprises based on a sample of bookkeeping farms. The Czech sample is surveyed annually in March and includes about 1,000 agricultural enterprises of physical and legal persons (Table 2). As in this paper the results for 1999 are mainly presented, data used for 1999 are discussed below.

The initial sample included 1,087 farms. After checking the individual data, 264 farms were excluded due to data inconsistency. Thus, the sample analysed included 823 farms. Fig 1 details the characteristics of the sample according to management form and average utilised area per management form. Considering management form, the largest group in the sample are the individual farms, 513 (62 per cent). Producer co-operatives are the second largest group, 154 (19 per cent). The rest of the sample is made up of 95 joint stock companies (12 per cent) and 61 limited liability companies (7 per cent). Due to their different history, different management forms have different average sizes. The most important difference is between the individual farms and the other management forms, with the former being much smaller than the successors of the previous state and collective farms.

Table 2 compares the characteristics of the individual and corporate farms included in the FADN sample (the 823 useable records for 1999) against returns from the Czech Republic's agricultural census. In the Czech Republic there are two main legal types of individual undertaking in agriculture: (a) trade law farmers, subjected to business regulations (Trade Law) like other full liability businesses and (b) solely operating farmers, with less strict regulations. Comparing the FADN sample with returns from the agricultural census, the former is biased to larger individual farms (in both the trade law and solely operating categories). For example, the average size of individual farms in the FADN sample is 134 ha compared against 18 ha in the agricultural census. In part this difference is derived from the fact that FADN is based on book-keeping records and so effectively includes only farms with commercial activities (although some farms also produce for own consumption). The FADN sample therefore excludes subsistence producers and this sector is not discussed in the scope of this paper. Comparing the average size of co-operatives and joint stock companies in the FADN and census samples, there are not significant differences although the mean size of limited liability companies in the FADN sample is larger than that recorded in the census.

The data for each farm in the sample contained total revenue and five cost items. The five cost items were: total labour costs for hired labour including wages and social insurance contributions, intermediate consumption (working capital), land rent, interest, and depreciation. In addition, land and labour were also given in physical units, annual work units (AWU) and hectares respectively. Labour was sub-divided into hired and family. Labour costs referred to hired labour only. Land area was given as a total and rented.

For the purposes of this analysis family labour was valued at regional farm unit labour costs (farms in FADN are classified in 76 administrative regions). The variation in the regional labour costs is between 62,000 CZK and 200,000 CZK per year (£ 1,250-3,900). As far as land is concerned, the regional rent was applied to the own land. In this case, agri-environmental regions were used as they better reflect the differences in land quality. The variation in rent is from 210 CZK/ha to 1,100 CZK/ha (from £ 4 to 22).