Credit allocation and farm structures in the Czech Republic, 1993-1997.

Bezemer, Dirk J.

INTRODUCTION

During the reforms in Central Europe, the emergence of market-oriented family farms, replacing communist-type wage-labour farms, has been limited compared to initial expectations. Although the early view was that `privatization in ... agriculture mainly concerns the breaking up of large units ...' (World Bank, 1995, p. 2), as early as in 1994 it could be noted that `already now it is clear that the process of farm restructuring ... is taking a course which appears to be different from the original expectations of many Western European observers.... It is remarkable that farm enterprises ... choose to reorganize as whole entities, without dismantling the collective structure' (Csaki and Lerman, 1994, pp. 566, 573). In most Central and Eastern European countries except Albania, such structures still cultivate considerable shares of agricultural land. This share is 89% in Slovakia, 62% in the Czech Republic, 48% in Bulgaria, 46% in Hungary, 37% in Estonia, 33% in Romania and Lithuania, and 5% in Latvia (Lerman et al., 2002). (1) In the (other) countries of the former Soviet Union, market-oriented family farming is typically even more marginal (see, for instance, Prosterman et al. (1998) and OECD (1998) for figures on Russia).

Various reasons for this development have been suggested (Sarris et al., 1999; Mathijs et al., 1999; Bezemer, 2002b). Oft-mentioned is the lack of functioning credit markets, which is a problem particularly in the agricultural sectors of transition economies (for a detailed discussion of the reasons, see Bezemer, 2002a). De novo and relatively small enterprises, such as most family farms are, are thought to be typically more sensitive to this than established farm structures. This link between the limited role of commercial family farming and hampering credit markets has been suggested in several empirical studies (Wolz, 1996; Swinnen and Gow, 1999; Davis et al., 1998; Shrieder and Heidhues, 1998; Pederson and Khitarishvili, 1997). More limited credit allocation may be one reason for the limited importance of individual farms in the Czech Republic and elsewhere in the economies in transition. A level playing field in this and other areas of farm operations would facilitate competition between different organisational alternatives in agricultural production, which, in turn, might lead to further improvements in allocative efficiency (as suggested by eg Mathijs and Swinnen, 2001). This level playing field was, and is, the official aim of structural policies in most transition economies--see OECD (1995, p. 90) and Ratinger and Rabinowicz (1997, p. 96) for the Czech case. Reality, however, may be different. In this paper, the relation between credit allocation and farm structures is investigated on the basis of primary farm-level data for the case of the Czech Republic in the mid-1990s.

BACKGROUND INFORMATION AND DATA

The organisational alternatives in farming in the Czech Republic, as in most of Central European agriculture, are what have been called `corporate' farms and `individual' farms (Sarris et al. 1999, p. 310). Corporate farms are successor organisations to collective or state farms, and in the Czech Republic they include cooperative farms (zemedelska drusztva), joint stock businesses (akciove spolecnosti), and limited liability companies (spolecnosti s.r.o.). They are defined by the corporate separation between farm ownership, control over the production process, and implementation of production tasks. The typical governance structure of Czech corporate farms consists of several hundreds of owners, up to 10 managers, some administrative and technical personnel, and between some tens and a few hundred workers. Czech cooperative farms are usually larger in terms of acreage and workers, and less profitable than are joint stock or limited liability companies (Csaki et al., 1999, p. 36).

Individually operated, or, for short, individual farms, are owned, managed, and operated by a limited number of people, usually united in one (extended) household. Their legal form is normally that of a `physical entity', which is usually the `physical person' (as distinct from a corporate structure) of the entrepreneur. The individual farm is sometimes the successor of a state farm, but typically newly founded as a production entity during the transformation, its owners working land that was acquired through restitution and lease. Wage labour is used on a very limited, often part-time or temporary basis, if at all. The production focus is on crop growing in which there is considerable product specialisation, while activities in animal husbandry are limited and non-farm production is usually absent. Also, individual farmers are more often supplemental-income farmers.

The limited importance of individual farms for market-oriented agricultural production is due not so much to their number as to their size. In the Czech Republic in 1998, the survey year, there were 2,208 joint stock/limited liability farms and 1,256 cooperative farms. In comparison, there were 92,845 persons registered as owning land and with the legal form of `physical person' (podniky fyzickych osob) in the Czech `Register of Economic Subjects' in 1998. However, `only' 22,971 persons were registered in another record, the `Agricultural Register', as producing food (ie for the market). The rest of the landowners plausibly either worked their land as gardeners, largely for home consumption, or rented it out to other people or to a farm. These 22,971 individual farmers had an average farm size of 36 ha, but over half of them worked less than 10 ha and only 6% used over a hundred hectares. Between them they worked 24% of agricultural land. The bulk of the remaining land was occupied by the 3,464 corporate farms, all working over a 100 ha, except for a minority of about 10% (Ministry, 1999; Csaki et al., 1999, p. 29). In short, individual farms were, and are, numerous, but very small, and therefore of limited importance for market-oriented farming compared to corporate farm structures.

This is the setting in which two surveys were implemented in 1998-1999. The survey questionnaires were developed on the basis of fieldwork in the Czech Republic during 1997 and 1998 (see Bezemer (1999) for findings). One survey addressed members of the Czech Association of Private Farmers (SSZ), that is, owners/operators of individual farms. This sample frame, implying a member of the SSZ, ensured that the sample was not saturated with extremely small farm structures (which are often actually gardens and hobby farms). Still, in view of the sectoral structure, a distinction was made in the sample between `professional' and `other' individual farmers. `Professional' individual farms were defined as farms where the owner both derives more than three-quarters of his or her income from farming and devotes over 40 hours weekly to it.

A parallel survey targeted management of corporate farms, and was conducted in cooperation with the Czech-Moravian Union of Agricultural Cooperatives (representing cooperative farms and farm companies with limited liability or joint-stock structure). Data on 193 individual farms (70 `professional' and 123 `other') and 69 corporate farms (29 farming companies and 40 cooperatives) were collected.

Data collected in the survey included: information on developments in farm size, structure, and governance during transition; experiences in the decollectivisation and restitution process; the nature of farm operations on input and output markets; farm profitability; and receipt of credit and subsidies, loan allocation criteria, and loan maturity and collateral. Most survey questions were identical for individual and corporate farms; some were specific to farm structure. A number of questions were retrospective, referring to the entire transition period to date (1990-1997) in connection to changes in farm structures, and to the last 5 years (1993-1997) in connection to credit market experiences.

During that period, Czech credit markets in general were characterised by three main problems. First, the bad debt problem had never been solved; debts accumulated during the central planning era, or shortly thereafter, of which repayment was long overdue, had not been systematically restructured. Only in 2000 did the problem begin to be addressed. By then, nonperforming loans accounted for 26% of GDP. Second, liquidation procedures in particular and court proceedings in general were very slow, and creditors could not effectively move against defaulting debtors. Third, capital market regulation was deficient and was not improved until 1998 (World Bank, 2000, pp. 4, 6, 37, 67). Each of these problems originated in the former economic system (Bezemer, 2002a).

In the segment of the Czech credit market servicing agriculture, there were four types of participants during the years of observation: individual farms, corporate farms, banks, and the state, represented by a credit support fund. A relatively large number (over 40) of the Czech commercial banks operated in agriculture, the most important being the majority state-owned Komercni Banka (KB). Through this bank went 38.5% of the loan guarantees and 34% of interest rate subsidies in this period. Second in importance was Agrobanka, a former state bank branch, handling only 15.6% of guaranteed loans and no subsidies. Agrobanka, although originally designated to become the main agricultural bank of the country, was planning to withdraw entirely from agriculture at the moment of the survey because of financial difficulties and a foreign take-over. The continuing bad debt problem was reflected in the fact that total outstanding credit to agriculture in 1998 (32,903 million, Czech Crowns or 1.8% of Czech GDP) was for a large part (15,373 million, Czech Crowns) long-term (Csaki et al., 1999; Pederson and Khitarishvili, 1997, p. 10). In 1998, the exchange rate was 32 Czech crowns to the US dollar (EBRD, 2001).

Subsidised credit handled by these banks originated with the Support and Guarantee Fund for Farmers and Forestry (PGRLF if abbreviated from the Czech), and accounted for most credit extended to agriculture (58% on 30 June 1998). Support took the form of either loan guarantees or subsidies on interest, reducing, for instance, the rate from the commercial 15% to 2.5% in 1996 (Agra Europe, 1996, p. 22). Csaki et al. (1999, p. 36) found that PGRLF-supported loans were more often allocated to cooperative farms, and were used to cross-subsidise the activities of parallel limited liability farms.

Credit allocation over farm structures

The survey data offer evidence for the notion, mentioned in the Introduction and in the descriptive literature, that experiences in the credit market of individual farms and corporate farms have differed during transition. Respondents were asked to evaluate the importance of various barriers in operating their businesses, among them `obtaining credit from a bank'. A ranking from 1 (very problematic) to 5 (no problem) produced an average of 1.9 among the 118 responding individual farmers (both professional and `other'). For 38 cooperative farm managers it was 3.1, and for 28 farm company representatives it was 2.5. The differences in means between the individual and corporate farm types are statistically significant ([alpha] = 5%).

Farm managers and operators also reported the amount of credit obtained in 1997 as a percentage of total financial resources in that year. Corporate and cooperate farms reported 20% (n = 26) and 15% (n = 35). Professional and `other' individual farms reported 4% (n = 93) and 8% (n = 40). It should be noted that these averages result from a very different distribution. No credit at all in 1997 (ie 0%) was reported by 23% and 20% of the corporate and cooperative farms, ie by 75% and 73% of professional and `other' individual farms respectively.

Other data suggest that this was true not only for 1997. Respondents were also asked if they took up credit in each of the years 1993-1997. The average number of years in which farmers did this was 0.7 for professional and of 0.3 for `other' individual farms (n = 112 and 46, respectively). The number was significantly ([alpha] = 1%) higher for corporate farms, which scored an average of 3.0 (n = 66), with negligible differences in average score between cooperative and other corporate farms.

This difference may have been because of fewer loan requests or more rejections in the group of individual farmers. Table 1 shows survey data on loan applications and rejections, suggesting that both factors played a role.

The figures show that almost all corporate farm managers applied for credit, and about half of them obtained it. The application rate in professional individual farms is about one-half of that in corporate farms. Over two-thirds of those individual farmers who applied, reported a rejection of their request. The frequency of applications was still lower in `other' individual farms, and here four-fifths had an application rejected. It should be noted that lower application rates may reflect lower demand, but may also indicate farmers' anticipation of a higher rejection probability. The data do not allow us to distinguish between these motives. One conclusion from the table is that individual farmers' lower participation rate in the credit system was at least partly a supply side phenomenon.

Allocation criteria

Respondents who had experienced a rejection of their loan request were asked to cite the reason that the bank gave, from an array of possible grounds suggested by the theoretical and empirical literature, as well as by interview experiences. Several options could be selected. The original question, the answer options, and the findings are reported in Table 2.

Almost all (97 out of 99) who experienced a rejection responded to this question. As to the overall importance of the various reasons, insufficient profitability and overly high risk are mentioned most frequently. This is true for both farm types, but corporate farm managers report this significantly more often than individual farmers ([alpha]=5%). Moreover, these are also practically the only reasons they give.

For individual farmers, three other reasons are mentioned with some frequency: inadequate security of some sort (either collateral or third-party guarantees), the absence of personal relations with the bank management, and the answer `because we work in agriculture'. It appears sensible to perceive a lack of specialist agricultural banks, well-known from other descriptive literature (eg Pederson and Khitarishvili, 1997), behind the reasons `because we work in agriculture' and `because of bank incompetence' (and, to some extent, `small farm size'). Again, this barrier to credit is more relevant for individual than traditional farmers.

Also the role of collateral was investigated in more detail. Collateral has a central function in credit allocation in market economies. It has often been noted (eg Davis et al., 1998, p. 2; OECD, 2001, pp. 125-135) that in credit markets for agriculture in the transition economies, there are typically serious problems connected with the use of collateral. These originate with insufficiently defined or overly complex owner relations and underdeveloped markets in collateral such as farmland and buildings. Respondents reported the type of collateral accepted by the bank in their latest loan transaction. They could check several options. Categories are listed, in order of descending overall frequency, in Table 3.

Overall, there is considerable diversity in the use of collateral, with differences over farm types. Buildings and a private, eg third-party guarantee (eg by a processing firm) rank highest overall; but that ranking mainly results from high frequencies in corporate farms. This is evidence that the traditionally strong links between farms and downstream industries also serve to improve the access to credit of this farm type, as also noted in other studies (eg OECD, 2001, p. 196-206).

The role of the PGRLF appears to have been to give access to credit for individual farmers who would otherwise have little alternative security to offer (buildings apart). Moveable assets and farm land were of comparable importance in this sample. The diversity in collateral used, which is particularly large in the individual farm sample, suggests that no single collateral type is seen as fully adequate. Reasons for this may include structural factors in the agricultural sector (eg the land market) and a learning process in the banks involved.

One oft-mentioned problem with transitional credit markets (from the point of view of the borrower) is the short maturity of loans extended. This is usually attributed to the great uncertainty on firm prospects in the medium and longer term and to the lack of adequate collateral. This problem did not appear to be dramatic in the sample, where the single most frequent maturity in all farm types was the medium term (1-4 years). This term does not allow farms to undertake substantial investments in technology that pay off only after a longer period, but neither does it preclude any investments beyond operational demands. Perhaps also, because of the small number of observations, there are no clear differences between the two alternative farm types, nor within those groups, in the distribution over loan terms.

Finally, an exploration of the importance of profitability as a loan allocation criterion is undertaken. As noted, profitability was the most important self-reported reason for loan rejection. Another reason to pursue this relation is that a basic function ascribed to credit markets is to cause financial resources to flow to the most profitable firms in a given market (eg Rother, 1999, p. 1). Credit allocation in line with some profitability measure would imply that there is a relation, for a given farm, between profitability on the one hand and access to or receipt of credit on the other.

Farm profitability was measured for each farm in the survey as the number of years out of the last five (1993-1997) in which profit was recorded. As to credit, in the survey both measures for access to and receipt of credit (already discussed above) were registered. Respondents reported the perceived difficulty in accessing credit by scoring a scaled variable, from 1 (accessing credit is very problematic) to 5 (accessing credit is no problem). They also reported the number of years in which they took up a loan out of the last 5 years (1993-1997).

Was there a relation between farm profit and access to credit, or between farm profit and receipt of credit, so measured? Computations of the bivariate Pearson coefficient between profit and either credit measures are presented in Table 4. Intuitively, one would expect positive and significant values for both correlation coefficients and for each type of farm. The figures suggest that this is indeed the case for corporate farms, and more strongly so for cooperative than for other corporate farms. For individual farms, however, there is no apparent relation between either pair of variables. Despite their larger number, positive correlation coefficients that are statistically significant are not observable.

Allocation by association?

The exploration of allocation criteria in the preceding section suggests two questions to be further investigated. Why do corporate farms more often obtain credit? And if the credit allocation process within the group of corporate farms reflects the pattern of farm profitability, why is this not the case in allocation to individual farms?

In addressing these questions, it may be suggested that individual farms are less profitable than corporate farms; or that differences in profitability within the group of individual farms are too small to be perceivable, as opposed to differences between corporate farms. Table 5 shows both explanations to be misplaced. This table presents the distribution of all farms over categories of profitability. Differences in profitability over categories, if perceived by lenders, may cause them to limit access to credit for certain farm types. The figures show that these differences could not be observed in the survey. Assuming lenders have the same perception, the ex ante perceived profitability of farm types cannot have been a reason for the more limited receipt of credit by individual farm operators. For no type of farm is the average number of profitable years significantly different, in the conventional statistical sense, from the grand mean. Only the difference in averages of the `professional' and `other' individual farms is statistically significant, but there is no difference between corporate and individual farms. Nor are there clear and explanatory differences between both farm types in the distributions of respondents over the six `number of profitable years' categories. Only the `other individual' farms are clearly more concentrated in the `0' and the `1 profitable year' categories than any of the other farm types. If anything, differences in profitability are easier to perceive there than in the group of corporate farms.

There is another explanation for both the larger allocation to corporate farms and the stronger correlation with profitability in that group. In most thinking on loan allocation decisions, the information available to the bank management on the applicant's characteristics is central. Taking, for instance, the criterion of farm profitability, it is perceived rather than actual profitability that directs the credit allocation decision. Bank management may obtain such information by requiring the loan applicant to prove sufficient (past or expected) farm or project profitability, or by using security instruments (such as collateral or accounting requirement) as tools to facilitate monitoring and ensure against default.

These methods require that accounting practices and standards are sufficiently developed that there exists collateral with sufficient resale value, and that the use of these instruments can be bindingly stipulated in a contract. These conditions obviously are not always fulfilled, particularly so in the transition setting. As is noted in the financial markets literature (eg Barry et al., 1995), relation-driven transacting, facilitating the transmission of information on farm performance, without exclusive reliance on formal contracts, is then an alternative way to obtain the information needed for an informed risk and profitability assessment.

A more general way to put this, which connects to recent theoretical work on economic systems, is to say that local economic networks are important for allocation decisions. Local economic networks consist of relations between economic agents--relations of a business nature, in civic society networks, through kinship, etc. Based on the trust that develops in such networks (Williamson, 1985; Fukuyama, 1995), information is more easily exchanged, because `[t]hrough the economic and social relations in the network, diverse information becomes less expensive to obtain' (Malecki, 2000). Through this information exchange, network relations facilitate coordinated action (Putnam et al., 1993). Relations are partly (never fully) replaceable by explicit contracts, but this is costlier (Arrow, 1972). Also, as was noted, effective replacement is not always possible, particularly in an institutionally deficient environment, such as a post-communist economy.

This theoretical perspective would be appropriate if investments and allocation of resources (credit, subsidies, inputs) occur in a setting where network relations, more than anything else, provide a basis for credible information transmission. This information, in turn, would be the basis for allocation and investment decisions. In this view, those making the investment decision, or those allocating resources (such as bank management), `will have two considerations in making a decision. First, is the applicant able to provide credibly the relevant information on returns to investment and on default risk, that is, is he/she part of the network? If not, a positive allocation decision is probably prohibited by the high uncertainty on key decision parameters such as default risk and return to investment. Second, if the relevant information can be credibly provided, a subsequent consideration is that information itself--does it give the bank management reason to actually allocate credit? In short, those within a network will more often obtain resources than do the outsiders; and if they do, resource allocation will be more in line with some criterion (such as profitability) on which information is now available.

Applying this perspective to this study is warranted if (1) network-based allocation of credit is important in Czech agriculture, and (2) corporate farm managers have better relations with bank management than individual farm operators. As to the first point, Grabher and Stark (1997) have argued the relevance of network-based transacting for post-communism in general, where adequate alternative coordination mechanisms are often absent or underdeveloped. Specifically, Koford and Tschoegl (1999), interviewing Bulgarian bankers in order to identify their loan allocation criteria, specifically report `significant difficulties in accumulating the information to evaluate borrowers'. One of their conclusions is that lending problems could be mitigated by the operation of reputation, which is a substitute for personal relations in information transmission.

In addition, it should be noted that post-communist villages and rural areas are typically local, relatively closed economies with a few large economic agents--the local administration, a bank (branch), a corporate farm--or, occasionally, firm(s), as for instance in mining areas. Both features, relative isolation and a limited number of economic agents, facilitate personalised network-supported allocation and exchange, rather than impersonal, pure-market transactions. This view of the local, rural economy is argued in more theoretical detail by the author in Bezemer (2002b). It appears sensible to apply the above theoretical view to transition economies in general and to credit markets in the rural economy in transition countries in particular.

A second step in the application of this argument is to argue that individual farmers are outside such networks, whereas corporate farm managers are typically part of it. A general observation in support of this is that corporate farms have nearly always been in existence for decades, while individual farms were newly established after the reforms. Typically, the management of the local branch of the agricultural bank was also in office before the reforms started. Indeed, previous qualitative fieldwork found that local 'bank management and farm management in Czech rural areas often have a long-standing relation (Bezemer, 1999).

Findings from the survey data indicate the relevance of this for loan allocation decisions. Those survey respondents who had experienced a loan application rejection were asked about perceived reasons for the rejection. None of the 29 corporate farm managers who responded to the suggestion that a lack of good personal relations was a reason for rejection of their loan application responded in the affirmative. Of the 63 respondents from individual farms, 11 agreed. It is interesting to note that among the answer options to this question were also 'insufficient profitability' and `too high risk'. Corporate farm managers most frequently reported these two factors as reasons for loan application rejection (21 and 12 times respectively, n = 29), while they were relatively less important in the individual-farms group (30 and 15, respectively, n = 69). If individual farms can provide less, or less adequate, information on profit and risk, one would indeed expect that this will less often be identified as a factor in loan decisions.

Additional evidence is contained in information from 132 individual farmers reporting their professional background. Former managers of corporate farms (n = 39) reported a higher average number of years in which credit was taken up in 1993-1997 than did other individual farmers. This was 1.0, which compares to 0.3 for former workers in agriculture (n = 45), 0.5 for former workers in other sectors (n = 35), and 0.7 for former managers in other sectors. In comparison, the average for present managers of corporate farms was 3.0. This supports the idea that credit transactions are partly relation-driven, given the fact that managers of corporate farms who start their own individual farms often do so in their own village or area, and can use the existing local relations in their new business context to credibly signal farm performance.

A qualification of this evidence is that it can also be interpreted as signifying differences in personal capabilities rather than relations. Also, individual farmers who used to be managers of corporate farms are still much less likely to receive credit than those who remained on the corporate farm. This suggests, unsurprisingly, that personal attributes (be they relations or capabilities) were not the only determinants.

It should be noted that the explanation suggested here is based on a view of the local economic system rather than on credit markets or the agricultural sector only. The systemic feature of this account can be exploited to further investigate its applicability using the survey data. In a local economic system, allocation of different resources (not just credit) occurs. Since similar allocation mechanisms can be expected to produce similar distributions, allocation predicated on personal relations as a systemic feature can be assumed to produce similar allocation patterns--even if the resources being allocated are not the same. With such network-based allocation, those within the network have better general access to resources of different types.

In our case, the survey data contain information on allocation of both subsidies and credit. If, as was argued, individual farmers are more often outside the local economic network, the implication would be that individual farmers would more often obtain neither credit nor subsidies, compared to corporate farm managers. Moreover, for the latter group the link between receiving credit and subsidies should be less clean They are assumed to be better able to provide relevant information (in both credit and subsidy allocation processes), and it is the content of that information more than anything else that would control their success in obtaining both credit and subsidies. Since the criteria for allocating credit and subsidies may differ, the allocation pattern may accordingly differ more among corporate farm managers than among individual farmers.

Respondents' evaluation of the extent to which access to credit and subsidies was a barrier to operating their business successfully, ranked between 1 (very problematic) and 5 (no problem), was analysed. For individual farms, the scores on the problematic nature of access to credit and to subsidies are significantly ([alpha] = 1%) correlated; the value of the Pearson correlation coefficient is 0.31. There is no significant correlation in the sample of corporate farms. This is in line with network-based allocation.

Furthermore, based on the information on receipt of credit and subsidies per year in 1993-1997, a variable was constructed that assumes value 1 if credit and subsidy, in a given year and for a given farm, are either both allocated or both not allocated (simultaneous allocation). For observations where only one type of funds is allocated but not the other, its value is zero (non-simultaneous allocation). A missing observation is registered if receipt of either or both was not reported in that year.

In a given subsample, the summation of values of this variable, divided by the total number of observations, would indicate the frequency of simultaneous allocation decisions. Table 6 presents these percentages for the four farm types and over the years 1993-1997.

The figures show that either type of individual farm experienced simultaneous (non-)allocation of credit and subsidies more often than either type of corporate farm, in line with the hypothesis that network participation was more determining in both allocation processes for individual farms. [chi square] tests showed that differences in means of the frequencies of simultaneous allocation of subsidy and credit were statistically significant in the comparisons between both types of individual farms with farm companies ([alpha] = 1%) and, more weakly, cooperative farms ([alpha] = 10%). The two types of corporate farms also differed significantly on average ([alpha] = 1%), but this was not true for both types of individual farms.

Moreover, individual farms more often experience simultaneous exclusion from, rather than simultaneous allocation of, funds. In corporate farms, the frequencies of both coincidences are more similar (Table 7).

For each farm type in Table 7, the sum of top-left and bottom-right figures is the percentage of simultaneous loan/subsidy allocation decisions. The distribution of those decisions over the top-left (yes/yes) and bottom-right (no/no) quarters is an indication of the extent to which such `network-based' decisions favour fund allocation to that farm type. The figures show that simultaneous credit allocation decisions occur in 75% of all observations. In 73% of all observations, these simultaneous decisions are negative. For corporate farms, simultaneous decisions are both less frequent (52% of all observations) and less often negative (30% of all observations). Similarly, non-simultaneous credit decisions (25% of all observations) are rarely negative for individual farms (3% of all observations), while for corporate farms they are often negative (40% out of 48%). In the interest of concise presentation, the figures in Table 7 are averages over 1993-1997; however, separate calculations, not presented here, showed that these conclusions apply for each single year in the series.

Four tentative conclusions from this section would follow. First, network-based allocation of credit is plausibly relevant in transitional agriculture. Second, such allocation, as signified by simultaneous credit and subsidy allocation, appears more prevalent for individual farms than corporate farms, as was hypothesised. Third, such credit decisions are nearly always negative for individual farms. In the present perspective, this would be in line with the assumption that they generally fail to satisfy the first criterion for allocation, which is adequate information provision based on network participation. Fourth, for corporate farms, simultaneous credit allocation decisions are rather evenly distributed over the positive and negative outcomes. This supports the idea that other criteria, such as the outcome of profitability assessments, are more important in credit and subsidy allocation within this group, compared to individual farms.

SUMMARY AND CONCLUSIONS

In this study, the allocation of credit in the Czech Republic to farms of different governance structures in the 1993-1997 years was investigated on the basis of primary survey data. Long-established, corporate farms appeared to have better access to credit than de novo, individual farms. Both the more frequent allocation to corporate farms and the distribution of loans within the individual farms group do not appear to be related to conventional loan allocation criteria or to the distribution of profitability over both farm types.

As an alternative explanation of the observed allocation patterns, it is suggested that in a situation of large uncertainty, and lacking conventional credit rating or security instruments, relation-driven transacting is important in the Czech credit market for agriculture. This helps bank management to reduce their uncertainty over the farm. This practice would indeed produce more credit allocation to the (longer established) corporate farms as well as a credit allocation pattern in line with profitability patterns observed in the survey. Observations of various survey data are in line with this account, although it should be noted that the empirical support for this explanation is indirect. This follows both from the systemic nature of the explanation and from limitations to the available data. Still, this account is offered as a suggestion for understanding the pattern of credit allocation in Czech agriculture.

The possible origins of the observed allocation pattern, which may restrict the scope for the development of farm efficiency, were explored analytically in this paper. It was shown how an institutionally defective environment may provide incentives for individually rational behaviour that produces an overall undesirable outcome. The analysis implies that a remedy to this outcome would primarily lie in the field of institutional development. This would include: better and more general accounting standards as well as training in this area, both for agricultural bank personnel and farm operators; development of more adequate legislation, for example, on residual ownership over collateral and on liquidation procedures; and better functioning courts. None of these points, which have a wider applicability than just Czech agriculture, is entirely original. The contribution of the present study is to once more point out their relevance in a concrete case, on the basis of firm-level data, and using a systemic perspective of the local economy.

Table 1: Individual farmers applied less often for loans and
were more often rejected

Farm type Did you apply for a loan in
 1993-1997?

 No Yes Total

 Counts (percentages in parentheses)

Corporate 1 (1) 27 (17) 28 (11)
Cooperative 3 (3) 37 (24) 40 (15)
Professional
 individual 58 (56) 64 (41) 122 (47)
`Other'
 individual 41 (40) 28 (18) 69 (27)

Total 103 (100) 156 (100) 259 (100)

Farm type Did you have one or more
 loan requests rejected
 during 1993-1997?

 No Yes Total

 Counts (percentages in parentheses)

Corporate 12 (21) 15 (15) 27 (17)
Cooperative 19 (24) 17 (17) 36 (23)
Professional
 individual 20 (36) 44 (44) 64 (41)
Other
 individual 5 (9) 23 (23) 28 (18)

Total 56 (100) 99 (100) 155 (100)

Source: Survey findings

Table 2: ProfitabilLity and risk were the main reasons for loan
rejection

 Farm type (no. of
Did the bank give a reason for the respondents)
rejection?
[] No, they gave no reason. Corporate Cooperative
Please indicate below what you (13) (16)
think the reason was.
[] Yes, they gave the following reason or
reasons (you can check several boxes):

 No. of responses
 (percentages in
 parentheses)

Profitability was not high enough 10 (50) 11 (58)
Risk was too high 4 (20) 8 (42)
There was no adequate collateral 2 (10) 0 (0)
Book-keeping information was not adequate 1 (5) 0 (0)
Nobody guaranteed my loan 1 (5) 0 (0)
I did not know the management personally 0 (0) 0 (0)
Other (not specified in questionnaire)
 My farm is too small 0 (0) 0 (0)
 The bank was incompetent 2 (10) 0 (0)
 We work in agriculture 0 (0) 0 (0)

Total number of responses in all answer 20 (100) 19 (100)
categories

 Farm type (no. of
Did the bank give a reason for the respondents)
rejection?
[] No, they gave no reason. Professional Other
Please indicate below what you individual individual
think the reason was. (46) (23)
[] Yes, they gave the following reason or
reasons (you can check several boxes):

 No. of responses
 (percentages in
 parentheses)

Profitability was not high enough 21 (36) 9 (35)
Risk was too high 12 (21) 3 (12)
There was no adequate collateral 3 (5) 1 (4)
Book-keeping information was not adequate 1 (2) 1 (4)
Nobody guaranteed my loan 5 (9) 3 (12)
I did not know the management personally 9 (16) 2 (8)
Other (not specified in questionnaire)
 My farm is too small 2 (3) 4 (15)
 The bank was incompetent 0 (0) 0 (0)
 We work in agriculture 5 (9) 3 (8)

Total number of responses in all answer 58 (100) 26 (100)
categories

 Farm type (no. of
Did the bank give a reason for the respondents)
rejection?
[] No, they gave no reason. All
Please indicate below what you (97)
think the reason was.
[] Yes, they gave the following reason or
reasons (you can check several boxes):

 No. of responses
 (percentages in
 parentheses)

Profitability was not high enough 51 (41)
Risk was too high 27 (22)
There was no adequate collateral 6 (5)
Book-keeping information was not adequate 3 (2)
Nobody guaranteed my loan 9 (7)
I did not know the management personally 11 (9)
Other (not specified in questionnaire)
 My farm is too small 6 (5)
 The bank was incompetent 2 (2)
 We work in agriculture 8 (7)

Total number of responses in all answer 123 (100)
categories

Note: Since respondents could select several reasons, the total number
of responses in all answer categories is larger than the number of
respondents.

Source: Survey findings

Table 3: Use of different types of collateral differs over farm types

Collateral accepted in the Farm type (no. of respondents)
last loan transaction
 Professional
 Corporate Cooperative individual
 (26) (36) (49)

 No. of responses (percentages in
 parentheses)

Buildings 21 (40) 30 (47) 33 (37)
Private third-party 19 (36) 23 (36) 0 (0)
guarantee
Machinery 8 (15) 6 (9) 15 (17)
PGRLF guarantee 1 (2) 3 (5) 19 (21)
Farm land 1 (2) 0 (0) 11 (12)
Cash 3 (6) 2 (3) 7 (8)
Home 0 (0) 0 (0) 2 (2)
Future revenues 0 (0) 0 (0) 2 (2)

Total number of responses 33 64 89
in all answer categories

Collateral accepted in the Farm type (no. of respondents)
last loan transaction
 Other individual Total
 (18) (129)

 No. of responses (percentages in
 parentheses)

Buildings 11 (44) 95 (41)
Private third-party 3 (12) 45 (19)
guarantee
Machinery 2 (8) 31 (13)
PGRLF guarantee 4 (16) 27 (12)
Farm land 4 (16) 16 (7)
Cash 0 (0) 12 (5)
Home 1 (4) 3 (1)
Future revenues 0 (0) 2 (1)

Total number of responses 25 232
in all answer categories

Note: Since respondents could select several reasons, the total number
of responses in all answer categories is larger than the number of
respondents.

Source: Survey findings

Table 4: Profitability is related to both credit obtained and to
access to credit for corporate, but not for individual farms

 Bivariate Pearson correlation
 coefficients, by farm type
 (no. of respondents)

 Cooperative Other Professional
 (40) corporate individual
 (29) (123)

Access to credit, 1-5 scale 0.344 * 0.259 0.137
(1=problematic, 5=no problem)
Number of years in which loan 0.415 * 0.211 -0.018
taken up in 1993-1997

 Bivariate Pearson correlation
 coefficients, by farm type
 (no. of respondents)

 Other All All
 individual corporate individual
 (70) (69) (193)

Access to credit, 1-5 scale 0.182 0.278 ** 0.155
(1=problematic, 5=no problem)
Number of years in which loan 0.024 0.282 ** 0.043
taken up in 1993-1997

* Correlation coefficient is statistically significant for
P [less than or equal to] 10%.

** Correlation coefficient is statistically significant for
P [less than or equal to] 5%.

Source Survey findings

Table 5: Differences in profitability over farm types are small

No. of profitable Farm type
years
 Other individual Professional individual

 No. of respondents (percentages in parentheses)

0 18 (36) 25 (21)
1 9 (18) 15 (13)
2 5 (10) 30 (26)
3 10 (20) 20 (17)
4 2 (4) 12 (10)
5 6 (12) 15 (13)

All categories 50 (100) 117 (100)

 Average/respondent *

 1.7 2.2

No. of profitable Farm type
years
 Corporate Cooperative All types

 No. of respondents (percentages in parentheses)

0 4 (16) 9 (24) 56 (24)
1 5 (20) 4 (11) 33 (14)
2 7 (28) 8 (22) 50 (22)
3 3 (12) 7 (19) 40 (17)
4 1 (4) 6 (16) 21 (9)
5 5 (20) 3 (8) 29 (13)

All categories 25 (100) 37 (100) 229 (100)

 Average/respondent *

 2.3 2.2 2.1

* Differences between farm types in the average number of profitable
years are not statistically significant at the 5% level, except for
the difference between `professional' and `other' individual farms.

Source: Survey findings

Table 6: Allocation of credit and subsidies to individual
farms (%) is often simultaneous

Year 1993 1994 1995 1996 1997

Farm type (no. of respondents) (a)
Professional individual (97) 73 71 72 71 79
Other individual (39) 79 72 79 84 84
Corporate (22) 64 48 56 24 48
Cooperative (37) 51 54 58 54 67

(a) In each group, the number of cases where there are observations on
both credit and subsidies (n) varies slightly over the years. Here
the lowest number is given.

Source: Survey findings

Table 7: For individual farms, allocation simultaneity
results mainly from exclusion

 % of farms receiving subsidies,
 averaged over 1993-1997

 Type Yes No

 Individual Yes 2 22
% of farms receiving credit, No 3 73
averaged over 1993-1997
 Corporate Yes 22 8
 No 40 30

Acknowledgements

This research was undertaken while the author was affiliated with the Department of General Economics, University of Amsterdam. Financial support from the Dutch Organisation for Scientific Research, the Tinbergen Institute, and the Chair of Transition Economics of the University of Amsterdam is gratefully acknowledged. The Mendel University at Brno and the Czech Research Institute for Agricultural Economics and the Agricultural University, both at Prague, provided various facilities. I thank Eva Kasova, Barbora Kysilkova, Martin Steflicek, Rianne Sybrandi, and Radka Vymlaticova for fieldwork assistance. Michael Ellman, Jeffrey Miller, and two anonymous referees provided helpful comments on earlier drafts. Any errors and all opinions remain mine.

(1) Figures are percentages of total agricultural land. The year of measurement varies between 1996 and 1998. Note that Poland and the former Yugoslavia, while geographically in the Central and Eastern European area, are not relevant here. In both countries (and successor states), family farms were dominant throughout the communist era, and de-collectivisation in the sense defined was not an issue in the transformation.

REFERENCES

Agra Europe. 1996: East Europe agriculture and food monthly, July issue. Agra Europe: London.

Arrow, K. 1972: Political and economic evaluation of social effects and externalities. In: Intriligator, M (ed). Frontiers of Quantitative Economics. North-Holland: Amsterdam.

Barry, P, Ellinger, P, Baker, C, and Hopkin, J. 1995: Financial management in agriculture. Interstate Publishers: Danville, Il.

Bezemer, DJ. 1999: Microeconomic institutions and the transformation of agribusiness: Evidence from the Czech Republic. Eastern European Countryside 5: 85-98.

Bezemer, DJ. 2002a: Credit markets for agriculture in the Czech Republic. Europe-Asia Studies 54(8): 1301-1317.

Bezemer, DJ. 2002b: De-collectivisation in Czech and Slovak agriculture: An institutional explanation. Journal of Economic Issues, 36(3): 1-23.

Csaki, C, Debatisse, M, and Honisch, O. 1999: Food and agriculture in the Czech Republic, from a `velvet' transition to the challenges of EU Accession. World Bank Technical paper no. 43 7. World Bank: Washington.

Csaki, C and Lerman, Z. 1994: Land reform and farm restructuring in the former socialist Countries in Europe. European Review of Agricultural Economics 21: 553-576.

Davis, J and Gaburici, A. 1999: Rural finance and private farming in Romania. Europe-Asia Studies 51: 843-870.

Davis, J, Gaburici, A and Hare, P. 1998: What's wrong with Romanian rural finance? Understanding the determinants of private farmers' access to credit. Working paper, Center for Economic Reform and Transformation. Heriot-Watt University: Edinburgh.

EBRD. 2001: Transition Report. European Bank for Reconstruction and Development: London. Fukuyama, F. 1995: Trust: The social virtues and the creation of prosperity. The Free Press: New York.

Grabher, G and Stark, D. 1997: Restructuring networks in post-socialism. Legacies, linkages and localities. Oxford University Press: Oxford.

Koford, K and Tschoegl, A. 1999: Problems of bank lending in Bulgaria: Information asymmetry and institutional learning. MOCT-MOST: Economic Policy in Transition Economies 9(2): 123-152.

Lerman, Z, Csaki, C and Feder G. 2002: Land policies and evolving farm structures in transition countries. World Bank Working paper 2794.

Malecki, E. 2000: Network models for technology-based growth. In: Acz, Z (ed). Regional Innovation, Knowledge and Global Change. Pinter: London.

Mathijs, E, Blaas, G and Doucha, T. 1999: Organisational form and technical efficiency of Czech and Slovak farms. MOCT-MOST: Economic Policy in Transition Economies 9(3): 331-344.

Mathijs, E and Swinnen, J. 2001: Production efficiency and organisation during transition: An empirical analysis of east German agriculture. The Review of Economics and Statistics 83(1): 100-107.

Ministry. 1999: Zprava O Stavu Zemedelstvz'CR Za Rok 1998. `Zelend Zprava'. Annual Agricultural Report, Ministry of Agriculture: Prague.

OECD. 1995: Review of agricultural policies: Czech Republic. Centre for Cooperation with Economies in Transition of the OECD: Paris.

OECD. 1998: Review of agricultural policies: Russian Federation. Centre for Cooperation with Economies in Transition of the Organisation for Economic Co-operation and Development: Paris.

OECD. 2001: Agricultural Finance and Credit Infrastructure in Transition Economies: Proceedings of OECD Expert Meeting, Portoroz, Slovenia, 2001. Centre for Co-operation with Economies in Transition of the Organisation for Economic Co-operation and Development: Paris.

Pederson, G and Khitarishvili, T. 1997: Challenges of agricultural and rural finance in CEE, NIS and Baltic Countries. Center for International Food and Agricultural Policy Working Paper WP97-5.

Prosterman, RL, Mitchell, RG and Rorem, BJ. 1998: Prospects for family farming in Russia. Europe-Asia Studies 50: 1383-1407.

Putnam, R, Leonardi, R and Nanetti, R. 1993: Making democracy work. Princeton University Press: Princeton, NJ.

Ratinger, T and Rabinowicz, E. 1997: Changes in the farming structures in the Czech Republic as a result of land reform and privatization. In: Swinnen, J, Buckwell, A and Mathijs E. (eds). Agricultural Privatization, Land Reform and Farm Restructuring in Central and Eastern Europe. Ashgate: Aldershot, UK.

Rother, P. 1999: Explaining the behavior of financial intermediation: evidence from transition countries. IMF Working Paper 99/36. IMF: Washington, DC.

Sarris, A, Doucha, T and Mathijs, E. 1999: Agricultural restructuring in central and eastern Europe: implications for competitiveness and rural development. European Review of Agricultural Economics 26: 305-329.

Shrieder, G and Heidhues, E 1998: Access constraints of Romanian peasants in relation to the formal financial sector. Discussion paper 97/12, Center for Economic Reform and Transformation of Heriot-Watt University: Edinburgh.

Swinnen, J and Gow, H. 1999: Agricultural credit problems and policies during the transition to a market economy in central and eastern Europe. Food Policy 24(I): 21-47.

Williamson, O. 1985: The economic institutions of capitalism. The Free Press: London and New York.

Wolz, A. 1996: State intervention in Slovak agriculture. Discussion paper no. 52, Institute for Research on International Agricultural Development of the University of Heidelberg, Germany, University of Heidelberg Press: Heidelberg.

World Bank. 1995: Farm restructuring and land tenure in reforming socialist economies. A comparative analysis of eastern and central Europe. World Bank Discussion papers no. 268, World Bank: Washington, DC.

World Bank. 2000: Czech Republic: Completing the transformation of banks and enterprises. World Bank Country Study 21440, World Bank: Washington, DC.

DIRK J. BEZEMER *

Received December 2001; revised July 2002; accepted September 2002

Imperial College at Wye, University of London, Wye, Kent TN25 5AH, UK. E-mail: d.bezemer@ic.ac.uk