Annals of Public and Cooperative Economics 81:1 2010
pp. 21–38
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY: SOME EMPIRICAL EVIDENCE FOR ITALIAN PRODUCERS’ COOPERATIVES by Ornella Wanda MAIETTA Universita’ degli studi di Napoli, Italy
and Vania SENA∗ Aston Business School, UK
ABSTRACT∗∗ :
In this paper, we test the extent to which producers’ cooperatives can experience an increase in technical efficiency following a tightening of financial constraints. This hypothesis is tested on a sample of Italian conventional and cooperative firms for the wine production and processing sector, using frontier analysis. The results support the hypothesis that increasing financial pressure can affect positively the cooperatives efficiency.
1
Introduction
How do increasing financial constraints affect the technical efficiency of producers’ cooperatives?1 Can they provide them with the incentives to improve their own performance as it happens with ∗
We want to thank Toke Aidt, Sergio Destefanis, Pasquale Lombardi, Mario Padula and Virginie Perotin for very valuable comments on previous versions of the work. We want also to thank the CSEF, University of Salerno, Italy for kindly providing the data. The usual disclaimer applies. E-mail:
[email protected]. ∗∗ R´esum´e en fin d’article; Zusammenfassung am Ende des Artikels; resumen al final del art´ıculo. 1 For the remainder of the paper, we will use the term cooperatives for producers’ cooperatives. C 2010 The Authors C CIRIEC 2010. Published by Blackwell Publishing Ltd. 9600 Garsington Road, Journal compilation Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
22
ORNELLA WANDA MAIETTA AND VANIA SENA
conventional firms or, on the contrary, push them out of the market so affecting adversely their chance of survival? For conventional firms, there exists substantial literature proving the role that financial constraints may play to improve efficiency. Nickell and Nicolitsas (1999) suggest that tightening financial constraints (due for instance to increases in the general level of interest rates) can sharpen incentives for managers of credit-constrained firms to increase productivity so to offset the fall in profits. In this paper we argue that cooperatives do not behave differently when facing the same type of constraints. More specifically we suggest that when the external availability of external finance reduces, members will try to cut back on investment and inefficiency so to counterbalance the potential adverse impact of financial constraints on the profitability of the cooperative. This may seem counterintuitive as it is usually believed that members of a cooperative do not have the incentive to cut inefficiencies as they will be able to appropriate only a fraction of the stream of added income generated by the extra effort and therefore may prefer to accept some inefficiency in the production process (Porter and Scully 1987). Our claim, though, is that increasing financial constraints may have the power to alter the incentives members have to act upon inefficiency. We test empirically the extent to which there is a positive relationship between increasing financial constraints and technical efficiency for a panel of conventional and cooperative Italian firms, specialized in the production of wine, over the time 1996–2001. Italy is the country with one of largest cooperative sectors among Western economies; therefore several studies on Italian cooperatives have been conducted, so providing a useful benchmark against which our results can be compared (Pencavel et al. 2005, Bartlett et al. 1992, Jones and Svejnar 1985). To estimate technical efficiency in our panel, we adopt the so-called ‘frontier’ approach to the measurement of technical efficiency where (in)efficiency is computed as the distance from an estimated optimal benchmark, that defines the optimal amount of output that can be produced in a sector given the available technology (the so-called ‘production frontier’). The closer a firm is to the frontier, the more efficient it is. More specifically we use the one-stage approach as in Battese and Coelli (1995) that estimates simultaneously the parameters of the production frontier and the efficiency scores, while simultaneously controlling for the factors that affect the distribution of the scores across the observations. The structure of the paper is the following. Section 2 discusses the relationship between financial constraints and technical efficiency in a cooperative. The empirical model and the results are presented C 2010 The Authors C CIRIEC 2010 Journal compilation
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY
23
in Sections 3 and 4, respectively. Finally some concluding remarks are offered in Section 5.
2
The theoretical framework
The central hypothesis of this paper is that the reduced availability of external finance (following a general increase in interest rate, for instance) should give the members of a cooperative the incentives to improve the efficiency of operations so to guarantee its profitability and therefore its survival. Obviously, this is possible as long as there is some initial inefficiency in the production process that can be acted upon as the availability of external finance decreases. This in turn requires a better understanding of the sources of inefficiencies in a cooperative. The theoretical literature has identified many of their institutional characteristics that can contribute to their inefficiency (Bonin et al. 1993, Porter and Scully 1987). In economic terms, a producers’ cooperative is a special type of cooperative where members supply inputs (other than labour) to the cooperative’s common productive process and this makes them different from the workers’ cooperatives (where members supply their work). The governance of a cooperative is quite simple: members (meeting in a General Assembly) form the so-called control group that coordinates the cooperative’s productive activities; they also are the residual claimants of the cooperatives’ surplus and therefore the bearers of the risk associated to its fluctuations due to the business cycle (Dow 2003). The cooperative is governed by the board of Directors and managed by the Chief Executive Manager. The board proposes to the annual meeting the amount of patronage dividends members are entitled to; these are usually calculated on the basis of the cooperative’s profits at the end of the financial year. Finally, in a producers’ cooperative, labour input is supplied by hired workers who, in terms of legal status, are not very different from the workers in a conventional firm. According to Porter and Scully (1987), technical inefficiency in a cooperative arises from the split between control and ownership in the same way as in a conventional firm. Indeed, managers of a cooperative can control the cash-flow generated by the cooperative process and therefore they have the incentive to use a part of the cash-flow for activities that do not necessarily create value to the members. At the same time though, unlike conventional firms, these can be reluctant to start any (costly) action that may curb C 2010 The Authors C CIRIEC 2010 Journal compilation
24
ORNELLA WANDA MAIETTA AND VANIA SENA
technical inefficiency as they are aware that they will be able to appropriate only a fraction of the stream of added income generated by the extra effort and therefore may prefer to accept some inefficiency. In this context, how can increasing financial constraints reduce technical inefficiency? To be able to answer this question we have first to clarify what type of financial constraint we are considering. A cooperative may experience financial constraints for several reasons. Traditionally, it is believed that cooperatives have limited access to risk capital (and so are financially constrained) as members are not willing to invest because of the way property rights are defined in a cooperative (Cook 1995, Hansmann 1996). However, this is not the only type of financial constraint a cooperative is exposed to. These may also arise because of the credit rationing cooperatives may experience due to the existence of asymmetric information in the credit market. In a very influential paper, Stiglitz and Weiss (1981) suggest that informational asymmetries between firms and banks in the credit market may generate credit rationing in equilibrium as there will be some borrowers who will have access to funds as long as their loans are fully collateralized (usually by the firm’s net worth).2 Asymmetric information in the credit market makes borrowing external funds more expensive than using internal funds for two reasons: Nickell and Nicolitsas 1999, first, the cost of borrowing external funds now includes the costs of the collateral evaluation and of monitoring. Second, the cost of external funds is now also increasing in the ratio of the size of the loan to collateralizeable net worth as the probability of bankruptcy increases when the debt rises relative to net worth (Nickell and Nicolitsas 1999). Now consider what happens if there is an increase of interest rates. This has two effects: first, it increases the cost of borrowing and therefore the costs of production go up while the overall profits of the cooperative will go down along with the share of the surplus each member receives at the end of the financial year. Second, the value of the cooperative’s net worth decreases (as large part of the net worth of a company is the present value of future profits) and so the ratio of debt to net worth rises implying that the overall availability of credit reduces further (Nickell and Nicolitsas 1999). 2 The argument is well-known. Banks trying to sort safe borrowers from risky ones cannot make loans more expensive by increasing the interest rate beyond a certain level as it will push borrowers with safer projects out of the market. Therefore banks may wish to hold the interest rate below the market clearing level to avoid adverse effects on their returns and rather prefer to screen potential borrowers. C 2010 The Authors C CIRIEC 2010 Journal compilation
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY
25
How will this impact a cooperative? There are two types of impact to consider. First, there is the direct consequence of the increase in the cost of borrowing. Indeed, any investment is adversely affected by a rise in cost of borrowing.3 Second, the fall of the share of surplus members receive implies that they have now the incentive to cut internal inefficiencies; in other words, as the debt position of the cooperative gets worse and the future of the cooperative seems uncertain,4 members may not only cut back on investment of various kinds but may also increase their efforts to cut costs and raise efficiency with the result that the cooperative’s actual output gets closer to the potential output. The aim of the empirical analysis is to provide new robust evidence in this regard by measuring the impact that increasing financial constraints have on the technical efficiency of producers’ cooperatives.
3
The empirical analysis
The format of our empirical analysis is straightforward enough. We apply parametric frontier techniques to compute the technical efficiency indexes for both cooperatives and conventional firms. In specifying the production frontier for our data-set, however, we control for the fact that we have two different types of firms (conventional and cooperative firms) that may not share the same technology by using insights from both the theoretical and empirical literature on cooperatives. Also, to measure the impact of increases in a firm’s financial constraints on its technical efficiency, we condition the mean of the technical efficiency scores’ distribution on a measure of the 3 It is important to recall that cooperatives may be reluctant to cut employment and may prefer to cut investment. A cooperative may decide not to make workers redundant as it may have the social objective of creating jobs in less developed areas (or for specific types of workers) and therefore during a recession may prefer to preserve employment so saving on labour turnover costs as well (Estrin and Jones 1995, Doucouliagos 1995). 4 Dissolving the cooperative is usually not an option as the opportunity cost of keeping the cooperative operating is only the unemployment allowance (in countries where the system exists, see Perotin 2006). Indeed there exists empirical evidence suggesting that cooperatives’ mortality is independent of economic downturns as there is some evidence of countercyclicality in the dissolution of the cooperatives (Russell and Hanneman, 1992). C 2010 The Authors C CIRIEC 2010 Journal compilation
26
ORNELLA WANDA MAIETTA AND VANIA SENA
firms’ financial constraints, in addition to a set of variables that control for additional sources of heterogeneity.
3.1
The production frontier specification
A central tenet of the frontier approach is that the units under analysis should be homogeneous in terms of the technology they use so to get meaningful frontier estimates. Indeed, theoretical work suggests that cooperatives tend to be under-capitalized due to the pervasive existence of financial constraints as well as a dissimilar structure of property rights (Mosheim 2002); this would imply that cooperatives may use a less capital-intensive technology than conventional firms. Not surprisingly, previous empirical research conducted on data-sets containing both cooperatives and conventional firms has tried to control for this heterogeneity in the technology by augmenting the production function specification for conventional firms with a set of interactions between the input variables and the dummy variable for cooperatives so that changes in input levels in cooperatives can affect output differently from conventional firms (see Jones 2007, for an example). So our frontier specification (assuming a translog functional form) is the following: ykt = β0 +
3
βi xkti + 0.5
i=1
+ COOP∗ 0.5
3 3 i=1 j=1
3 3
βi j xkti xktj + COOP∗
3
ρi xkti
i=1
ρi j xkti xktj + β10 YEAR + β11 SOUTH
i=1 j=1
+ β12 COOP + (vkt − ukt )k = 1, . . . , K t = 1, . . . , T
(1)
where y kt is the log of production of the kth firm at time t, x kt is the vector of the log of inputs (capital, labour and materials, respectively) of the kth firm at time t and β is a vector of unknown parameters. Notice that both the terms in level and the interactions have been multiplied by a dummy COOP, taking the value of 1 if the firm is a cooperatives and 0 otherwise. Also we introduce among the regressors the dummy COOP (on its own, not interacted with other variables) in case there are features of the cooperative status that can affect directly the output level. By testing whether the coefficients ρ i are significant, we can test whether this specification fits better our data. We allow for the possibility of disembodied technical progress by introducing a continuous time trend (YEAR) in the model. Finally we control for the firm’s location by using a C 2010 The Authors C CIRIEC 2010 Journal compilation
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY
27
dummy variable related to its geographic location (South); this takes the value of 1 if the firm is located in the South of Italy and 0 otherwise. This variable is traditionally introduced to capture the impact that different environmental conditions have on the firms’ level of production. This is to be probably ascribed to the operation of local factors such as infrastructure endowment, external economies linked to the local technological potential or level of industrialization, the presence of organized crime, and so on. The v kt are independent and identically distributed random errors, assumed to be distributed as a N(0,σ 2v ), and independent of the u kt ; in turn, these are non-negative random variables assumed to account for technical inefficiency in the production and to be independently distributed as truncations at zero of the N(μ kt , σ 2u ) distribution with μ kt = z kt δ, where z kt (or technical efficiency effects) is a p × 1 vector of variables which may influence the efficiency of a firm and δ is a p × 1 vector of parameters to be estimated together with σ 2 = σ 2v + σ 2u and γ = σ 2u /(σ 2v + σ 2u ). The model for the technical efficiency effects in the stochastic frontier model is: ukt = δ0 + δ1 SOUTH + δ2 COOP + δ3 YEAR + δ4 FINANCE CONSTRk,t−1 + δ5
k,t−1
FINANCE CONSTR∗ COOP
(2)
where δ 0 is a constant term. In the specification of the inefficiency model, we control again for the firm’s location with the dummy variable SOUTH. It is a well established piece of evidence in the Italian literature that location matters for productive efficiency. We also assume that the cooperative status has an impact on the firms’ efficiency, as suggested by the literature started by Porter and Scully (1987). We introduce the time trend variable (YEAR) in the specification of the inefficiency model so to control for the impact on technical efficiency of factors that affect both cooperatives and conventional firms equally every year (INEA 2001). Finally we introduce our variable of interest: FINANCE CONSTR. This variable is introduced on its own first and interacted with the dummy variable COOP. Our expectation is that the impact of increasing financial constraints on technical efficiency is positive in the case of cooperatives. Also, this variable is lagged of one period: this has been done to avoid potential endogeneity problems in the regression model. (1) and (2) can then be estimated simultaneously by using the procedure suggested by Battese and Coelli (1995), based on Maximum Likelihood estimation. C 2010 The Authors C CIRIEC 2010 Journal compilation
28
3.3
ORNELLA WANDA MAIETTA AND VANIA SENA
The institutional framework, the data-set and the variables
The Italian cooperative system is one of the largest in the Western economies and not surprisingly it has been the object of several empirical studies in the field (Pencavel et al. 2005, Bartlett et al. 1992, Jones and Svejnar 1985 among the others). The cooperative sector contributes 7–8 per cent of GDP (Lega delle Cooperative 2006). Our data set is an unbalanced panel of Italian conventional and cooperative firms from 1996 to 2001, belonging to the sector of Wine Production and Processing5 (corresponding to the 4-digit sector A01.13/1 of the SIC92 classification). The data-set has been extracted from AIDA,6 a database collecting all the annual balance sheets of those Italian companies whose operating revenues are larger than 1 million euros. So the database represents the universe of firms above this threshold. In addition to the information contained in the annual reports, the database reports information on companies’ location, their legal status and additional financial data. The initial data-set has been cleaned so to eliminate all the observations containing both missing and negative values; also, we have cut the 1 per cent extreme observations. So, the final number of observations over the five years is now 413. According to their legal status, 63 firms (corresponding to 250 observations over the whole time period) are cooperatives, while 40 firms (corresponding to 163 observations) are conventional firms. The wine industry has been selected for a number of reasons: first, the firms’ output mix is limited compared to that of firms belonging to other sectors as they produce only wine.7 In addition, the number of cooperatives in the Italian wine industry has always been rather substantial and this implies that their size has always been slightly larger than that of the conventional firms (van Bekkum and van Dijk 1998, Huffman 2001). Output is measured by the company’s sales plus the change in inventories deflated by the appropriate production index (ISTAT 2002). Among the inputs, we include intermediate consumption, capital and labour. Intermediate consumption is defined as the sum 5 The firms classified in this sector include firms that both grow and process grapes to produce wine. For the remainder of the paper, we will refer to this sector interchangeably as the wine sector or wine industry. 6 More information on this database can be found at http://www. bvdep.com/browse5.asp. 7 In particular, the firms (both cooperatives and conventional) included in our sample are specialized in the production of medium quality wine. C 2010 The Authors C CIRIEC 2010 Journal compilation
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY
29
Table 1 – Sample descriptive statistic: median Intermediate consumption∗
Capital∗
Labour∗∗
ICR∗
Conventional firms 1997 3729 1998 3592 1999 3068 2000 2723 2001 2681
3115 2401 2419 1843 1855
1915 1749 1923 2450 2601
14 14 14 12 13
0.31 0.30 0.27 0.25 0.26
Cooperatives 1997 1998 1999 2000 2001
4005 3254 3369 3076 3125
1412 1336 1310 1543 1732
12 10 9 11 10
0.48 0.39 0.32 0.34 0.33
Statistics
Output∗
4148 3827 3684 3564 3452
∗ 1995 ml Italian liras. The Table reports the yearly median value of each variable by type of firm. ∗∗ Labour is the median number of employees per firm and per year.
of materials and services while capital is the sum (at book value) of land, buildings, machinery and other fixed assets. Both variables have been deflated by the price index of materials and of investment goods for the beverage industry, respectively (ISTAT 2002). All these variables are expressed in 1995 million Italian liras. Labour is the number of employees in each firm at the end of the fiscal year and includes both full-time and seasonal workers. Our proxy of financial constraints is the interest coverage ratio (ICR) (Whited 1992, Ng and Schaller 1996). This is the ratio between the interest expenses and the sum of the interest expenses and the firm’s cash-flow. It is a commonly used measure of a firm’s likelihood of ending up in financial distress. So it measures how well the firm’s earnings can cover the interest payments on its debt. Indeed the higher the ratio, the larger the proportion of cash-flow that is used to service the debt, giving this way an indication of the long-term solvency of the firm. ICR will be higher for a firm that has experienced negative shocks that reduce cash-flows and/or increase interest payments. Table 1 reports the sample statistics of our variables for both conventional and cooperative firms, respectively. On average, cooperatives produce more than conventional firms, use less capital and labour but have more intermediate consumption than conventional firms for each year under consideration. This relative C 2010 The Authors C CIRIEC 2010 Journal compilation
30
ORNELLA WANDA MAIETTA AND VANIA SENA
undercapitalization of cooperatives is quite common8 and not only limited to the Italian cooperatives (see also Mosheim, 2002). This is usually explained by the fact that members of cooperatives do not have an incentive to invest in capital equipment as they may not appropriate the increase in value following the investment, in case they decide to leave the cooperative (Mosheim, 2002). ICR increases steadily for conventional firms, while it does not show any trend for cooperative firms.
4
The results
4.1
The production frontier estimates
The Maximum Likelihood estimates of (1) are reported in Table 2. The estimates are generally statistically different from zero at the 5 per cent level. The regularity condition (monotonicity and quasi-concavity)9 of the estimated translog production function have been checked in each sample point: the fitted production elasticities are positive for almost all observations and the bordered Hessian matrix is negative semi-definite at almost all data points. We thus conclude that our estimated translog production function is appropriate to represent the state of technology of Italian wine industry. We have mean-corrected our sample data before the estimation so that we can interpret the 1st order coefficients listed in Table 2 as the partial output elasticities evaluated at the sample mean. Table 3 summarizes these partial output elasticities for both cooperatives10 8 Tortia (2007) for instance finds evidence of a lower ratio of fixed assets per head in cooperatives. 9 Monotonicity of the translog requires the logarithmic marginal products to be positive for all inputs; in turn this requires the logarithmic partial elasticity for each input to be positive. The isoquants of the translog function are strictly convex if the corresponding bordered Hessian is negative definite. The above mentioned conditions of positive monotonicity and quasi-concavity depend on the values of the inputs, the output and the individual coefficients of the estimated translog function. 10 However, it is important to recall that for the cooperatives the partial output elasticity formulas are different. For instance, the output elasticity to labour is computed as: β3 + β6 ln L + β8 ln K + β9 ln M + ρ3 COOP + ρ6 ln L ∗ COOP + ρ7 ln K ∗ COOP + ρ9 ln M ∗ COOP where β3 + ρ3 can be interpreted as the output elasticity of labour for cooperatives evaluated at the geometric mean of the sample. C 2010 The Authors C CIRIEC 2010 Journal compilation
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY
31
Table 2 – MLE estimates Variable
Coefficient
t-ratio
Constant β1 β2 β3 β 11 β 22 β 33 β 12 β 13 β 23 ρ1 ρ2 ρ3 ρ 11 ρ 22 ρ 33 ρ 12 ρ 13 ρ 23 β 10 β 11 β 12
8.310 0.775 0.058 0.148 0.080 0.011 0.066 −0.019 −0.080 0.007 0.073 −0.035 −0.053 0.047 −0.010 −0.007 0.021 −0.028 0.019 0.010 −0.039 −0.117
372.507 61.642 6.991 11.309 4.190 1.692 3.244 −2.237 −5.277 0.623 3.890 −2.261 −2.739 1.672 −0.790 −0.237 1.198 −1.343 1.015 1.572 −2.490 −5.785
Inefficiency model parameters δ0 δ1 δ2 δ3 δ4 δ5 σ2 γ
−1.338 0.261 −0.273 −0.243 −0.065 −1.347 0.118 0.941
−5.942 8.956 −3.481 −4.981 −6.133 −14.688 6.204 63.874
Table 3 – Output partial elasticities
Cooperatives Conventional Firms
Capital
Materials
Labour
Scale returns
0.023 0.058
0.848 0.775
0.095 0.148
0.966 0.981
Note: The elasticities are computed at the mean value of the respective variable.
and conventional firms. On average, the cooperatives’ elasticity of output to intermediate consumption is higher than for the conventional firms, while conversely, conventional firms’ elasticity of output to capital and labour is higher than the cooperatives’. In general a C 2010 The Authors C CIRIEC 2010 Journal compilation
32
ORNELLA WANDA MAIETTA AND VANIA SENA
Table 4 – Hypothesis tests on the parameters of the stochastic production frontier Null Hypothesis
Ho Ho Ho Ho Ho
: ρo = 0 : ρ i = ρ ij = 0 i = 1 , . . . , 3; j = 1 , . . . , 3 : ρ ij = β ij = 0 i = 1 , . . . , 3; j = 1 , . . . , 3 : β 10 = 0 : γ = 0, δ i = 0 i = 0 , . . . , 5
Critical value (5% level)
Decision
59.34 55.26 225.72 1.86 119.31
3.84 16.92 21.03 3.84 13.40
Reject Null Reject Null Reject Null Do not Reject Null Reject Null
Note: The critical values for the test that technical efficiency effects are not random are obtained from Table 1 of Kodde and Palm (1986) where the degrees of freedom are q + 1 where q is the number of parameters assumed equal to zero, but are not boundary values.
10 per cent increase of materials implies an 8.5 per cent increase in output for cooperatives while for conventional firms this is only equal to a 7.7 per cent increase in output. This is consistent with the fact that cooperatives tend to use more materials than conventional firms. Returns to scale tend to be constant for both types of firms. These findings are also in accordance with both theory and previous empirical analysis: Coppola et al. (2008), by estimating a non parametric cost function, find that the shadow prices of capital and labour for cooperatives are significantly lower than those of conventional firms in the Italian wine industry. Finally, scale elasticity is given by the sum of the three first-order output elasticities and it is measured at the mean values of the input set. Scale elasticity for cooperatives is equal to 0.966 while for cooperatives it is equal to 0.981. This is consistent with the traditional result suggesting that cooperatives are less capable of exploiting economies of scale as they have access to worse technology than conventional firms. As for the estimation’s results, cooperatives tend to produce less on average; also firms (whether cooperatives or not) located in the South tend to produce less output. Several hypothesis tests have been carried out on the estimated production frontier. Table 4 summarizes the main results. We have tested whether cooperatives and conventional firms share the same technology with three tests. First of all, we have tested whether the dummy COOP is significant or not. The test rejects the null hypothesis (COOP = 0) and therefore we can conclude that the cooperative status can affect the level of output directly. We have also tested specification (1) versus a specification where the terms interacted with COOP appear. The test rejects the null hypothesis (namely that the interaction terms are equal to zero). We also tested the translog functional form against the C 2010 The Authors C CIRIEC 2010 Journal compilation
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY
33
alternative (common) Cobb-Douglas. The null hypothesis is rejected: indeed the LR test is equal to 225.72 with a number of restrictions equal to 6 (against a critical value of 12.59 at the 5 per cent significance level). The hypothesis of no neutral technical change cannot be rejected. In addition we have tested whether the technical inefficiency effects in the model are not random and whether the explanatory variables in our technical efficiency effects model are significant. For this test, the null hypothesis is formulated as H o : γ = 2 0, where γ = σ 2σ+σ 2 , and δ z = 0, where z = 0, . . . 5; also in this case v u the null hypothesis is strongly rejected by the data. The result of the test indicates that the frontier model is a significant improvement over the standard average production function. The estimated coefficients of the z’s variables are listed at the end of Table 2. Recall that a positive coefficient associated to one of these variables implies that the variable has a negative impact on the firms’ technical efficiency. From these results, we can see that the dummy for the South of Italy is significant: therefore firms located there tend to be less inefficient than the ones located in the North of Italy. The dummy COOP is also significant implying that the cooperative status decreases the inefficiency of the firm. Also, technical inefficiency grows over time as the variable YEAR is significant and positive. The variable relative to the firm’s ICR is statistically significant and negative. Its interaction with the dummy for cooperatives is significant and negative; also the parameter of the interaction term is larger than the parameter of the variable ICR: this implies that inefficiency generally decreases as firms face increasing financial constraints but the effect is larger for cooperatives. Table 5 shows the main descriptive statistics of the technical efficiency scores computed for both cooperatives and conventional firms. First, it is clear that on average cooperatives tend to be more efficient than conventional firms and that the dispersion of the efficiency scores around the mean is not very large. This is in contrast with what happens with conventional firms where, on the contrary, the efficiency dispersion around the mean is rather large. It seems clear that cooperatives appear to cope better with higher financial pressure. These results are not in contrast with the findings from previous studies on Italian cooperatives. Jones and Svejnar (1985) find that the superior performance of Italian producers’ cooperatives could be ascribed to structural characteristics of cooperatives like profit-sharing and participation. Also, Bartlett et al. (1992) find that Italian cooperatives achieve higher levels of both labour and capital productivity than comparable private firms. C 2010 The Authors C CIRIEC 2010 Journal compilation
34
ORNELLA WANDA MAIETTA AND VANIA SENA
Table 5 – Technical efficiency scores Years
Cooperatives
Conventional firms Average
1998 1999 2000 2001
0.953 0.949 0.937 0.920
0.922 0.903 0.876 0.863 standard deviation
1998 1999 2000 2001
5
0.015 0.013 0.077 0.013
0.069 0.075 0.080 0.078
Concluding remarks
In this paper, we have analyzed the nature of the relationship between financial constraints and technical efficiency in a panel of Italian conventional and producers’ cooperatives specialized in the production of wine over the period 1996–2001. This is a well established research topic for conventional firms and indeed several studies (both theoretical and empirical) have defined the channels through which tighter financial constraints can improve technical efficiency. On the contrary, very little research on this topic has been done for cooperatives. Still, from a theoretical standpoint, it is possible to argue that increasing financial constraints may have a beneficial impact on the technical inefficiency of producers’ cooperatives: as these are appear to be inefficient, a reduction in the availability of external financial resources has the effect of inducing members to cut inefficiencies in their productive process so to minimize losses. Our findings are quite interesting: a) cooperatives appear to be systematically more efficient than conventional firms; also the dispersion of the efficiency indicator is quite small, unlike conventional firms; b) cooperatives and conventional firms use technologies with different capital-labour ratios; c) the output elasticity of materials is relatively higher for cooperatives than for conventional firms; d) cooperatives (along with conventional firms) experience an improvement in technical efficiency following an increase in the financial constraints. What are the policy implications of our study? It seems that producers’ cooperatives are well-equipped to cope with increasing C 2010 The Authors C CIRIEC 2010 Journal compilation
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY
35
financial constraints or, more correctly, their institutional structure allows them to have some ‘in-built’ mechanisms that give them some slack that can be used to deal successfully with financial pressure. Like any other type of firm, cooperatives can devise strategies that can help them to survive when there are increasing credit constraints and one of these is to reduce the internal inefficiency where it is possible. One limitation of this study is that we only consider a specific type of cooperative firm (producers’ cooperatives) in one sector (the wine sector). Therefore further research is needed to test whether this positive relationship between financial pressure and cooperatives’ technical efficiency still holds in other sectors and for other types of cooperatives. Also additional work is required to identify exactly the channels through which increasing financial constraints may exert such a positive impact on cooperatives’ efficiency in different types of cooperatives as it is plausible to assume that these may be quite different for each type of cooperative. REFERENCES
BARTLETT W., CABLE J., ESTRIN S., JONES D. and SMITH S., 1992, ‘Labor-managed cooperatives and private firms in North Central Italy: an empirical comparison’, Industrial and Labor Relations Review, 46(1), 103–118. BATTESE G.E. and COELLIT J., 1995, ‘A model for technical inefficiency effects in a stochastic frontier production for panel data’, Empirical Economics, 20(2), 325–332. BONIN J., JONES D. C., and PUTTERMAN L., 1993, ‘Theoretical and empirical studies of producer cooperatives: will ever the twain meet?’ Journal of Economic Literature, 31(3), 1290–1320. COOK M. L., 1995, ‘The future of U.S. agricultural cooperatives: a neo-institutional approach’, American Journal of Agricultural Economics, 77(5), 1153–1159. COPPOLA A., MAIETTA O.W. and PASCUCCI S., 2008, ‘Corporate governance e performance: un’analisi comparata delle imprese del sistema agro-alimentare italiano’, in Cella, G. and Zago, A., eds, Spillover sistemici, efficienza e competitivita` dell’economia italiana: una valutazione quantitativa per le politiche di settore, Bologna, Il Mulino. DOUCOULIAGOS C., 1995, ‘Worker participation and productivity in labor-managed and participatory capitalist firms: a metaanalysis’, Industrial and Labor Relations Review, 49(1), p 58–77. C 2010 The Authors C CIRIEC 2010 Journal compilation
36
ORNELLA WANDA MAIETTA AND VANIA SENA
DOW G. K., 2003, Governing the Firm. Workers’ Control in Theory and Practice, CUP, Cambridge. ESTRIN S. and JONES D.C., 1995, ‘Worker participation, employee ownership and productivity’, Advances in the Economic Analysis of Participation and Labor-managed Firms, 5, 3–24. HANSMANN H., 1996, The Ownership of Enterprise, HUP, Cambridge, MA. HUFFMAN W. E., 2001, ‘Human capital: education and agriculture’, in Gardner B. L. and Rausser G. C., eds, Handbook of Agricultural Economics, Amsterdam, Elsevier Science. INEA 2001, Annuario dell’agricoltura italiana, Napoli, Edizioni Scientifiche Italiane. ISTAT 2002, ‘Numeri indici dei prezzi alla produzione e al consumo’, Collana Informazioni, N◦ 54, Roma, Italy. JONES D., 2007, ‘The productive efficiency of Italian producer cooperatives: evidence from conventional and cooperative firms’, Advances in the Economic Analysis of Participatory and Labour Managed Firms, 10, 3–28. JONES D. and SVEJNAR J., 1985, ‘Participation, profit sharing, worker ownership and efficiency in Italian producer cooperatives, Economica, 52 (208), 449–465. LEGA DELLE COOPERATIVE, 2006, ‘Internationalisation of cooperatives: some experiences and reflections by Legacoop’, mimeo, Rome, Italy. MOSHEIM R., 2002, ‘Organizational type and efficiency in the Costa Rican coffee processing sector’, Journal of Comparative Economics, 30, 296–316. NG, S. and SCHALLER, H., 1996, ‘The risky spread, investment and monetary policy transmission: evidence on the role of asymmetric information’, Review of Economics and Statistics, 375– 383. NICKELL S. and NICOLITSAS D., 1999, ‘How does financial pressure affect firms’, European Economic Review, 43(8), 1435–1457. PENCAVEL J., PISTAFERRI L. and SCHIVARDI F., 2005, ‘Wages, employment and capital in capitalist and worker-owned firms’, mimeo, University of Stanford. PEROTIN V., 2006, ‘Entry, exit and the business cycle: are cooperatives different?’ Journal of Comparative Economics, 34(2). C 2010 The Authors C CIRIEC 2010 Journal compilation
FINANCIAL CONSTRAINTS AND TECHNICAL EFFICIENCY
37
PORTER P. and SCULLY G. W., 1987, ‘Economic efficiency in cooperatives’, Journal of Law and Economics, 30, October, 489– 512. RUSSELL R. and HANNEMAN R., 1992, ‘Cooperatives and business cycle: the Israeli case’, Journal of Comparative Economics, 16(4), 701–715. STIGLITZ J. and WEISS A., 1981, ‘Credit rationing in markets with imperfect information’, American Economic Review, 71(3), 393– 410. VAN BEKKUM O. F. and VAN DIJK G., 1998, Lo sviluppo delle cooperative agricole nell’Unione Europea, Ancona: Edizioni CLUA. TORTIA E., 2007, ‘Self-financing in labor-managed firms (lmfs): individual capital accounts and bonds’, Advances in the Economic Analysis of Participatory and Labor-Managed Firms, 10, 233–261. WHITED T. M., 1992, ‘Debt, liquidity constraints and corporate investment: evidence from panel data’, Journal of Finance, 47, 1425–1460.
` Contraintes financieres et efficacite´ technique. Le cas des ´ cooperatives de production italiennes
Dans cet article, les auteurs explorent la mesure dans laquelle les coop´eratives de production peuvent augmenter leur efficacit´ technique suite a` un resserrement des contraintes financi`eres. Cette hypoth`ese est test´ee a` l’aide d’une analyse de fronti`ere d’efficacit´e sur un e´ chantillon d’entreprises classiques et coop´eratives dans le secteur de la production et de la distribution de vin. Les r´esultats confirment l’hypoth`ese qu’une pression financi`ere croissante peut positivement affecter la performance des coop´eratives.
¨ Finanzielle Zwange und technische Effizienz. Empirische ¨ Evidenz bezuglich italienischer Produktionsgenossenschaften
In diesem Beitrag wird getestet, in welchem Ausmaß bei Produktionsgenossenschaften eine Verscharfung ¨ finanzieller Zwange ¨ zu einem Anstieg der technischen Effizienz fuhren ¨ kann. Diese Hypothese wird getestetan einer Stichprobe konventioneller und genossenschaftlicher Unternehmen im Sektor der Weinproduktion und -verarbeitung in Italien, unter Anwendung der Frontier Analyse. Die Ergebnisse C 2010 The Authors C CIRIEC 2010 Journal compilation
38
ORNELLA WANDA MAIETTA AND VANIA SENA
stutzen ¨ die Hypothese, dass zunehmender finanzieller Druck die Effizienz von Genossenschaften positiv beeinflussen kann.
´ Apremios financieros y eficiencia tecnica. El caso de las ´ italianas cooperativas de produccion
En este art´ıculo los autores exploran la medida en la que las cooperativas de producci´on pueden aumentar su eficiencia t´ecnica tras una mayor presi´on de los apremios financieros. Esta hip´otesis se ha verificado con la ayuda de un analisis ´ de frontera de eficiencia sobre una muestra de empresas clasicas ´ y de cooperativas en el sector de la producci´on y de la distribuci´on de vino. Los resultados confirman la hip´otesis de que una presi´on financiera creciente puede afectar positivamente los resultados de las cooperativas.
C 2010 The Authors C CIRIEC 2010 Journal compilation
