The significance of market transaction costs and technical efficiency for economic performance
Heinrich Hockmann* and Ekaterina Gataulina**
*Institute of Agricultural Development in Central and Eastern Europe (IAMO)
Theodor Lieser Straße 2
06120 Halle, Germany
**All-Russian Institute of Agrarian Problems and Information Theory (VIAPI)
PO Box 342 Bolshoi Kharitonievsky per., 21-6
Moscow, 105064, Russian Federation
The paper investigates the significance of internal transaction costs (or inefficiency) and risk in agriculture in the Tatarstan Republic. The analysis is conducted for independent farms and farms which are members of a business group. Inefficiency in agroholding members is higher than in independent farms. However, the estimation suggests that this result is due to more intense risk management in farms which belong to a business group. Thus, members of a business group have a more intense use of inputs; however, these are rather allocated to reduce uncertainty of production than to increase production.
Keywords: Risk production function, internal transaction costs, efficiency
JEL Classification: Q110, D220, P230
Several studies have revealed that Russia's agricultural sector is lagging behind the development of other sectors of the economy (Voigt and Hockmann 2008). The reasons for this phenomenon are intensively discussed among economists and politicians. In this paper we will contribute to this debate and analyse the significance of risk and internal transaction costs as well as their influence on production growth. The analysis will be conducted for Tatarstan Republic.
Internal transaction costs determine the degree to which producers are able to exploit production possibilities. Thus, technical inefficiency can be regarded as an indicator of internal transaction costs. Risk leads to a variation of agricultural production around the average. This component basically results from the variation of natural conditions, e. g. weather. These indicators will be investigated for different organisational forms in order to assess whether productivity differences among agricultural enterprises are determined by the choice of technology or basically by ownership and governance structures. Thus we will also contribute to the question whether the occurrence of horizontally and vertically integrated structures (often called agroholdings or business groups) have had a positive effect on agricultural production.1 The methodological framework is provided by stochastic frontier analysis.
The paper is organised in five chapters. The next chapter presents the theoretical and methodological background applied in the empirical part. Chapter 3 describes the data set used in our analysis. The results are presented in Chapter 4 where we discuss estimation results and the implication of the contribution of technology, risk and inefficiency to production. In Chapter 5 a summary of our finding is presented and policy conclusions are drawn.
2 Methodological considerations
In the empirical analysis we use a risk production function. This concept was originally introduced by Just and Pope (1978) and extended by Kumbhakar (2002):2
(1) , with and .
f(x,t,m) mean production function
g(x,d,m) risk function
q(x,m) inefficiency function
We decompose the variation of output into three components. First there is technology or the mean production function f, which represents the average impacts of inputs (x) on production. In addition we add a trend variable (t) which captures the impact of technical change and a dummy variable for organisational form (m). We assume that the natural logarithm of the mean production has a translog from.
The second component g is assumed to capture the effects of risk on production. Due to poor or favourable weather conditions actual output can be lower or higher than its average level. Thus, it is straightforward to connect the risk function with a two-sided error component (v). The risk function is assumed to consist of two parts. The first affects all farms similarly, e.g. thus is generic. We follow Bokusheva and Hockmann (2006) and consider this kind of risk by a constant and dummy variables for the years 2006 and 2008 (d06, d08). The second part of g is farm-specific or idiosyncratic and depends on the intensity and structure of input use. The idiosyncratic component is represented by a Cobb Douglas function:
The function q captures the impact of factor use on the exploitation of the production possibilities or the efficiency of production. This function transforms the one-sided error term u. The inefficiency function q isalso represented by a Cobb Douglas function:
Estimation is conducted using a modification of the standard workhorse of stochastic frontier analysis. Details on the implementation and the likelihood function can be found in Kumbhakar (2002).
We use accountancy data of agricultural enterprises in the Tatarstan Republic for the period of 2006-2008 (Rosstat data provided by VIAPI). First, we excluded farms for which we had only one observation. Second, we excluded observations with nonsense partial productivities, e. g. when land productivity or milk production per cows was 100 time larger than average. This cleaning resulted in a data set of 277 farms and 636 observations. The set contains 41 members of agroholdings; they account for 101 observations.
The data set contains detailed information on production structures, specialisation and factor input. In addition, the data provide information on organisational forms and thus governance structures.
Inputs comprise land (Lan), labour (Lab), capital (Cap) and materials (Mat). The first and second are given by used agricultural area and the number of workers, respectively. Capital input was approximated by depreciation. We constructed the variable by adding depreciation of capital use in crop and animal production, each deflated by the corresponding regional price indices for machinery. Materials comprise all expenses for variable inputs. The data base provided only information in current prices. Volumes were constructed by (a) weighting the individual components (seed, fertilizer, feedstuff ...) by corresponding regional price indices and (b) adding up the individual volumes.
Our output variable represents the volume of gross production. This variable was constructed in several steps. First, implicit firm specific product prices can be calculated from the data set using the quantities and sales of marketed products. Second, gross production in current price was estimated by adding the products of gross production in physical terms and firm specific product prices. We distinguish between nine categories of production (cereals, sugar beet, sunflower, potatoes and vegetables as well as beef, pork, lamb, poultry, milk, meat, egg, wool and dairy). In a third step, we calculated firm specific multi-lateral consistent price indices using the approach developed by Caves et al. (1982).3 In doing so, we used firm specific product prices and firm specific revenue shares. Finally, we deflated gross production in current values by the firm specific output price indices.
All variables were normalized by their geometric mean. Due to this procedure, the parameter estimates for the first order terms can be directly interpreted as the production elasticities at the sample mean.
4 The sources of quantity variation
4.1 Estimation results
Parameter estimates of the risk production function are given in Table 1. Most parameters in the mean production as well as the risk and inefficiency functions are highly significant. Thus, it can be concluded that the omission of the two latter would have produced biased estimates for the production function.
The mean production function fulfils the monotonicity requirement (αi > 0, for i = A, L, C, V), e. g. an increase in input use leads to higher production. However, the mean production function does not exhibit the quasi-concave requirement in all inputs. We found that materials and capital follow the "law" of diminishing returns (αii + αi2 - αi < 0, for i = C, V) for independent farms, though land and labour showed increasing returns (αii + αi2 - αi > 0, for i = A, L) at the sample mean. However, the estimates indicate that the implementation of management techniques cannot be denied. In turn, these might be responsible for the violation of quasi-concavity.
Elasticity is defined as the relation of marginal and average productivities. Because of the normalisation average productivities at the sample mean are equal to one. Thus, the estimated values provide direct information about the marginal products. The low estimate of labour suggests that on average, independent farms might operate with a suboptimal high labour input and thus are putting on average a relatively high weight to their social function in rural areas4. Agroholding membership have a higher marginal product of labour (αAM > 0). Since the "law" of diminishing returns holds for agroholdings, this suggests labour input is considerably lower in agroholdings than in independent farms. This result supports the often heard proposition that agroholdings are more profit-oriented and release underemployed labour.
The results also support the view that agroholdings have a better access to material inputs than independent farms. Marginal products in agroholdings are significantly lower than in independent farms (αVM < 0). Again, the law of diminishing returns implies that agroholding members apply material inputs more intensively than independent farms. Similar conclusions can be deduced for land and capital input, though the corresponding estimates are not significant.
Table 4: Parameter estimates of the risk production function
Note: t-ratios of 1.66, 2.04, and 2.72 are the critical values for the 10 %, 5 % and 1 % levels of significance, respectively.
Source: Own estimations.
Independent farms as well as agroholdings operate at almost constant returns to scale (Σαi ≈ 1, i = A, L, C, V). In addition, we neither find a significant impact of neutral technical change (αT and αTT) nor a special impact of agroholding membership on technical change (αTM). However, the estimates suggest that technical change was strongly biased (αiT, i = A, L, C, V). It was labour and land using and capital and material input saving.
All inputs except land had a risk increasing effect. Since the coefficients differ in sign, it has to be concluded that the farms apply risk management techniques. Agroholding membership appears to increase production risk. This result is consistent with the more intense use of purchased inputs in this organisational structure. In addition, generic risk did not differ sigificantly in the years under investigation.
Moreover, the estimate of the constant implies that generic risk is more important than generic inefficiency (σu). The estimate of -1.611 implies a standard deviation of the two-sided error term of about 0.2.5 This is almost two times higher than the standard deviation of the efficiency distribution. According to our estimates efficiency is significantly affected by all inputs except capital. Labour and land input increase efficiency while materials tend to decrease efficiency. The effect regarding labour is consistent with the conclusion derived for the mean production function where differences in labour input were justified by the different perception of social function in independent farms and agroholding members.
In addition, the results suggest that group members were better positioned than independent farms regarding the exploitation of production possibilities. This result is consistent with the findings of Hockmann et al. (2009) who pointed out that agroholdings usually change the managerial structure and adopt modern management structures that allow for better monitoring of production processes.
4.2 Variance decomposition
Table 2 contains our results regarding the variance decomposition of total production6. The first lines for the organisational forms indicate how much of the variance of production is explained by the variance of the individual factors. First, in most years, considering variances only would overestimate the variance of total production. Furthermore, in most of the years the variance of the mean production function is already larger than the variance of production. This implies that the covariances between mean production, risk and inefficiency have a homogenizing effect on the variation of total production.
Table 3: Decomposition of the variance of total production
Note: Mean production was decomposed using equation (1). The contributions are calculated in relation to the total variance of mean production.
Source: Own estimations.
Second the by far major part of production variance stems from mean production. This result simply underlines the importance of farm size and the implied effects on specialisation and input intensities. The differences in productivity play a less important role and explain only about one fourth of the variance of mean production. However, the variation of partial productivities e. g. technology is far more important than the variation of risk and inefficiency.
Risk contributes considerably more to production variability than inefficiency. Interestingly, the impact of risk in agroholdings is lower than the impact in independent farms. At first glance this is surprising since Table 1 reports that agroholding membership has a positive impact on risk (γm). This was explained by the higher intensity of input use, in particular, purchased inputs. Table 2 implies that this intensity effect on risk was more than compensated by employing risk management techniques. However, this benefit was bought at the cost of higher input use which induces a lower level of efficiency7.
5 Conclusion and interpretation
We analysed the significance of risk and technical efficiency on agricultural development and production growth in Tatarstan Republic using bookkeeping data for the period of 2006-2008. Both effects are highly important. Moreover, agroholding membership significantly affects the production structures. Members of a business group have better access to purchased inputs and use them more intensively than independent farms. Labour input is lower in holding members suggesting that this group pays less attention to the social function of farms in rural areas. Technical change was strongly biased, e. g. labour and land using as well as capital and material input saving. Neutral technical change was insignificant.
The estimates confirm that the agricultural enterprises conduct risk management. Consistent with the more intense use of inputs, agroholding membership appears to have a higher (generic) production risk. However, the idiosyncratic effects imply that agroholdings apply risk management techniques more intensively than independent farms. This results in a lower contribution of the risk component to total production variance. Moreover, this finding is consistent with outcomes from earlier studies which highlight the change in management and the adoption of modern management strategies observed for agroholdings.
Generic inefficiency in agroholdings is significantly smaller than for independent farms. However, the idiosyncratic component changes the relation leading to higher inefficiency for group members. The estimation suggests that this result is due to more intense risk management in agroholding members. Thus, members of a business group use inputs more intensively; however, these are rather allocated to reduce uncertainty of production than to increase production.
The estimates provide further that risk is of higher relevance for the variation of production than inefficiency. This finds its expression not only in the generic but also in the idiosyncratic component. Moreover, for independent farms risk and inefficiency explain about the same amount of production variance than technology differences. For holding members the differences in technology were the major source of output variations.
Thus, for improving the production conditions, agricultural policy is required to tackle both issues using a mix of appropriate policy measures. One option would be the support of holding membership. However, the benefits come at the cost of higher inefficiency. However, nor farms should not be forced be become a member of an agroholding, neither agroholdings should be forced to accommodate more members. The term "support" is meant more in an evolutionary meaning in such a way that policy should provide an environment in which farms and holding can decide about their strategies. Moreover, risk, inefficiency, and technology are not necessarily substitutes. A system of measures needs to be defined and implemented which improves all three indicators. In this sense, fostering membership in agroholdings might be only regarded as a second best solution.
Bokusheva, R. and Hockmann, H. (2006): Production risk and technical inefficiency in Russian agriculture. European Review of Agricultural Economics 33: 93-118.
Caves, D. W., Christensen, L. R., and Diewert, W. E. (1982): Multilateral comparisons of output, input, and productivity using superlative index numbers. The Economic Journal 92: 73-86
Hockmann, H., Bokusheva, R., Bezlepkina, I. (2009): Agroholding membership: does that make a difference in performance? Quarterly Journal of International Agriculture. 48 (1): 25-46.
Jondrow, J., Lovell, C. A. K., Materov I. S. and Schmidt, P. (1982): On the estimation of technical inefficiency in the stochastic frontier production function model. Journal of Econometrics 19: 233-38.
Just, R. E. and Pope, R. D. (1978): Stochastic representation of production functions and econometric implications. Journal of Econometrics 7: 67–86.
Koester, U. (2005): A revival of large farms in Eastern Europe? - How important are institutions? in: Colman, D. and N. Vink (eds.), Reshaping Agriculture's Contributions to Society. Proceedings of the Twenty-Fifth International Conference of Agricultural Economists 2003 in Durban, South Africa. Blackwell Publishing 2005, pp.103-114.
Kumbhakar, S. C. (2002): Specification and estimation of production risk, risk preferences and technical efficiency. American Journal of Agricultural Economics 84: 8–22.
Mood, A. M., Graybill, F. A., and Boes, D. C. (1974): Introduction to the Theory of Statistics. McGraw-Hill.
Voigt, P. and Hockmann. H. (2008): Russia's transition process in the light of a rising economy: Economic trajectories in Russia's industry and agriculture. European Journal of Comparative Economics 2 (5): 251-267
Wandel, J. (2010): Integrierte Strukturen im Agrar- und Ernährungssektor Russlands: Entstehungsgründe, Funktionsweise, Entwicklungsperspektiven und volkswirtschaftliche Auswirkungen". Habilitationsschrift. Martin - Luther - Universität Halle - Wittenberg.
1On this issue see for instance Kolnesnikov (2009), Wandel (2010) and Hockmann et al. (2009)
2 In the following, bold symbols indicate vectors or matrices; all other variables are scalars. Subscripts will be omitted in order to improve readability.
3 Assuming a translog aggregator function, the result is a Törnquist-Theil Index. Basically, by this approach each observation is compared to the average in the sample.
4See Koester (2005) for more details.
5The assumptions in (1) and the functional forms in (1b) provide .
6 Output variance was decomposed with by , where c(y,p) represents the aggregated effects of the covariance (Mood et al. 1974).
7 Lower efficiency of agroholding members is due to the impact of the idiosyncratic effects (i, i=A.L.C,V). These overcompensate the positive generic effect of agroholding membership (M).
Назад в раздел