SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY Marie Pechrová, Tomáš Hlavsa, Martin Hruška Institute of Agriculture Economics and Information [email protected] , [email protected], [email protected] Keywords: mountain areas – subsidies – Stochastic Frontier Analysis – technical efficiency – “True” Fixed Effects model JEL classification: H71, H21 Abstract: The aim of the paper is to assess the influence of subsidies to farms in mountainous areas under Common Agricultural Policy of the EU on their technical efficiency. Parametric Stochastic Frontier Analysis is used to estimate production function of farms in Cobb-Douglas form. “True” Fixed Effects model for panel data assumed truncatednormal distribution of the inefficiency term. The subsidies per hectare (SAPS, Top-Up and LFA for mountains areas) were included into variance of inefficiency function in order to assess the direction of their influence. Average efficiency during 2008–2012 amounted to 81.16% suggesting that there is still space for improvement of farm management. All subsidies had undesirable impact as they increased the variance of inefficiency. Also the correlation between LFA subsidies and efficiency was negative. Introduction The implementation of the principles of Common Agricultural Policy (CAP) after the Entrance of the Czech Republic to European Union (EU) had many consequences for agricultural sector. Beside other facts, Czech farmers gain the possibility to obtain subsidies in a form of direct payments (single area payment scheme – SAPS toped-up by national support – Top-Up) and from Rural Development Programmes (mostly of project character). A special category of subsidies is area payments (on acreage of agricultural land) provided on permanent grasslands in Less Favoured Areas (LFA). Those areas represent around 50% of the total agricultural land and are divided into three types: mountainous areas, other LFA, and the areas affected by specific handicaps (Štolbová and Hlavsa, 2008). Pechrová (2014) shows that after the entrance to the EU “the subsidies helped to increase the profitability of a sector as a whole as same as of particular commodities. The justification for support of farms in mountainous areas is given by their unfavourable conditions. The aim of the payments is to maintain agriculture and the use of land by the means of adequate but not excessive financial compensations (Štolbová and Hlavsa, 2008). Together with Kroupová and Malý (2010) we may state stated “it is necessary to continuously analyse the efficiency of spent

800

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

finances in relation to the gained added value.” Therefore, the aim of the paper is to assess the influence of subsidies provided to farms in mountainous areas under Common Agricultural Policy of the European Union on their technical efficiency. The paper is structure as follows. First section review previous researches on the topic. Second introduces used methods for the analysis. Then the results are described and discussed in the next section. Last section concludes. 1. Literature overview The subsidies should support the performance of the agricultural holdings, should enhance the efficiency of their functioning (Technical efficiency in this sense is understood as defined by Pitt and Lee (1981) as the ability to produce “the maximum quantity of output attainable from given inputs”.) and increase their resilience. It is expected that they should contribute to better farm management. However, this is not always true and additional financial means can have negative consequences. For example Lososová and Zdeněk (2014) found out that “the dependence of profit on subsidies is greatest in the mountain LFA.” Štolbová and Hlavsa (2008) discovered that economic results of even all agricultural holdings (not only those located in LFA) are more and more dependent on subsidies. Influence of agro-environmental subsidies on eco-efficiency of farms in Campos region, Spain, was assessed by Picazo-Tadeo, Gómez-Limón and Reig-Martínez (2011). Using truncated regression and bootstraping procedure they came to the conclusion that the eco-efficiency is positively influence by the fact that the farmer participates in agro-environmental programs. It this sense, the subsidies fulfil their objectives. Similarly Juan, Sperlich, Kleinhanss and Murillo (2005) examined the determinants influencing farms’ DEA efficiency index. They found out that direct payments tend (in absolute terms) to increase the efficiency. This type of subsidies positively correlates with environmental friendly production and awarding positive externalities, which are generated in agriculture. On the other hand, direct payments are not sufficient instrument which would correct the fact that the most efficient farms produce less environmental friendly. 2. Methods First approaches towards the estimation of the production function’s frontier were based on linear and quadratic programming techniques. The parametric approach – Stochastic Frontier Analysis (SFA) – originated with work of Meeusen and van den Broeck (1977) and Aigner, Lovell and Schmidt (1977). In this paper, we utilized SFA method as it accounts for panel data. Firstly, the production function was estimated in Cobb-Douglas form (i.e. power function – see Pechrová, 2015) which has several important features: it assumes constant returns to scale (with increasing inputs, the output grows proportionally), can be linearized by natural logarithms, and (therefore) the coefficients can be interpreted as elasticity. The functional form is stated below (1).

801

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

yit = a i x1b,it1 x2b,2it ...xkb,kit e it

,

(1)

where a i is a group specific constant, yit denotes production of a farm i in time t, and x k ,it (k = 1,... K) represents K explanatory variables powered by elasticity coefficients

b k . A stochastic term ε is time and individual variant and consists of idiosyncratic term ( vit ) and inefficiency term ( u it ). We assumed that the first mentioned follows normal distribution and the later one truncated-normal distribution. Production ( yit – where i (i = 1, … n) denotes particular farm in time t) was represented by sales of own products and services in thous. CZK. Its amount was deflated by the agricultural producers’ prices (2005 = 100). Explanatory variables were: consumed material and services ( x1,it ) and capital ( x 2 ,it ), both deflated by industrial producers’ prices (2005 = 100), labour measured in number of workers ( x 3,it ), and land in hectares ( x 4 ,it ). SAPS ( z1,it ), TopUp ( z 2 ,it ), and LFA subsidies ( z 3,it ) together with constant were included as explanatory variables in the function of variance of the inefficiency term ( s u2 ), which i

model heteroscedasticity in farms’ data. The subsidies should have negative sign as we suppose that they decrease the variance of inefficiency. Mean of inefficiency ( m u2 ) i

explained heterogeneity among farms only by a constant. “True” fixed-effects model (TFE) elaborated by Greene (2002) was estimated in the form (2).

yit = ai + b T x it + vit - uit

(2)

where the variables have the same meaning as above. We applied maximum likelihood estimation method. Secondly, the efficiency was calculated as suggested by Jondrow, Lovell, Materov and Schmidt (1982). The data and results were described by descriptive statistics. The normality of distribution of efficiency was tested by Shapiro-Wilk test. As it had non-normal distribution, the correlation was assessed by Spearman’s rank coefficient. Difference in technical efficiency between farms which receive above average and below average subsidies for LFA was tested using Wilcoxon rank sum test. Null hypothesis stated that there are no differences in average technical efficiency between them. We utilized panel data of 223 farms observed during years 2008 to 2012. Farms were selected according to their location in mountainous areas. There were 962 observations ranging from 2 to 5 with average 4.3. Accountancy data were obtained from database Albertina of Bisnode ltd., data about subsidies from State Agricultural Interventional Fund and about land from Land Parcel Identification System. The sample is described at TAB. 1. Average farm produced goods in value of 42.3 mil. CZK per year. Use of capital (long term assets was almost five times higher. Regarding the size of the farm, it

802

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

is unbalanced in the sample. There are few big farms and a lot of smaller ones. Average agricultural holding had 52 employees, but 90% of them had less than 75 workers. There were few bigger farms as average acreage amounted to 1 022 hectares, but more than half of the farms were smaller than 846 hectares. All firms receive SAPS (on average 4186 CZK/ha), but not all of them obtained Top-Up. The sample was divided to the farms which receive more than average of LFA subsidies paid in mountainous areas (2 568 CZK/ha) and to those who receive less. The biggest part of the farms is located in mountainous area, the higher subsidies per hectare it obtains (Pearson correlation coefficient = 0.46 is statistically significant and points on relatively strong relation). TAB. 1: Descriptive statistics of the sample of Czech farms in mountainous areas

Source: own elaboration

3. Results Firstly, a production function was constructed. Secondly, the technical efficiency of the agricultural holdings was assessed. Finally, the direction of the influence of subsidies on technical efficiency was assessed from the function of the variance of technical inefficiency, by correlation coefficient, and by testing whether the technical efficiency of agricultural holdings statistically significantly differ when they receive above or below average LFA subsidies for mountainous areas. The results of TFE model estimation are displayed at TAB 2. It can be seen that all parameters were statistically significant at 1% level of significance and have expected sign. Increase of material and services, capital, labour or land by 1% cause increase of production by 0.92%, 0.07%, 0.001% and 0.14%, respectively. The sign for subsidies was positive showing that the increase of subsidies increases the variance of inefficiency among farms. This is not desirable situation. However, the coefficients had quite small values. Therefore, the correlation between the subsidies and the efficiency was examined next. Average efficiency was estimated at 81.16% which shows that farms still can improve their performance by 18.84%. The efficiency decreased between 2008 and 2009, but grew later (to 86.49% in 2012). Shapiro-Wilk test rejected the hypothesis that the data were normally distributed. Spearman correlation coefficients

803

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

between technical efficiency and SAPS, Top-Up and LFA in mountainous areas amounted to 0.07 ** , -0.14 *** , and -0.09 *** . As all of them were statistically significant, we may conclude that there was weak positive relationship between SAPS and technical efficiency and weak negative relationship between Top-Up and LFA paid to farms in mountainous areas and the efficiency. TAB. 2: Results of TFE model estimation, truncated-normal distribution of u it

Source: own elaboration; Note: Statistical significance: *** at α = 0.01, ** at α = 0.05 and * at α = 0.1

The first result is in contrast to previous observation from inefficiency variance function. When dividing the sample in two groups, having tested it by Wilcoxon ranksum test, we found that probability that the efficiency of farms which receives above average subsidies is equal to technical efficiency of farms that obtain below average LFA payments is 0.03%. On 5% level of significance we may reject the null hypothesis, but on 1% level, the null hypothesis still holds. Hence, the ambiguous results do not allow us to clearly assess the impact of subsidies. 4. Discussion Comparing our results to similar researches, we may conclude that finding ambiguous or not positive influence of subsidies on the performance of farms is common. On one hand, Pechrová (2015) found out that the RDP subsidies have a positive and statistically significant impact on the technical efficiency. On the other hand, she admits that the effect was statistically significant only at 90% level of significance. Kroupová and Malý (2010) suggested cancelling the payment on permanent grassland and decrease of payments on arable land while keeping the level of support for organic agriculture in order to increase the profit and production of organic farms. Otherwise, their results “indicate negative impact of subsidies on production, profit and technical efficiency of organic farmers and refer to the reality that actual level of subsidy discourages organic farmers from rational behaviour and implicates their dependence on state support” (Kroupová and Malý, 2010). Similarly results of the analysis done by Štolbová and Hlavsa (2008) show “...the ever rising dependence of the holdings economic results on

804

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

subsidies, and not only those for less-favoured areas, but even for the group of the FADN holdings operating outside the LFA”. Conclusion The aim of the paper was to assess the influence of LFA subsidies paid to farms in mountainous areas on their technical efficiency. Using Stochastic Frontier Analysis with production function in Cobb-Douglas form a “True” Fixed Effects model for panel data was estimated. The subsidies per hectare (SAPS, Top-Up and LFA for mountains areas) were included into variance of inefficiency function. A negative direction of their influence was found. Average efficiency during 2008–2012 was 81.16% suggesting that there is still space for improvement of farm management. The results of the analysis of the influence of LFA subsidies on technical efficiency of farms in mountainous areas in the Czech Republic are ambiguous. On one hand, they increased the variance of technical inefficiency and the correlation between Top-Up and LFA in mountainous areas and technical efficiency was negative, but, on the other hand, the correlation between SAPS and technical efficiency was found positive. In both cases, the values of the coefficients are around zero. Therefore, the examination of the influence on other farms’ performance indicators is needed. Acknowledgement: The research was financed from internal research grant (IVP) number 1297/2015 of Institute of Agricultural Economics and Information. References: Aigner, D., Lovell, C. A. K., & Schmidt, P. (1977). Formulation and estimation of stochastic frontier production function models. Journal of Econometrics, 6(1), 21-37. Greene, W. (2002). Fixed and Random Effects in Stochastic Frontier Models. New York: Stern School of Business. Retrieved from http://people.stern.nyu.edu/wgreene/ fixedandrandomeffects.pdf. Jondrow, J., Lovell, C. A. K, Materov, I. S., & Schmidt, P. (1982). On the Estimation of Technical Inefficiency in the Stochastic Frontier Production Function Model. Journal of Econometrics, 19(1982), 233-238. Juan, C. S., Sperlich, S., Kleinhanss, W., & Murillo, C. (2005). Eƥciency, subsidies and environmental adaptation of animal farming under CAP. Retrieved from http://128.118.178.162/eps/othr/papers/0512/0512015.pdf. Kroupová, Z. and Malý, M. (2010). Analýza nástrojů zemědělské dotační politiky aplikace produkčních funkcí. Politická ekonomie, 6(2010), 774-794.

805

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

Lososová , J., & Zdeněk, R. (2014). Key Factors Affecting the Profitability of Farms in the Czech Republic. Agris on-line Papers in Economics and Informatics, 6(1), 21-36. Meeusen, W., & van den Broeck, J. (1977). Technical efficiency and dimension of the firm: Some results on the use of frontier production functions. Empirical Economics, 2(2), 109-122. Pechrová, M. (2014). Impact of EU Membership on the Development of the Czech Agriculture. In Proceedings of the 2nd International Conference on European Integration, Ostrava, Czech Republic, June 2014 (pp. 545 – 552). VŠB –Technical University of Ostrava. Pechrová, M. (2015). Impact of the Rural Development Programme Subsidies on the farms’ inefficiency and efficiency. Agricultural Economics, 61(5), 197-204. Picazo-Tadeo, A. J., Gómez-Limón, J. A., & Reig-Martínez, E. (2011). Assessing farming eco-efficiency: A Data Envelopment Analysis approach. Journal of Environmental Management, 92(2011), 1154-1164. Pitt, M., & Lee, L-F. (1981). The Measurement and Sources of Technical Inefficiency in the Indonesian weaving Industry. Journal of Development Economics, 1981(9), 4364. Štolbová, M., & Hlavsa, T. (2008). The impact of the LFA payments on the FADN farms in the Czech Republic. Agricultural Economics - Czech, 54 (10), 489–497.

806

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY Marie Pechrová, Tomáš Hlavsa, Martin Hruška Institute of Agriculture Economics and Information [email protected] , [email protected], [email protected] Keywords: mountain areas – subsidies – Stochastic Frontier Analysis – technical efficiency – “True” Fixed Effects model JEL classification: H71, H21 Abstract: The aim of the paper is to assess the influence of subsidies to farms in mountainous areas under Common Agricultural Policy of the EU on their technical efficiency. Parametric Stochastic Frontier Analysis is used to estimate production function of farms in Cobb-Douglas form. “True” Fixed Effects model for panel data assumed truncatednormal distribution of the inefficiency term. The subsidies per hectare (SAPS, Top-Up and LFA for mountains areas) were included into variance of inefficiency function in order to assess the direction of their influence. Average efficiency during 2008–2012 amounted to 81.16% suggesting that there is still space for improvement of farm management. All subsidies had undesirable impact as they increased the variance of inefficiency. Also the correlation between LFA subsidies and efficiency was negative. Introduction The implementation of the principles of Common Agricultural Policy (CAP) after the Entrance of the Czech Republic to European Union (EU) had many consequences for agricultural sector. Beside other facts, Czech farmers gain the possibility to obtain subsidies in a form of direct payments (single area payment scheme – SAPS toped-up by national support – Top-Up) and from Rural Development Programmes (mostly of project character). A special category of subsidies is area payments (on acreage of agricultural land) provided on permanent grasslands in Less Favoured Areas (LFA). Those areas represent around 50% of the total agricultural land and are divided into three types: mountainous areas, other LFA, and the areas affected by specific handicaps (Štolbová and Hlavsa, 2008). Pechrová (2014) shows that after the entrance to the EU “the subsidies helped to increase the profitability of a sector as a whole as same as of particular commodities. The justification for support of farms in mountainous areas is given by their unfavourable conditions. The aim of the payments is to maintain agriculture and the use of land by the means of adequate but not excessive financial compensations (Štolbová and Hlavsa, 2008). Together with Kroupová and Malý (2010) we may state stated “it is necessary to continuously analyse the efficiency of spent

800

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

finances in relation to the gained added value.” Therefore, the aim of the paper is to assess the influence of subsidies provided to farms in mountainous areas under Common Agricultural Policy of the European Union on their technical efficiency. The paper is structure as follows. First section review previous researches on the topic. Second introduces used methods for the analysis. Then the results are described and discussed in the next section. Last section concludes. 1. Literature overview The subsidies should support the performance of the agricultural holdings, should enhance the efficiency of their functioning (Technical efficiency in this sense is understood as defined by Pitt and Lee (1981) as the ability to produce “the maximum quantity of output attainable from given inputs”.) and increase their resilience. It is expected that they should contribute to better farm management. However, this is not always true and additional financial means can have negative consequences. For example Lososová and Zdeněk (2014) found out that “the dependence of profit on subsidies is greatest in the mountain LFA.” Štolbová and Hlavsa (2008) discovered that economic results of even all agricultural holdings (not only those located in LFA) are more and more dependent on subsidies. Influence of agro-environmental subsidies on eco-efficiency of farms in Campos region, Spain, was assessed by Picazo-Tadeo, Gómez-Limón and Reig-Martínez (2011). Using truncated regression and bootstraping procedure they came to the conclusion that the eco-efficiency is positively influence by the fact that the farmer participates in agro-environmental programs. It this sense, the subsidies fulfil their objectives. Similarly Juan, Sperlich, Kleinhanss and Murillo (2005) examined the determinants influencing farms’ DEA efficiency index. They found out that direct payments tend (in absolute terms) to increase the efficiency. This type of subsidies positively correlates with environmental friendly production and awarding positive externalities, which are generated in agriculture. On the other hand, direct payments are not sufficient instrument which would correct the fact that the most efficient farms produce less environmental friendly. 2. Methods First approaches towards the estimation of the production function’s frontier were based on linear and quadratic programming techniques. The parametric approach – Stochastic Frontier Analysis (SFA) – originated with work of Meeusen and van den Broeck (1977) and Aigner, Lovell and Schmidt (1977). In this paper, we utilized SFA method as it accounts for panel data. Firstly, the production function was estimated in Cobb-Douglas form (i.e. power function – see Pechrová, 2015) which has several important features: it assumes constant returns to scale (with increasing inputs, the output grows proportionally), can be linearized by natural logarithms, and (therefore) the coefficients can be interpreted as elasticity. The functional form is stated below (1).

801

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

yit = a i x1b,it1 x2b,2it ...xkb,kit e it

,

(1)

where a i is a group specific constant, yit denotes production of a farm i in time t, and x k ,it (k = 1,... K) represents K explanatory variables powered by elasticity coefficients

b k . A stochastic term ε is time and individual variant and consists of idiosyncratic term ( vit ) and inefficiency term ( u it ). We assumed that the first mentioned follows normal distribution and the later one truncated-normal distribution. Production ( yit – where i (i = 1, … n) denotes particular farm in time t) was represented by sales of own products and services in thous. CZK. Its amount was deflated by the agricultural producers’ prices (2005 = 100). Explanatory variables were: consumed material and services ( x1,it ) and capital ( x 2 ,it ), both deflated by industrial producers’ prices (2005 = 100), labour measured in number of workers ( x 3,it ), and land in hectares ( x 4 ,it ). SAPS ( z1,it ), TopUp ( z 2 ,it ), and LFA subsidies ( z 3,it ) together with constant were included as explanatory variables in the function of variance of the inefficiency term ( s u2 ), which i

model heteroscedasticity in farms’ data. The subsidies should have negative sign as we suppose that they decrease the variance of inefficiency. Mean of inefficiency ( m u2 ) i

explained heterogeneity among farms only by a constant. “True” fixed-effects model (TFE) elaborated by Greene (2002) was estimated in the form (2).

yit = ai + b T x it + vit - uit

(2)

where the variables have the same meaning as above. We applied maximum likelihood estimation method. Secondly, the efficiency was calculated as suggested by Jondrow, Lovell, Materov and Schmidt (1982). The data and results were described by descriptive statistics. The normality of distribution of efficiency was tested by Shapiro-Wilk test. As it had non-normal distribution, the correlation was assessed by Spearman’s rank coefficient. Difference in technical efficiency between farms which receive above average and below average subsidies for LFA was tested using Wilcoxon rank sum test. Null hypothesis stated that there are no differences in average technical efficiency between them. We utilized panel data of 223 farms observed during years 2008 to 2012. Farms were selected according to their location in mountainous areas. There were 962 observations ranging from 2 to 5 with average 4.3. Accountancy data were obtained from database Albertina of Bisnode ltd., data about subsidies from State Agricultural Interventional Fund and about land from Land Parcel Identification System. The sample is described at TAB. 1. Average farm produced goods in value of 42.3 mil. CZK per year. Use of capital (long term assets was almost five times higher. Regarding the size of the farm, it

802

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

is unbalanced in the sample. There are few big farms and a lot of smaller ones. Average agricultural holding had 52 employees, but 90% of them had less than 75 workers. There were few bigger farms as average acreage amounted to 1 022 hectares, but more than half of the farms were smaller than 846 hectares. All firms receive SAPS (on average 4186 CZK/ha), but not all of them obtained Top-Up. The sample was divided to the farms which receive more than average of LFA subsidies paid in mountainous areas (2 568 CZK/ha) and to those who receive less. The biggest part of the farms is located in mountainous area, the higher subsidies per hectare it obtains (Pearson correlation coefficient = 0.46 is statistically significant and points on relatively strong relation). TAB. 1: Descriptive statistics of the sample of Czech farms in mountainous areas

Source: own elaboration

3. Results Firstly, a production function was constructed. Secondly, the technical efficiency of the agricultural holdings was assessed. Finally, the direction of the influence of subsidies on technical efficiency was assessed from the function of the variance of technical inefficiency, by correlation coefficient, and by testing whether the technical efficiency of agricultural holdings statistically significantly differ when they receive above or below average LFA subsidies for mountainous areas. The results of TFE model estimation are displayed at TAB 2. It can be seen that all parameters were statistically significant at 1% level of significance and have expected sign. Increase of material and services, capital, labour or land by 1% cause increase of production by 0.92%, 0.07%, 0.001% and 0.14%, respectively. The sign for subsidies was positive showing that the increase of subsidies increases the variance of inefficiency among farms. This is not desirable situation. However, the coefficients had quite small values. Therefore, the correlation between the subsidies and the efficiency was examined next. Average efficiency was estimated at 81.16% which shows that farms still can improve their performance by 18.84%. The efficiency decreased between 2008 and 2009, but grew later (to 86.49% in 2012). Shapiro-Wilk test rejected the hypothesis that the data were normally distributed. Spearman correlation coefficients

803

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

between technical efficiency and SAPS, Top-Up and LFA in mountainous areas amounted to 0.07 ** , -0.14 *** , and -0.09 *** . As all of them were statistically significant, we may conclude that there was weak positive relationship between SAPS and technical efficiency and weak negative relationship between Top-Up and LFA paid to farms in mountainous areas and the efficiency. TAB. 2: Results of TFE model estimation, truncated-normal distribution of u it

Source: own elaboration; Note: Statistical significance: *** at α = 0.01, ** at α = 0.05 and * at α = 0.1

The first result is in contrast to previous observation from inefficiency variance function. When dividing the sample in two groups, having tested it by Wilcoxon ranksum test, we found that probability that the efficiency of farms which receives above average subsidies is equal to technical efficiency of farms that obtain below average LFA payments is 0.03%. On 5% level of significance we may reject the null hypothesis, but on 1% level, the null hypothesis still holds. Hence, the ambiguous results do not allow us to clearly assess the impact of subsidies. 4. Discussion Comparing our results to similar researches, we may conclude that finding ambiguous or not positive influence of subsidies on the performance of farms is common. On one hand, Pechrová (2015) found out that the RDP subsidies have a positive and statistically significant impact on the technical efficiency. On the other hand, she admits that the effect was statistically significant only at 90% level of significance. Kroupová and Malý (2010) suggested cancelling the payment on permanent grassland and decrease of payments on arable land while keeping the level of support for organic agriculture in order to increase the profit and production of organic farms. Otherwise, their results “indicate negative impact of subsidies on production, profit and technical efficiency of organic farmers and refer to the reality that actual level of subsidy discourages organic farmers from rational behaviour and implicates their dependence on state support” (Kroupová and Malý, 2010). Similarly results of the analysis done by Štolbová and Hlavsa (2008) show “...the ever rising dependence of the holdings economic results on

804

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

subsidies, and not only those for less-favoured areas, but even for the group of the FADN holdings operating outside the LFA”. Conclusion The aim of the paper was to assess the influence of LFA subsidies paid to farms in mountainous areas on their technical efficiency. Using Stochastic Frontier Analysis with production function in Cobb-Douglas form a “True” Fixed Effects model for panel data was estimated. The subsidies per hectare (SAPS, Top-Up and LFA for mountains areas) were included into variance of inefficiency function. A negative direction of their influence was found. Average efficiency during 2008–2012 was 81.16% suggesting that there is still space for improvement of farm management. The results of the analysis of the influence of LFA subsidies on technical efficiency of farms in mountainous areas in the Czech Republic are ambiguous. On one hand, they increased the variance of technical inefficiency and the correlation between Top-Up and LFA in mountainous areas and technical efficiency was negative, but, on the other hand, the correlation between SAPS and technical efficiency was found positive. In both cases, the values of the coefficients are around zero. Therefore, the examination of the influence on other farms’ performance indicators is needed. Acknowledgement: The research was financed from internal research grant (IVP) number 1297/2015 of Institute of Agricultural Economics and Information. References: Aigner, D., Lovell, C. A. K., & Schmidt, P. (1977). Formulation and estimation of stochastic frontier production function models. Journal of Econometrics, 6(1), 21-37. Greene, W. (2002). Fixed and Random Effects in Stochastic Frontier Models. New York: Stern School of Business. Retrieved from http://people.stern.nyu.edu/wgreene/ fixedandrandomeffects.pdf. Jondrow, J., Lovell, C. A. K, Materov, I. S., & Schmidt, P. (1982). On the Estimation of Technical Inefficiency in the Stochastic Frontier Production Function Model. Journal of Econometrics, 19(1982), 233-238. Juan, C. S., Sperlich, S., Kleinhanss, W., & Murillo, C. (2005). Eƥciency, subsidies and environmental adaptation of animal farming under CAP. Retrieved from http://128.118.178.162/eps/othr/papers/0512/0512015.pdf. Kroupová, Z. and Malý, M. (2010). Analýza nástrojů zemědělské dotační politiky aplikace produkčních funkcí. Politická ekonomie, 6(2010), 774-794.

805

SUBSIDIES AND FARMS’ TECHNICAL EFFICIENCY

Lososová , J., & Zdeněk, R. (2014). Key Factors Affecting the Profitability of Farms in the Czech Republic. Agris on-line Papers in Economics and Informatics, 6(1), 21-36. Meeusen, W., & van den Broeck, J. (1977). Technical efficiency and dimension of the firm: Some results on the use of frontier production functions. Empirical Economics, 2(2), 109-122. Pechrová, M. (2014). Impact of EU Membership on the Development of the Czech Agriculture. In Proceedings of the 2nd International Conference on European Integration, Ostrava, Czech Republic, June 2014 (pp. 545 – 552). VŠB –Technical University of Ostrava. Pechrová, M. (2015). Impact of the Rural Development Programme Subsidies on the farms’ inefficiency and efficiency. Agricultural Economics, 61(5), 197-204. Picazo-Tadeo, A. J., Gómez-Limón, J. A., & Reig-Martínez, E. (2011). Assessing farming eco-efficiency: A Data Envelopment Analysis approach. Journal of Environmental Management, 92(2011), 1154-1164. Pitt, M., & Lee, L-F. (1981). The Measurement and Sources of Technical Inefficiency in the Indonesian weaving Industry. Journal of Development Economics, 1981(9), 4364. Štolbová, M., & Hlavsa, T. (2008). The impact of the LFA payments on the FADN farms in the Czech Republic. Agricultural Economics - Czech, 54 (10), 489–497.

806