Printer Friendly

Allocative and technical efficiency of corporate farms in Russia.

INTRODUCTION

Russian agriculture appears to have recovered from the general collapse of the 1990s. Estimation of the current level of technical efficiency of agricultural producers will reveal the potential for growth through increasing the productivity of resources without increasing their volume. Identification of factors that have a negative influence on efficiency is a prerequisite for the development of productivity improvement programmes for agriculture.

Russian and Western studies of efficiency trends in Russian agriculture in the 1990s produce widely differing results. Some researchers find that efficiency increased during the transition, while others conclude that efficiency decreased. (1) We can only say with certainty that the huge improvements in efficiency originally expected at the beginning of reforms in 1991-92 have not materialised. The objective of this study is to analyse the efficiency of Russian corporate farms (also called agricultural enterprises or farm enterprises) using the data of a survey conducted in 2002-03. The study examines two forms of efficiency: technical efficiency and allocative efficiency. Technical efficiency measures show how efficiently the farm uses the available inputs to produce a given output. In other words, technical efficiency determines whether the farm achieves maximum output using a given bundle of factors of production. Allocative efficiency measures show how far the farm is from the point of maximum profitability given the existing market prices for inputs and products. Thus, allocative efficiency determines whether the factors of production are used in proportions that ensure maximum output at given market prices.

The basic approach to estimating allocative efficiency is through the value of marginal product (VMP). VMP is calculated from econometrically estimated production functions. The use of an agricultural input is allocatively efficient if the value of marginal product is equal to its price. Technical efficiency is usually estimated by one of two approaches: parametric (stochastic frontier analysis (SFA)) or non-parametric (data envelopment analysis (DEA)). Both methods calculate an efficiency index, which measures the distance of the observed firm (corporate farm in our case) from a point on the production frontier. Firms lying on the production frontier are absolutely efficient, and the 'inefficiency' of the remaining firms increases with the distance from the production frontier (for details of the two approaches to technical efficiency see Coelli et al. (1998)).

DATA AND METHODOLOGY

The analysis reported in this article uses the data of 144 corporate farms (generally former kolkhozes and sovkhozes) that participated in the 2003 BASIS survey in three oblasts (Rostov, Ivanovo, and Nizhnii Novgorod). Table 1 presents the basic characteristics of the surveyed corporate farms.

For allocative efficiency analysis, we estimated Cobb-Douglas production functions by the OLS method and then calculated the value of marginal product for each factor of production (VM[P.sub.i]). The value of marginal product was compared with the marginal cost. Given the estimated Cobb-Douglas production function

Y = a[X.sup.b1.sub.1][X.sup.b2.sub.2] ... [X.sup.bk.sub.k]

the marginal product of factor i is calculated as

M[P.sub.i] = [DELTA]Y/[DELTA][X.sub.i] = [b.sub.i]Y/[X.sub.i]

where g is the geometrical mean of the output (the mean of its natural logarithms); [X.sub.i] is the geometrical mean of input i; b, is the OLS estimated coefficient of input i (the elasticity of input i). The marginal product obtained in physical units is then multiplied by the price of the output [P.sub.Y] which gives the value of marginal product of input i: VM[P.sub.i] = M[P.sub.i] x [P.sub.Y]. Allocative efficiency is determined by comparing the value of marginal product of factor i (VM[P.sub.i]) with the marginal factor cost (MF[C.sub.i]). We assume that farms are price takers in the input market, so that the price of factor i ([P.sub.i]) approximates MF[C.sub.i]. If VM[P.sub.i] > [P.sub.i], input i is underused and farm profits can be raised by increasing the use of this input. If, conversely, VM[P.sub.i] < [P.sub.i], the input is overused and to raise farm profits its use should be reduced. The point of allocative efficiency (and maximum profit) is reached when VM[P.sub.i] = [P.sub.i].

The technical efficiency of production of specific commodities was estimated using input-oriented DEA models with variable returns to scale (VRS). Input-oriented estimation is more appropriate than the output-oriented alternative because one of the objectives of the study is to determine the efficiency of input use for the production of a given output and find ways to optimise input use. The impact of external factors on technical efficiency was then estimated by second-stage regression analysis. (2)

The efficiency was estimated for 'total farm' production as well as for crop and livestock production separately. The 'total farm' production function was estimated using the value of output in money units ('gross output' model). Crop and livestock production functions in turn were estimated on two levels of aggregation: using the value of output in money units to estimate 'whole crop' and 'whole livestock' models; and using the data on the physical output of each commodity (in physical units) to estimate specific commodity models (grain, sunflower, beef, milk, and pork). The commodity models should be robust to intentional and unintentional distortions of financial performance measures that affect the value of output. This consideration is particularly relevant for transition countries with a large shadow economy.

ALLOCATIVE EFFICIENCY ANALYSIS

The estimated coefficients of production functions for specific crops and for crop production as a whole are presented in Table 2. The different models produce similar results. All the coefficients are positive, which means that the output increases with the increasing use of each input (keeping the use of all other inputs constant). Most models described 70%-80% of the observed variability.

Land is estimated to be the most significant factor of crop production. Purchased inputs, and in particular fertiliser and seeds, also have a significant positive impact on output. Labour is statistically significant in the aggregated crop production models and not significant in the commodity models. This is probably due to the high correlation between sown land and labour, so that in some models the combined effect of both these factors is reflected in the estimated coefficient of one of them (land).

Crop output is seen to depend on weather conditions. Pre-reform performance is also significant: corporate farms that were profitable (and marginally profitable) in the past are estimated to produce more than the formerly unprofitable farms.

Farm machinery (tractors, grain harvesters, feed combines, trucks) is not statistically significant in any of the crop production models. This result shows that farm machinery (acquired mainly in the pre-reform period) is not a limiting factor in the sample. Machinery, however, can prove to be a limiting factor for the most efficient farms. This indirectly follows from the observation that a substantial number of corporate farms in the sample (although definitely not all the farms) actually purchase machinery.

The estimated coefficients of production functions for specific animal products and for livestock production as a whole are presented in Table 3. As in crop production, the livestock models have a satisfactory explanatory power. Labour and cattle numbers are the most significant factors in livestock production functions, which confirms the extensive character of the Russian livestock sector. The high estimated coefficients of these factors probably incorporate also the effect of fixed assets on output. Fixed assets were not directly included in the models because of low reliability of the available data.

The distance of the farm from the oblast capital does not have a significant effect on livestock output (the same result is observed in crop production). This is probably the outcome of two oppositely directed effects.

On the one hand, the farm's distance from the oblast centre limits its access to the markets for inputs and products, and to some extent reduces the availability of government support. On the other hand, remote farms do not have to compete for skilled labour. The pre-reform status of the farm is not statistically significant in livestock production (contrary to crop production, see Table 2).

The VMP for commodity models and for crop and livestock production as a whole is presented in Table 4. The estimated coefficients of the inputs are taken from Tables 2 and 3. For comparison, Table 4 also shows the value of marginal product for the 'gross output' production function, which aggregates the values of both crop and livestock products.

In the grain and sunflower models, the value of marginal product of land varies from 682 rubles/ha for all grain to 1,738 rubles/ha for grain in Rostov Oblast. These estimates are higher than the value of marginal product of land in crop production as a whole and in the gross output model, where VMP varies from 292 to 446 rubles/ha. The estimates are close to the market price of land, which suggests that land prices adjust to the actual economic productivity of land. This means that, on the whole, land is used efficiently by corporate farms.

The value of marginal product for fertiliser is 1,469 rubles/ton for crop production as a whole and 2,496 rubles/ton for grain. The return to fertiliser in grain production is much higher than the average because grain is one of the most profitable commodities. In the Rostov grain model, however, the value of fertiliser marginal product is lower (1,835 rubles per ton), which probably reflects the effect of decreasing returns to scale (the use of fertiliser is much higher in Rostov Oblast than in the other regions). The weighted average price of fertiliser is 2,619 rubles/ton, higher than the value of marginal product. This indicates that fertiliser is overused given the market prices of inputs and farm products. This is a paradoxical result, because the use of fertilisers has decreased by more than by 50% since the beginning of the 1990s. Perhaps our result does not represent the equilibrium in the domestic fertiliser market, but instead points to overpricing of fertiliser due to the emphasis of Russian fertiliser manufacturers on exports and the price-boosting impact of input subsidy programmes (Serova and Shick, 2005).

The value of marginal product of seeds is 2,239 rubles/ton in the grain model and 3,702 rubles/ton in the total crop production model. This does not mean, however, that the return on seeds for other crop products is higher than for grain. In the grain model, the explanatory variable is the quantity of seeds used in tons, whereas in the total crop production model the explanatory variable is the value of purchased seeds. The cost of purchased seeds is only 18% of the total cost of seeds used in the sample. The mean price of purchased seeds is 3,518 rubles/ton, while the mean cost of seeds from own production is 1,500 rubles/ton. Given the estimates for the value of marginal product, it is inefficient to use purchased seeds at the current market price. The most probable explanation of this phenomenon is the almost total absence of a seed market in Russia. Faced with severe working-capital constraints, farms have sharply curtailed their seed purchases and have shifted to seeds from own production. These are lower-quality seeds, but at least their use does not require 'ready money'. The reduced demand for seeds has led to a sharp contraction of commercial seed production and virtual disappearance of the seed market, which is currently in an inefficient equilibrium characterised by low quantities and high prices.

Estimation of the value of marginal product of labour presents some difficulties. In most commodity models, the coefficient of labour is statistically not significant and the value of marginal product of labour cannot be calculated. In total production models (two crop production models, livestock model, and 'gross output' model), the coefficient of labour is statistically significant and the value of marginal product varies from 19,000 to 26,000 rubles per year. This is close to the actual level of agricultural wages, and we conclude that the use of labour in Russian agriculture today is allocatively efficient (despite the fact that agricultural wages are less than the minimum standard of living and many farm managers estimate that three-quarters of the employed are redundant).

The value of marginal product of purchased inputs estimated from the aggregated models in money units (total crop production, livestock production, and gross output) is between 0.439 and 0.704 million rubles per each million spent on purchased inputs. This again suggests that purchased inputs are overpriced relative to their marginal product, and the use of purchased inputs is inefficient at the current market prices.

TECHNICAL EFFICIENCY ANALYSIS

Table 5 presents the technical efficiency (TE) scores for specific commodities. The mean technical efficiency over all commodities is fairly high (about 0.80), and for more than 50% of corporate farms it is greater than 0.70. These farms are close to the efficiency frontier, where technical efficiency reaches its maximum value 1.

The high mean TE scores and the high frequency of 'best practice' technologies in the sample limit the potential impact that can be expected from the adoption of 'best practice' technologies by the inefficient farms. Thus, the adoption of 'best practice' technologies will increase the production of beef by 22%-36% (Models 2 and 1, respectively). Production gains are obviously desirable; however, the magnitude of the potential gains will not be sufficient for Russian agriculture to close the productivity gap observed between Russia and developed market economies (Trueblood and Arnade, 2001).

Figure 1 shows the distribution of corporate farms by TE scores (averaged over all crop and livestock models). The distribution is clearly bimodal. The main mode includes farms that form the efficiency frontier (technical efficiency > 0.9). This mode contains 49% of farms in the sample (averaged over all commodity models). The second mode includes less efficient producers. Their technical efficiency is 0.4-0.6 (23% of farms in the sample). The results of Model 1 for livestock products (in these models, animal feed is expressed in feed units) shed some light on the differences between the farms in the two modes. In corporate farms forming the efficiency frontier (the main mode, TE > 0.9), the share of purchased feed is relatively high. it is 19% for the beef model, 14% for the milk model, and 10% for the pork model. For farms in the second (inefficient) mode, the respective purchased feed shares are 2.6%, 2.0%, and 3.4%. The mean share of purchased feed for all corporate farms in the sample is 7%. Purchased feed is mainly high-quality concentrated feed, whereas feed from own production is basically hay, pasture grasses, or low-quality concentrated feed. The farms at the efficiency frontier, using a high proportion of purchased feed in the ration, probably optimise the ration by analysing the market price of feed, the cost of on-farm production, and the value of the end product. The inefficient farms, using a small proportion of purchased feed in their ration, probably follow the strategy of cost minimisation for purchased inputs because of financial constraints.

[FIGURE 1 OMITTED]

Once the TE scores have been determined by DEA, it may be useful to examine their relationship with additional factors, such as farm size, location, or financial conditions. Such a relationship is usually determined by regressing the TE scores on the relevant external factors. Table 6 shows the estimation results for a 'second-stage' regression model in which the dependent variable is the TE score and the explanatory variables are various external factors.

Many factors that a priori were expected to affect the technical efficiency proved to be not statistically significant and are not shown in Table 6. Thus, the farm size (measured by hectares of land used) does not have a statistically significant effect on technical efficiency. State subsidies and borrowing of any kind do not affect technical efficiency (however, very few farms in the sample provided information on these variables). The managerial qualification variable expressing knowledge of tax laws, lease payments, and land allocation procedures is not statistically significant in the model. The land utilization ratio (ie, the share of land actually used in agricultural production) does not affect technical efficiency. It is quite possible that keeping agricultural land in the farm's possession without cultivating it is the most efficient strategy under the present circumstances because of complex alienation procedures and high transaction costs.

On the other hand, most models show a positive association between enlargement of holdings and technical efficiency. Farm enlargement (a yes/no variable that indicates if the farm has added new agricultural land to its holdings) is statistically significant in both crop and livestock production models. Farm enlargement not only represents expansion of the sown area, but it is also an indicator of management quality.

As suggested by a priori considerations, wage arrears have a negative effect on technical efficiency (in half the commodity models). If this factor is accepted as a proxy for financial health, we conclude that financially ailing farms are less efficient than the rest. This factor is also closely connected with management quality. The absence of wage arrears and good financial health reflect highly qualified management.

The impact of management structure is also reflected by the farms' pre-reform status. Farms that were profitable in the pre-reform period demonstrate higher technical efficiency in some models. The existence of a controlling packet of shares, which a priori led us to expect greater efficiency due to more effective control of the majority owner over management, is not statistically significant in most models, perhaps because of the very small number of farms with a controlling packet in the sample.

Specialisation in crop production does not improve the efficiency of crop enterprises, while significantly reducing the efficiency of subsidiary livestock production. The regional factor has a statistically significant effect: farms in Ivanovo Oblast are the least efficient and Rostov farms are the most efficient in the sample. Distance from the oblast capital in general does not have a statistically significant effect on technical efficiency. The existence of surplus labour does not affect the technical efficiency of corporate farms either.

CONCLUSION

Land is unquestionably the cornerstone of agricultural production. Its impact on output is positive and highly significant. However, technical efficiency of corporate farms is not associated with land endowment. The share of idle land is substantial and the price of laud is close to the value of marginal product. In order to intensify laud use and prevent degradation of land quality (which is typical for unprofitable farms), it is essential to encourage redistribution of land from least efficient to most efficient users. The government should actively move to simplify the procedures for buying, selling, and leasing of land so as to reduce the extremely high transaction costs in the land market.

Efficient production requires purchased inputs. Their positive impact on output is proved by production functions. Nevertheless, the actual use of purchased inputs in Russia is substantially lower than in developed market economies. The analysis of allocative efficiency for purchased inputs produces a mixed picture. However, this analysis is based on actual market prices for products and inputs. Underpricing of farm products combined with overpriced inputs may lead to allocatively efficient use of purchased inputs, but at a very low level of quantities. In the medium- and long-term perspective, this may cause degradation of land quality and a range of other negative phenomena.

Price distortions in input markets may be caused by various reasons. Thus, in fertiliser and fuel markets, pervasive monopolisation (especially the existence of local monopolies) combined with high export demand strengthens the bargaining power of input manufacturers and leads to overpricing. To correct this distortion, it is necessary to encourage deliveries to the domestic market, while carefully avoiding direct bans that have already demonstrated their absolute inefficiency. Input subsidies do not correct price distortions, and sometimes even aggravate them. The solution of the problem probably requires development of appropriate market institutions that will increase the bargaining power of agricultural producers and at the same time lower the suppliers' risk in dealing with agricultural producers. The most obvious approach is through the creation of farmers' input-supply cooperatives capable of pooling the flows, reducing the risks, and possibly developing short-term credit schemes for input purchases.

The low productivity of Russian agriculture is mainly attributable to management factors, and not to technological or allocative factors. This conclusion is supported by the analysis of external factors that affect technical efficiency. Corporate farms with a proactive management policy (including, for instance, acquisition of additional land) demonstrate higher technical efficiency than others.

Finally, simple extension of 'best practice' production will not affect the relatively large contingent of best performers and thus will not eliminate the large productivity gap between Russia and the developed market economies. Growth requires shifting the production frontier outward, which in turn requires the government to develop intelligent long-term policies for agriculture.
Table 1: Basic characteristics of surveyed corporate farms

 Unit of Number of Mean
 measurement respondents

Agricultural land Hectares 144 4,093
Land used in agriculture Hectares 141 3,351
Total number of workers People 137 122
Fertiliser (used) Tons 106 99
Gasoline (purchased) Tons 141 65
Diesel fuel (purchased) Tons 142 173
Tractors Pieces 130 20
Grain harvesters Pieces 132 5
Trucks Pieces 137 11
Equity Million rubles 131 5
Debt Million rubles 90 59

 Standard Min Max
 deviation

Agricultural land 2,624 4 14,242
Land used in agriculture 2,504 0 14,242
Total number of workers 83 10 408
Fertiliser (used) 142 0 811
Gasoline (purchased) 70 2 636
Diesel fuel (purchased) 165 5 1,015
Tractors 12 0 68
Grain harvesters 4 0 21
Trucks 8 0 34
Equity 6 0 46
Debt 522 0 4,959

Table 2: Crop production functions

 Models in physical units

 Grain
 Grain (Rostov (b)) Sunflower

Constant 1.848 0.920 --
Land 0.415 0.861 1.161
Labour --
Purchased inputs (total value) X X X
Fertiliser 0.101 0.060 --
Seeds 0.328 -- --
Weather 0.370 0.482 --
Pre-reform status
 Marginally profitable 0.319 0.601 0.587
 Profitable 0.283 0.473 0.505
 Unprofitable 0 0 0
Rostov regional factor -- X
[R.sub.adj.sup.2] 0.73 0.64 0.80
Number of observations 103 46 45

 Models in value units (a)

 Model 1 Model 2

Constant 1.062 1.510
Land 0.207 0.178
Labour 0.314 0.257
Purchased inputs (total value) 0.516 X
Fertiliser X 0.076
Seeds X 0.566
Weather X X
Pre-reform status
 Marginally profitable X X
 Profitable X X
 Unprofitable X X
Rostov regional factor 0.483 0.688
[R.sub.adj.sup.2] 0.70 0.70
Number of observations 114 107

Table shows only factors that are statistically significant at 10%
level. X=not included in the model; --=not statistically significant.

(a) In Model 1, all purchased inputs are aggregated by value; in Model
2, fertiliser and seeds are included separately (in value units).
Labour defined as the average annual number of workers engaged in crop
production. Output defined as the value of crop production.

(b) Estimated only for Rostov Oblast.

Table 3: Livestock production functions (a)

 Beef Milk

 Model 1 Model 2 Model 1 Model 2

Constant -1.731 -1.840 1.575 --
Labour -- -- 0.366 0.267
Number of cattle 0.763 0.680 0.705 0.801
Purchased inputs X 0.054 (c) X --
Weather 0.397 -- 0.401 --
Ivanovo regional factor 0.454 -- -- --
Feed quality (d) -0.673 -0.731 -- --
[R.sub.adj.sup.2] 0.851 0.815 0.909 0.938
Number of observations 77 75 72 50

 Pork Livestock (b)

 Model 1 Model 2

Constant -- -4.867 1.319
Labour 0.791 0.507 0.280
Number of cattle 0.613 0.844 0.303
Purchased inputs X -- 0.481
Weather -- -- X
Ivanovo regional factor -- X
Feed quality (d) -- -- X
[R.sub.adj.sup.2] 0.811 0.752 0.760
Number of observations 45 47 98

Table shows only factors that are statistically significant at 10%
level.

X=not included in the model; --=not statistically significant.

(a) In Model 1, feed is expressed in feed units from the survey; in
Model 2, feed is expressed in terms of inputs used for its production
(labour, fertilisers, and agricultural machinery). Feed is not
statistically significant in any of the models. Output defined as
production of beef, milk, or pork in physical units.

(b) Output defined as the value of livestock production.

(c) Fertiliser only.

(d) Share of concentrated feed in ration.

Table 4: Comparison of values of marginal product with factor prices

Factor Units Model Coefficient

Land Hectares Grain 0.414
 Grain (Rostov) 0.860
 Sunflower 0.410
 Crop production M1 0.207
 Crop production M2 0.178
 Gross output 0.401
Fertiliser Tons (active Grain 0.100
 substance) Grain (Rostov) 0.060
 Rubles Crop production M2 0.076
Seeds Tons Grain 0.330
 Rubles Crop production M2 0.566
Labour Man-years Crop production M1 0.314
 Crop production M2 0.257
 Livestock production 0.280
 Gross output 0.410
Livestock Head Milk 0.705
 Beef and veal 0.760
 Gross output 0.303
Purchased inputs Million Crop production M1 0.516
 rubles Livestock production 0.481
 Gross output 0.163

Factor VMP Factor price
 (rubles) (rubles)

Land 682 500-1500 (a)
 1,738 1,500-2,000 (a)
 1,319 1,500-2,000 (a)
 332 500-1,500 (a)
 292 500-1,500 (a)
 446 500-1,500 (a)
Fertiliser 2,496 (b) 2,619 (c)
 1,835 (b) 2,619 (c)
 1,469 2,619 (c)
Seeds 2,239 1,500-3,518 (d)
 3,702 NA
Labour 22,120 12,000-18,000 (a)
 19,270 12,000-18,000 (a)
 25,628 12,000-18,000 (a)
 25,700 12,000-18,000 (a)
Livestock 4,297 NA
 23,691 NA
 2,037 NA
Purchased inputs 0.704 1
 0.439 1
 0.546 1

(a) Expert estimate.

(b) Calculated per ton of fertiliser based on the average active
substance content of one ton of purchased fertilisers.

(c) Weighted average price per ton for fertiliser purchases in the
survey.

(d) Lower limit is the cast of seeds from own production; upper limit
is the average market price.

NA: not applicable.

Table 5: TE of corporate farms for selected commodities (estimated by
the DEA method)

 TE (mean) St. dev. TE > 0.9

 % of farms Number of farms

Grain 0.77 0.23 39 44
Sunflower 0.71 0.24 35 17
Beef 1 (a) 0.64 0.29 31 22
Beef 2 (a) 0.78 0.26 54 39
Milk 1 (a) 0.65 0.24 25 19
Milk 2 (a) 0.82 0.22 55 41
Beef+milk 1 (b) 0.78 0.20 42 33
Beef+milk 2 (b) 0.88 0.16 65 51
Pork 1 (a) 0.75 0.28 50 24
Pork 2 (a) 0.88 0.18 67 32

 TE > 0.7, TE > 0.3,
 % of farms % of farms

Grain 64 2
Sunflower 49 2
Beef 1 (a) 44 17
Beef 2 (a) 64 6
Milk 1 (a) 37 2
Milk 2 (a) 69 1
Beef+milk 1 (b) 59 0
Beef+milk 2 (b) 80 0
Pork 1 (a) 63 6
Pork 2 (a) 79 0

(a) See notes on Models 1 and 2 in Table 3.

(b) Two-output DEA models (beef and milk). Without fattening
operations, the inputs are shared by the two outputs.

Table 6: Factors that influence technical efficiency (a)

Factor Grain Sunflower Beef 1 Beef 2

Ivanovo regional factor -0.134 -0.337 -0.296
Distance from oblast capital 0.243
Augmentation of farm holdings 0.122 0.321
Wage arrears -0.017
Pre-reform status (profitable) 0.298
Crop specialisation -0.496 -0.613
Controlling packet of shares -0.175
Surplus labour

Factor Milk 1 Milk 2 Beef+milk 1

Ivanovo regional factor -0.279 -0.294
Distance from oblast capital 0.134
Augmentation of farm holdings 0.281 0.298
Wage arrears -0.022 -0.017 -0.018
Pre-reform status (profitable) 0.214 0.185
Crop specialisation -0.777 -0.613 -0.533
Controlling packet of shares
Surplus labour

Factor Beef+milk 2 Pork 1 Pork 2

Ivanovo regional factor -0.492 -0.284
Distance from oblast capital
Augmentation of farm holdings 0.167 0.267
Wage arrears -0.025
Pre-reform status (profitable) 0.171
Crop specialisation -0.343 -1.179 -0.485
Controlling packet of shares
Surplus labour 0.002

(a) Table shows only the factors that are statistically significant at
10% level in one of the commodity models.


(1) The first view is advanced in Lerman et al. (2003) and Voigt and Uvarovsky (2001); the second view in Osborne and Trueblood (2002) and Sedik et al. (1999).

(2) The DEA programme used in this study has been developed by Aleksandr Usol'tsev on the basis of standard linear programming algorithms published in the literature. The work has been carried out at the Analytical Centre for Agri-Food Economics in Moscow as part of the BASIS Russia project.

REFERENCES

Coelli, T, Rao, DSP and Battese, G. 1998: An introduction to efficiency and productivity analysis. Kluwer: Boston.

Lerman, Z, Kislev, Y, Kriss, A and Biton, D. 2003: Agricultural output and productivity in the former Soviet republics. Economic Development and Cultural Change 51(4): 999-1018.

Osborne, S and Trueblood, M. 2002: Agricultural productivity and efficiency in Russia and Ukraine: Building on a decade of reform. Agricultural Economic Report 813, ERS/USDA: Washington, DC.

Sedik, D, Trueblood, M and Arnade, C. 1999: Corporate farm performance in Russia, 1991-95: an efficiency analysis. Journal off Comparative Economics 27: 514-533.

Serova, E and Shick, O. 2005: Markets for purchased farm inputs in Russia. Comparative Economic Studies 47(1): 154-166.

Trueblood, M and Arnade, C. 2001: Crop yield convergence: How Russia's yield performance has compared to global yield leaders. Comparative Economic Studies 43(2): 59-81.

Voigt, P and Uvarovsky, V. 2001: Developments in productivity and efficiency in Russia's agriculture: the transition period. Quarterly Journal of International Agriculture 40(1): 45-66.

MARGARITA GRAZHDANINOVA (1) & ZVI LERMAN (2)

(1) AFE--Analytical Centre for Agri-Food Economics, Moscow, Russia. E-mail: margo@iet.ru;

(2) Department of Agricultural Economics and Management, The Hebrew University of Jerusalem, Israel E-mail: lerman@agri.huji.ac.il
COPYRIGHT 2005 Association for Comparative Economic Studies
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2005 Gale, Cengage Learning. All rights reserved.

 Reader Opinion

Title:

Comment:



 

Article Details
Printer friendly Cite/link Email Feedback
Author:Grazhdaninova, Margarita; Lerman, Zvi
Publication:Comparative Economic Studies
Article Type:Industry Overview
Geographic Code:4EXRU
Date:Mar 1, 2005
Words:5068
Previous Article:Financial performance and efficiency of corporate farms in northwest Russia.
Next Article:The allocative efficiency of material input use in Russian agriculture.
Topics:


Related Articles
A parametric approach to efficiency measurement using a flexible profit function.
Symposium on Russian agriculture factor market constraints on economic growth in Russian agriculture--Golitsino papers: an introduction to special...
Large and small business in Russian agriculture: adaptation to market.
Development of peasant farms in Central Russia.
Russia's new agricultural operators: their emergence, growth and impact.
Agricultural employment in Russia 1990-2003.
The allocative efficiency of material input use in Russian agriculture.
First-order economizing: organizational adaptation and the elimination of waste in the U.S. pharmaceutical industry *.
Individual farming as a labour sink: evidence from Poland and Russia.
Cost efficiency, technological progress and productivity growth of banks in GCC countries.

Terms of use | Copyright © 2014 Farlex, Inc. | Feedback | For webmasters