Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1598-7248 (Print)
ISSN : 2234-6473 (Online)
Industrial Engineering & Management Systems Vol.19 No.4 pp.857-865

Econometric Models of Cattle-Breeding Production Cost

Aigul Hazieva*, Nuria Rafikova, Gamir Habirov, Zariya Zalilova, Elza Sagadeeva
Federal State Budgetary Educational Establishment of Higher Education “Bashkir State Agrarian University”, Ufa, Russian Federation
*Corresponding Author, E-mail:
September 20, 2020 October 9, 2020 October 15, 2020


The paper presents the results of multiple correlations and regression modeling of the cattle weight gain production cost. Based on actual data of farms, a comparative assessment of developed models was carried out for the first time with the use of panel data models and so called ‘farms-years’ method. The most significant factors affecting the change in the production cost of cattle weight gain were identified both for individual stages and for the entire period of the study in agricultural organizations of the Republic of Bashkortostan. Specific effects that reflect the characteristics of individual farms, and temporary effects, showing the influence of internal and external factors, state programs for the development of the industry, were determined. The introduction of the proposed analysis methodology contributes to further increase in live-stock production. At the same time, it helps to reduce the cost of production and improve production efficiency.



    Improving the efficiency of production costs control, analyzing the cost of production and searching for reserves to reduce it is an important and relevant direction for ensuring the effective use of enterprise’s production and resource potential and improving the competitiveness of domestic products (Shim et al., 1996; Ableeva et al., 2019;Zalilova et al., 2020).

    In modern conditions, much attention is paid to a theoretical justification of the terms ‘production costs’, ‘expenses of production’, to their identity and differences. Enterprises and firms more often use ‘production cost’ category instead of the term ‘production expenses’. The subject of the research is the economic aspects of using these terms and the method of econometric analysis and modeling of agricultural production cost (Rafikova et al,, 1999).

    Implementation of management accounting and cost budgeting primarily depends on how carefully the basic principles of accounting, cost analysis and product cost modeling adopted in domestic practice are maintained (Dumnov et al., 2019;Zakirova et al., 2019;Hung, 2020).

    ‘Direct-costing’ method is considered in scientific publications and in practice. This method is used to study costs depending on the amount of output production with the division of costs into semi-fixed and semi-variable (Shim et al., 1996; Mueller et al., 2003; Oblad, 2019;Kovanov and Iakovleva, 2016).

    In agriculture, the cost of production is highly dependent on natural and climatic factors and resource prices. The proportionality of costs to the quantity of products produced is not always maintained. Different cost factors require their optimal combination and detailed analysis to determine the conditions for figuring out costs and finding ways to reduce them. In market economy conditions the influence of microlevel factors that directly depend on the level of work of individual manufacturers increases (Rafikova et al., 2019). However, in market economy, there is not enough open access information on agricultural organizations. This makes it difficult to analyze and forecast costs. Foreign authors, in particular authors from the Czech Republic, also write about this (Hloušková et al., 2018). Forecasting methods based on time series analysis are mainly used.

    Thus, three approaches to cost forecasting were compared using the macroeconomic indicators of CZCO data (Czech Statistical Office) and FADN CZ microeconomic data. In the first approach a price index for agricultural products is used. The disadvantage of this approach is the late availability of data. The data is published quarterly with a delay of one and a half month. This means that information on the current year's index may be available in the middle of February of the next year. The second approach takes into account the time series of cost items (data from the FADN database in CR). Panel data has been used in the developed process since 2001. The main advantage of using panel data is the reduction of the impact of changes in farms within the sample on forecast results. The third approach is based on the time series of cost items in current prices according to EAA data. According to their opinion, the second approach, based on time series analysis of FADN data, is the most appropriate. The advantage of this approach is the availability of data (Hloušková et al., 2018;Riyadi, 2020).

    Econometric modeling of production costs remains poorly studied in Economics. The authors of this study have many years of experience in research on economic and statistical analysis of the cost of agricultural products (Rafikova et al., 2019).

    In 2019, the main authors of this study presented the results of economic and statistical analysis of the influence of zonal conditions on the production and cost of production of growing and fattening of cattle without taking into account dairy and meat cows. Typological groupings of agricultural organizations in the Republic of Bashkortostan were used as analysis tools. They allowed obtaining more reliable estimates compared to the grouping for a single year. A significant difference between zones and between districts in the context of individual zones was determined. Three groups of areas were determined according to the cost level, productivity and cost of production. Cost structure features, productivity level and wages in cattle breeding were determined. Multiple correlational and regressive models of livestock productivity and production costs for all regions and zones were constructed, which allowed establishing more significant factors (Rafikova et al., 2019).

    The use of econometric calculations should be based on knowledge and understanding of the essence of economic processes and occurrences, features of economic relationships and patterns of their development (Akhmetova et al., 2018).

    Many articles describe the processes of developing a set of mathematical models for evaluating agricultural production technologies and managing technological processes in crop production and animal breeding (Gianelle et al., 2018;Skevas et al., 2018;Tkachev et al., 2018).

    In this study, we consider the method of analysis and construction of models of the cost of production of growing and fattening of dairy and meat cattle, which reflect the features of the growing technology at the level of individual agricultural organizations.

    The purpose of this study is econometric modeling of the livestock gain cost with the use of the ‘farms-years’ methods and panel data models to identify significant factors.


    The research was carried out using statistical, econometric, calculative and constructive methods and evaluation of the activities of agricultural producers based on the obtained models. At the first stage of the study, a statistically significant regression model of the production cost of cattle growth was obtained using the ‘economyyears’ method and using panel data for 2009-2013. In the future, the study was continued according to data for 2014-2017, as well as in general for 2009-2017.

    To build a cost model for 1 dt of increment the following factors are selected:

    • x1 is the reverse productivity indicator - the number of heads of cattle required for the production of 10 kg of weight gain, per head. As a productivity factor of cattle, its reverse indicator was taken. This indicator represents the number of heads of cattle needed to produce 10 kg of weight gain;

    • x2 is the feed consumption per 1 head, dt feed units;

    • x3 is the cost of 1 dt of feed units, RUB.

    Let’s consider the obtained statistically significant regression model of the cattle weight gain cost using the ‘farms-years’ method for 2009-2013:

    A developed model (1) gives evidence of the statistical significance of both the model and the x1, x2, and x3 factors.

    By comparing actual values of the weight gain cost with the values calculated according to the model (1), we determined a group of farms, where the actual weight gain cost was lower compared to the calculated value. Using the date from these farms, the following regression model of the livestock weight gain cost was developed. To do this the ‘farm-years’ method was applied:

    According to the model parameters (2), the significance level of such factors as x3 and x1 is very high. At the same time, there is a decrease in the significance of x2 factor in comparison with the model parameters (1). To compare the factors by their effect on the weight gain cost, β-coefficients were calculated, and the reliability of the factors for materiality was checked according to the Student’s t-test. When ranking factors by their effect on the cost of livestock weight gain, it should be noted that the cost of feed units (β3 = 0.966), feed consumption per head (β2 = 0.587), and the inverse indicator of cattle productivity (β1 = 0.507) had the greatest impact. Thus, the given model proves that the cost of feed units is determinant of the three factors which are considered.

    At the second stage, a statistically significant regression model of the production cost of livestock weight gain for 2014-2017 was also developed using the ‘farms-years’ method:

    A comparison of models (1) and (3) shows that the strength of relationship and significance of the model for 2014-2017 dramatically increased compared to 2009- 2013. The significance of the coefficients of the regression equation for all three factors according to the Student’s t-test has also increased.

    At the third stage, a statistically significant regression model of the production cost of cattle weight gain for 2009-2017 was developed using the ‘farms-years’ method:

    3. RESULTS

    The model developed over the entire study period was even more significant both in general and for individual factors, especially for -x3 and x2.

    Thus, the use of this tool made it possible to obtain adequate models. They can be used to evaluate the performance of individual farms not only among themselves, but also in the dynamics.

    The panel data method was used to continue analysis. The method has certain advantages:

    • - first of all, a larger number of observations and greater efficiency in estimating the parameters of the econometric model;

    • - second of all, panel data allows to build more flexible and meaningful models, as well as enables to identify effects and get answers to questions that are not available only within the spatial data;

    • - third of all, it becomes possible to take into account and analyze individual differences between economic units, which cannot be done in the framework of standard regression models.

    During the implementation of this stage, the following tasks were set and solved:

    • 1) to identify differences between the time periods under consideration;

    • 2) to assess the development of individual farms and their groups.

    Initially, a visual data analysis was carried out. Time series graphs of farms for each variable were constructed. Let us consider the change in the productivity factor of livestock (x1), (Figure 1). The inverse indicator of productivity for 2009-2013 has a clear tendency to decrease. After 2010 there was a trend of increasing productivity, which contributed to food security.

    As it can be seen from Figure 2, the feed consumption per head factor (x2) is in general distributed more evenly. Variations of this factor are observed in three farms. In other farms the feed consumption per head varies from 15 to 30 dt to a single unit (Figure 2).

    It can be seen from the data in Figure 3 that during the study period the cost of 1dt of feed units in Chekmagush district fluctuates sharply, which is caused by low yields and high cost of feed crops due to the drought in 2010.

    In 2010-2011, there was a sharp cost decline in three farms. Moreover, there was also a significant decrease in the cost of 1 dt of feed units in three other organizations. There has been a more moderate period since 2012. This is due to an increase in the support from the government in recent years.

    When developing panel data models, various factors were included. Types of models were chosen both by including or excluding fixed temporary and individual effects. Coefficients were estimated using the Eviews package (Molchanov et al., 2016). Coefficients of the regression equation that were obtained are significant both in general and for individual factors, and do not contradict their economic content.

    According to the results of panel data modeling, the model with fixed individual and temporary effects was recognized the best (Molchanov et al., 2016;Hung, 2020). More significant parameters’ estimates of the included factors x1, x2, and x3, as well as constants were obtained:

    The value of the determination coefficient is the evidence of the model good quality (5). Analysis of the obtained coefficient estimates for the explicative variables x1, x2 and x3 allows recognizing them as significant explicative variables.

    Having analyzed the obtained values, we identified significant factors, which, as in the model (1), were the cost of 1 dt of feed units (x3) and feed consumption per 1 head (x2). According to the Student's t-test, the significance of x1 factor has decreased, but it is still important. At the same time, the signs of the estimated coefficients fully correspond to the signs of the cost model of livestock weight gain, developed with the use of the ‘farms-years’ method.

    According to the obtained model (5), the maximum estimated cost of livestock weight gain was observed in 2010, amounting to 13,096 rubles. The actual cost was 12,210 rubles. This indicates the effective farm operation. The minimum cost of 4,370 rubles was revealed in 2009, although it was actually 1,111 rubles higher (Figure 4).

    Having evaluated the panel data model, fixed individual effects that influence the cost of cattle weight gain and are individual for each farm were obtained. They include unrecorded or unobserved factors that reflect individual characteristics of farms (for example, quality of management, service life, quality of equipment, etc.) (Molchanov et al., 2016). Analysis of the obtained values of fixed individual effects indicates that positive individual effects are observed in six farms. This indicates an increase in the cost of weight gain of cattle. According to the results of the correlation and regression analysis of economic activity, two farms in Chekmagush district were included in the group of effectively working farms (Table 1).

    The analysis of temporary effects shows that in 2009-2013 the cost of livestock weight gain in all farms of Chekmagush district was affected by such unaccounted factors as state programs, economic crisis, natural and climatic phenomena etc. As a result, negative assessments of temporary effects were observed in 2009-2011, and positive estimates were observed in 2012-2013 (Table 2). A negative assessment of temporary effects may point to an increase in the efficiency of government programs and measures, which are aimed at agricultural activities development.

    Besides, a graph of the forecast pointwise values for the model with fixed individual and time effects was constructed (Figure 5). The confidence interval of the forecast in two standard deviations is marked by a dotted line on this graph.

    It should be noted that the panel data model confirmed previously obtained comparative results on cost reduction of individual farms performance.

    It should be noted that it is almost impossible to measure the influence of certain factors in 10-45% of multiple models. Moreover, these factors cannot be included to the model. Panel data allows to partially taking this dissimilarity into account, since individual effects reflect the influence of all (observed or unobserved) variables (Molchanov et al., 2016).

    At the next stage of the analysis, after checking necessary conditions, we will consider and evaluate how the coefficients of the cost model of livestock weight gain change with the change in random effect:

    Model (6) includes factors that explain 56.3% of the change in the cost of livestock weight gain. The differences between the obtained coefficients of fixed and random effects are insignificant. It is important to pay attention to the parameters of the estimated model, which in their turn are very close to the parameters of the model obtained using the ‘farms-years’ method.

    Hausman test used to check this model showed that specification of the resulting model with random effects cannot be applied (Molchanov et al., 2016).

    At the end of the study, a panel analysis was made based on data from agricultural organizations of Chekmagush district for 2014-2017. As a result, an adequate model with fixed individual and temporary effects was obtained:

    In model (7) significance of factor x1 decreased (ttabl. = 2.02 at a significance level of 0.05) and the influence of factor x3. Increased.

    At the end, an adequate model with fixed individual and temporary effects for 2009-2017 was obtained:

    The obtained model (8) confirms model results (5) and (7). For the study period from 2009 to 2017 x3 and x2 were more significant factors when using the ‘farms-years’ method. A similar result was also obtained when using the panel data method.

    Figure 6 shows the change in the cost of cattle weight gain in different agricultural enterprises for the analyzed period from 2009 to 2017.

    The advantage of the panel data method is that it can be used to study the trend of changes in the cost of livestock gain, as well as factors that affect its decline in particular organizations. The method also helps identify the potential of each factor.


    The paper presents the changes in resource potential of livestock industries according to the data of agricultural censuses of 2006 and 2016. Differences between regions of Russia in terms of conditions and efficiency of livestock production are identified based on the typification of regions, which is done by grouping them based on a multidimensional analysis. The practical significance of the research lies in the possibility for regional and federal agricultural bodies to apply this methodology (Kagirova et al., 2018).

    Klychova et al. (2014a, 2014b) consider special aspects of calculating the cost of horse breeding products and offer a scientific and methodological approach to its optimization. Creating the database of management of costs for livestock development, as well as for horse breeding, is a complex process that includes planning, assessment, accounting, analysis, effective control and management of costs, forecasting and making appropriate decisions to prevent negative results. In our study, analysis, modeling and forecasting of production costs were the emphasis when making management decision (Klychova et al., 2014a, 2014b).

    Data selected by random sampling and questionnaires are used in the following work to apply multiple regression (Ugwumba et al., 2010). The ‘farms-years’ and the panel data methods have been applied for the first time in our research to study agricultural organizations for each year during the following periods: 2009-2013, 2014- 2017 and 2009-2017. This method allowed to maintain the principles of regression modeling and made it possible to replenish the block of data, increase the reliability of the study and establish significant differences in the main indicators that characterize the production of cattle breeding and feeding in agricultural enterprises. In addition, when developing analysis tools, authors took into account the complex effect of differences that characterize the level of activity of agricultural enterprises on the production cost reduction.

    In works of Mohammadi et al. (2010) and Banaeian et al. (2011) costs are studied as energy units, but not in monetary terms. This is due to the fact that production of cattle weight gain is less energy-intensive compared to glasshouse production of fruits and vegetables, which includes the costs for fertilizers, diesel fuel, water for irrigation and labor.

    In the works of Banaeian et al. (2011), the Cobb- Douglas production function was used to determine the distribution of energy resources. This function reflects the multiplicative dependence of the volume of output on capital and labor costs. In our study, we used multiple regression models of product cost using the ‘the farmyears’ method and panel data obtained over a long period. They allowed us to determine not only the elasticity coefficients, but also the coefficients of the multiple regression equation, individual and temporal effects.

    The more significant factors in the obtained models of cost of 1 dt of cattle weight gain were: the cost of 1 dt of feed units, feed consumption per head, livestock productivity.

    4.1 Practical Application and Recommendations

    As a result, based on the identified significant factors of the model (1), which we constructed using the ‘farmyears’ method, we calculated forecasts of the cost of cattle weight gain in three variants:

    • - when using the average values of the factors of farms the actual cost values of which are lower than the calculated ones, the forecast value of the production cost will be 7060 rubles. This is 16.3 % less than the forecast obtained by using the model (1).

    • - when using the average values of the included factors for the analyzed agricultural organizations in the resulting model, the projected cost of 1 dt of weight gain will be 8,438 rubles;

    • - when the productivity is low, the values of the cost of 1 dt of feed unit and feed consumption per head of cattle are high, the projected cost of the cattle weight gain will be 9,818 rubles, which is 16.4% higher than the forecast cost obtained using the model (1).

    Based on the coefficients of models with fixed individual and temporary effects, it can be concluded that if the number of heads of cattle required for the production of 10 kg of weight gain per 1 head increases, the cost of weight gain of cattle will increase in the range of 1268- 1740 rubles. With an increase in feed consumption per 1 head by 1 dt of feed units, the cost of livestock weight gain will increase by 133-275 rubles. If the cost of 1 dt of feed units increases by 1 RUB, the cost of weight gain will increase by 8-18.5 RUB.


    Based on the analysis, we can draw the following conclusions:

    • 1. The models developed using the ‘farms-years’ and the panel data methods were significant in general for individual stages and for the entire period of the study.

    • 2. Comparative analysis of models of the cost of weight gain of cattle showed that the closest relationship is achieved in the model with fixed individual and temporary effects.

    • 3. The results of the analysis show that the factors effecting the change in the cost of weight gain of cattle in the farms of Chekmagush district range by their strength according to β-coefficients: the cost of 1 dt of feed units (x3), feed consumption per head (x2), livestock productivity (x1).

    • 4. Using models based on panel data, more accurate estimates of the analyzed factors and the model as a whole were obtained. Individual effects, reflecting characteristics of farms and a temporary effect that is associated with the influence of external factors were established. All the above should be taken into account when evaluating state programs aimed at improving the efficiency of economic activity.



    Change in the inverse cattle breeding productivity indicator for 2009-2013.


    Change in feed consumption per head.


    Change in the cost of 1 dt of feed units for 2009- 2013.


    Actual, calculated values and cost balances of livestock weight gain.


    Statistical forecast of the cost of livestock weight gain.


    Change in the cost of livestock weight gain in farms of Chekmagush district for 2009-2017.


    Individual effects

    Temporary effects


    1. Ableeva, A. M. , Salimova, G. A. , Rafikova, N. T. , Fazrahmanov, I. I. , Zalilova, Z. A. , Lubova, T. N. , and Hazieva, A. M. (2019), Economic evaluation of the efficiency of supply chain management in agricultural production based on multidimensional research methods, International Journal of Supply Chain Management, 8(1), 328-338.
    2. Akhmetova, D. T. , Utibayeva, G. B. , Utibayev, B. S. , Zhunusova, R. M. , Baidakov, A. , and Tukenova, B. I. (2018), 6 forecasting production performance, European Research Studies Journal, 21(2), 134-146.
    3. Banaeian, N. , Omid, M. , and Ahmadi, H. (2011), Energy and economic analysis of greenhouse strawberry production in Tehran province of Iran, Energy Conversion and Management, 52(2), 1020-1025.
    4. Dumnov, A. D. , Khoruzhi, L. I. , Kharitonova, A. E. E. , Ukolova, A. V. , and Skachkova, S. A. (2019), Eco-economic systems of russian agriculture: Statistical analysis, Journal of Reviews on Global Economics, 8, 362-372.
    5. Gianelle, D. , Romanzin, A. , Clementel, F. , Vescovo, L. , and Bovolenta, S. (2018), Feeding management of dairy cattle affect grassland dynamics in an alpine pasture, International journal of agricultural sustainability, 16(1), 64-73.
    6. Hloušková, Z. , Ženíšková, P. , and Prášilová, M. (2018), Comparison of agricultural costs prediction approaches, AGRIS on-line Papers in Economics and Informatics, 10(1), 3-13.
    7. Hung, T. T. (2020), Preventing human trafficking in Vietnam through economic empowerment programmes, International Journal of Criminology and Sociology, 9, 1-8.
    8. Kagirova, M. V. , Rodionova, O. A. , Dumnov, A. D. , Kerimov, V. E. , and Kolomeeva, E. S. (2018), Statistical analysis of differentiation of Russian regions in terms of ensuring intensive import substitution in the livestock sector, Journal of Applied Economic Sciences, 13(4), 1069-1080.
    9. Klychova, G. S. , Safiullin, L. N. , and Zakirova, A. R. (2014a), Information-analitical support of cost management in horse breeding, Mediterranean Journal of Social Sciences, 5(18), 193-193.
    10. Klychova, G. S. , Zakirova, A. R. , Khametova, M. V. , and Sadrieva, E. R. (2014b), Special aspects of horse husbandry production costs calculation, Mediterranean Journal of Social Sciences, 5(24), 91-91.
    11. Kovanov, D. I. and Iakovleva, N. A. (2016), Application of statistical methods in the analysis of factors that determine the level of milk cost, Innovative Economy: Prospects for Development and Improvement, 8, 76-81.
    12. Mohammadi, A. , Rafiee, S. , Mohtasebi, S. S. , and Rafiee, H. (2010), Energy inputs-yield relationship and cost analysis of kiwifruit production in Iran, Renewable Energy, 35(5), 1071-1075.
    13. Molchanov, P. , Tyree, S. , Karras, T. , Aila, T. , and Kautz, J. (2016), Pruning convolutional neural networks for resource efficient inference, arXiv preprint arXiv:1611.06440.
    14. Oblad, T. (2019), Cyberbullying among emerging adults: Exploring prevalence, impact, and coping methods, International Journal of Criminology and Sociology, 8, 45-54.
    15. Rafikova, N. , Hazieva, A. , Khabirov, G. , Ableeva, A. , Salimova, G. , Zalilova, Z. , Kabashova, E. , and Valishina, N. (2019), Statistical and economic analysis of zonal conditions impact on efficiency of cattle production supply chain management, International Journal of Supply Chain Management, 8(4), 692-702.
    16. Riyadi, B. S. (2020), Culture of abuse of power due to conflict of interest to corruption for too long on the management form resources of oil and gas in Indonesia, International Journal of Criminology and Sociology, 9, 247-254.
    17. Skevas, I. , Emvalomatis, G. , and Brümmer, B. (2018), Heterogeneity of long‐run technical efficiency of german dairy farms: A bayesian approach, Journal of Agricultural Economics, 69(1), 58-75.
    18. Tkachev, S. I. , Voloshchuk, L. A. , Melnikova, Y. V. , Pakhomova, T. V. , and Rubtsova, S. N. (2018), Economic and mathematical modeling of quantitative assessment of financial risks of agricultural enterprises, Journal of Applied Economic Sciences, 13(3), 823-729.
    19. Ugwumba, C. O. A. , Okoh, R. N. , Ike, P. C. , Nnabuife, E. L. C. , and Orji, E. C. (2010), Integrated farming system and its effect on farm cash income in Awka south agricultural zone of Anambra state, Nigeria, American-Eurasian Journal Agricultural and Environt Science, 8(1), 1-06.
    20. Zakirova, A. , Klychova, G. , Doroshina, O. , Safiullin, I. , Nurieva, R. , and Zalilova, Z. (2019), Improvement of the procedure for assessing the personnel of the agricultural organization, E3S Web of Conferences, 110, 02073.
    21. Zalilova, Z. , Lukyanova, M. , Kovshov, V. , and Sharafutdinov, A. (2020), Strategic development and use of agro-food sector’s potential in rural areas, In Advances in Social Science, Education and Humanities Research, 392, 156-161.