Trends and Determinants of Indian Agricultural Prodution: An Empirical Study
It is necessary to know the trends in the total food grains of country to learn about the reasons why Indian total food grain production / yield / area are increased or decreased at a specific period. The study is based on secondary data and data ranges for the period 1990−91 to 2010−11. In order to satisfy the objectives the study employs simple descriptive as well as some econometric techniques. The study uses the Annual Compound Growth Rate (ACGR), coefficient of variation, simple ordinary least square techniques and Multicollinearity.
Key words: Indian food grains, Trends, coefficient of variation, and determinates.
Agriculture is the primary sector and it plays a vibrant role in the economic development of India. Nearly 14.6 percent of the gross domestic product is accounted by agriculture sector where about 50 percent of the workforce is engaged. India has a variety of climatic and soil conditions. This results in diverse agro−climatic zones and makes it possible to grow a wide variety of agricultural products. India has become self−sufficient in food grains. Its agricultural exports accounted for around 9.1 percent of India’s total export for the year 2010−11. Now Indian agriculture has been going through a serious crisis during the post−reform period. Besides domestic concerns, such as decline in productivity, high input cost, stagnant net sown area, declining public sector investment, inadequate availability of institutional credit, and rising agricultural imports, Indian agriculture has been facing external challenges under WTO (World Trade Organization) regime. (Kamat, et al, 2007).
OBJECTIVES OF THE STUDY
The study focuses on temporal changes in food grains production during 1990−91 to 2010− 11, and also inquires the role of agriculture determinants in increasing production output.
RESEARCH METHODOLOGY Data source
The data for the study has been collected from Economic Survey ( various issues), publication of ministry of finance, GOI and the agriculture statistics at a glance ( various issues), a publication of ministry of agriculture. Besides, other sources are Handbook of Statistics on Indian Economy, publication of Reserve Bank of India (RBI).
Trend Analysis− Growth rate Analysis
Various statistical tools were used for the analysis of the secondary data about annual compound growth rate of area, production, and yield of total food grains from 1990−91 to 2010− 11. The annual compound growth rate is computed by employing following formula: Y = abt By using logarithm, it may be written as: Log y = log a + t log b Y* = a* + t.b* (where log y = y*, log a = a* and log b = b*) The value of b* is computed by using OLS Method. Further, the value of ACGR can be calculated by following method: ACGR = (Antilog b* −1) x 100
The instability was measured by estimating the coefficient of variation of production, area, and yield. The coefficients of variation of these parameters were calculated as under:
Standard deviation CV (% ) = −−−−−−−−−−−−−−−−−−−−−−−−−−−−−− X 100
Mean Analysis of determinates of Indian Food Grain Production
In order to find out the, determinates of Indian Food Grains Production since 1990−91 to 2010−11 we use the following model:
lnY = + ln NSA+ £ lnNIA+ lnCF + lnP + ¥ lnCE + U .. (1) Where Y is the Indian Food Grains Production NSA is the Net Sown Area NIA is the Net Irrigated Area CF is the Consumption of Fertilizer P is the use of Pesticides CE is the Consumption of Electricity Equation 1 the parameters , , £, , , and ¥ are linear and parameters are the respective elasticity. This model is also known as log− log model and finds out the Multicollinearity in this study.
RESULT AND DISCUSSION Trends Analysis
In the earlier years of planning, food availability was a serious problem in India. Total food grains production was hardly 95.32 million ton in the year 1990−91, which increased to 241.56 million ton at the end of 2010−11. Total wheat production was 55.14 million ton in the year 1990−91, which increased to 95.32 million ton at the end of 2010−11. Total rice production was 74.29 million ton in the year 1990−91, which increased to 85.93 million ton at the end of 2010− 11. Total Coarse Cereals production was 32.7 million ton in the year 1990−91, which increased to 42.22 million ton at the end of 2010−11. Total Pulses production was 14.26 million ton in the year 1990−91, which increased to 18.09 million ton at the end of 2010−11 (Ministry of Agriculture, Government of India).
The annual compound growth rate or trends of area, production and yield of food grains are shown in table no.1. The data reveals that during 1990−91 to 2010−11, the growth rate in total food grains production has shown an increasing trend and is at 1.66 percent. The rate of rice is 1.254 percent. Similarly increase in growth rate for wheat is 2.24 percent, total cereals are 1.285 percent, and pulses are 1.58 percent. For entire period of 1990−91 to 2010−11, the highest growth rate of area is noticed for wheat (0.958 percent) and negative for rice (−0.02 percent), coarse (− 1.36 percent) and total food grains (−0.08). The growth of yield is highest in coarse (2.68 percent) and the lowest in pulses (0.882).
During the 2 decades, there have been some shifts in the areas from rice, coarse, pulses and total food grains (exclude wheat). The lower area under food grains has been due to shortfall in the area under jowar in Maharashtra, Rajasthan, and Gujarat, and bajra in Maharashtra, Gujarat and Haryana, and in case of pulses in Maharashtra, Uttar Pradesh, Andhra Pradesh and Rajasthan (Economic Survey of India, 2011−12). This Shift in cropping pattern was taking place due to remunerative price being offered to commercial crops and better market access given to growers ( Kamat, et al, 2007).
It is very necessary to analyze instability
of Indian food grains production in term of area, yield and production. Results from table no.2 indicate the instability of term in the entire period (1990−91 to 2010−11) based on the coefficient of variation (CV) of food grains´ area, yield and production.
Results in table 2 show that in case of rice area the coefficient of variation is more stable at (CV=2.81 percent) and similarly total food grains also stable at (CV= 2.18 percent). The other food grains are observed unstable with high value of coefficient of variation. The reason for the high instability in these food grains may be due to the deflections in the growth rate in the area.
In case of yield, it is found that more instability is prevailing in all food grains with high value of coefficient of variation. This is due to the poor agricultural growth rate which led to more instability. In case of production, it is observed that mostly food grains are in an unstable growth with high coefficient of variation. For the entire period (1990−91 to 2010−11) analysis shows that all area, yield and production are showing high instability. Production is more unstable than yield and yield is more unstable than area. It means that there are considerable quantitative increases in terms of production.
DETERMINANTS OF INDIAN FOOD GRAINS PRODUCTION
In order to check for the determinants of food grains production, the present study uses five explanatory variables to find out its relationship with output. These determinants are net sown area, net irrigated area, consumption of fertilizers, use of pesticides, and consumption of electricity. These explanatory variables determine the food grains production in the four different models. Our study is thus limited due to unavailability of variables.
The matrix of pair wise correlation coefficients is presented here. The last column stands for the variable numbers. Thus, the correlation between variable 2 and 3 is 0.934. Note that correlation coefficients show some high values. Diagonal elements are all 1.00 because the correlation between a variable and itself is one. As expected, Variable 2 `Net Irrigated Area´ is positively correlated with Chemical Fertilizer and negatively correlated with Pesticides. Chemical Fertilizer is negatively correlated with consumption of Electricity. We can expect these high correlations to introduced multicollinearity among these variables and affected regression results.
Model−1 (1990−91 to 2010−11) uses only five variables for measuring food grain production function because we cannot get complete data on remaining variables for the same period. Our results point out that all explanatory variables have an insignificant effect on the food grains production. However, the consumption of electricity has negative impact on the total food grains production, and net sown area, net irrigated area, consumption of fertilizers, and pesticide have positive effect on food grains production, but these variables remain statistically insignificant. The ¯R2 value of this model is 0.81 which means that about 81 percent of the variation in the log of dependent variable is explained by the log of these five variables. Here variable 5 consumption of electricity is excluded from the model to remove the problem of multicollinearity because it has high value of Prob.
Model−II uses only four variables for measuring production function. In this model all variables have positive effect on food grains production but these are statistically insignificant. Here variable 4 `use of pesticide´ is excluded from the model to remove the problem of multicollinearity because it has high value of Prob. After excluding the variable 4 there is no effect on the ¯R2 value. The ¯R2 value is 0.82 percent.
Model−3 uses only three variables for measuring the production function. In this model net sown area shows positive and significant effect on the food grains production at 5 percent level, It means that though the net sown area is increased by 1 percent food grains increased by 0.402 percent and net irrigated area shows positive and significant effect on the food grains production at 10 percent level. It means that though the net irrigated area increased by 1 percent, food grains increased by 0.6 percent. After excluding the variable 4 there is no effect on the ¯R2 value. The ¯R2 value is o.83 percent. Here variable 3 `consumption of fertilizers´ is excluded from the model to remove the problem of multicollinearity because this has high value of Prob. Model−IV uses only two variables for measuring the production function. Both variables net sown area and net irrigated area are positive and statistically significant at 10 percent and 1 percent levels. The ¯R2 value is o.82 percent.
1. GOI (2010), ’ Agricultural Statistics at a Glance 2010 and earlier issue’, Directorate of Economics & Statistics, Department of Agriculture & Cooperation, Ministry of Agriculture and Economics survey of India, Govt. of India, and New Delhi.
2. Kamat, et al (2007), ’Indian Agriculture in the New Economic Regime, 1971−2003: Empirics based on the Cobb Douglas Production Function’ Available at: http://mpra.ub.uni−muenchen.de/6150/
3. Ramanatam, Ramu (2002) ’ Introductory Econometrics with Application’ Fifth Edition Published by Thomson Asia Pte Ltd., Singapure Reserve Bank of India (2010), ’Hand book on India Statistics, Department of Economic Policy and Research, Mumbai
4. Thanh. C. Nguyen and Singh Baldeo (2006), Trend in Rice Production and Export in Vietnam’Available:http://docs.google.com/-viewer?a=v&q=cache:OFDpTjqpEcMJ:clrri.org/-lib/omonrice1413.pdf+trend+in+of+rice+production+and+export&hl=en&gl=in&pid=bl&srcid=ADGEESjHcrPHla79N