Abstract
Rapid rise in the prices of foodgrains and their continued upsurge is a matter of concern not only for the government and policy makers but also for anyone concerned with social welfare. This is particularly so because the increased prices of basic food items cause great distress to the poor sections of society who spend a large part of their income on food. Thus, a clear understanding of the causes of inflation is necessary for framing the right policy to tackle the problem. The current study tries to examine how prices are determined in the Indian foodgrain market. This requires a slightly different approach from the conventional demand and supply framework as government intervenes in the market through open-market operations. To this end, we propose a structural model, explaining the behaviour of foodgrain prices for the period 1980–81 through 2011–12 incorporating the role of government interventions. Our results confirm that there is a strong impact of demand- as well as supply-side factors. However, when it comes to controlling inflation, demand-side management turns out to be highly significant. Under supply-side management, increased capital stock is found to be important, in so far as it significantly boosts production. Government intervention through procurement and off-take plays a stabilising role.
JEL
INTRODUCTION
Inflation is widely seen as a major problem that must be quickly and effectively dealt with. It is argued that persistent inflation may destabilise the process of growth in general and also have specific and significant adverse welfare implications. With regard to growth, inflation may be seen to induce a measure of uncertainty which may cause a dent in the investment process in addition to the impact of higher costs of borrowing. As far as welfare is concerned much depends on the structure of inflation, because prices are bound to rise at different rates. In countries like India, where a large proportion of the population is below the poverty line, an increase in the prices of basic goods has an excessively adverse impact on the standard of living of the poorer sections of the society. Moreover, since the increased prices paid by the final consumer are seldom passed on to the basic producers, who may also belong to the poorer sections of society, the impact is largely adverse. In any case, many in the poorer sections of society are buyers rather than sellers of these basic goods. Quite likely, the real incomes of the poorer sections would record a sharper decline than that of the more privileged sections. With more than 30 per cent of its population living in poverty, an increase food prices is particularly serious for India, and this motivates the present exercise.
An in-depth analysis of the policies to control food inflation is of concern to government agencies, policy makers and professional economists for two main reasons. First, the solution is neither obvious nor one-way. Attempts to minimise the impact of higher food prices by subsidising them or meeting domestic demand through imports are bound to increase public expenditures and result in high import bills, respectively. With India’s fiscal deficit currently above 5 per cent of its gross domestic product (GDP), it faces a particularly difficult situation. Moreover, high food prices would typically raise wages and the material costs of most industries which in turn mean higher inflation all around. Moreover, the use of monetary policy which may require raising interest rates and reducing money supply may have a seriously adverse impact on investment. Thus, a wrong diagnosis of the problem and corresponding policies may have spillover effects on other sectors, leading to macroeconomic imbalances. Clearly the likely solutions are not straightforward.
FOODGRAINS PRICES: OVERALL TRENDS
Foodgrain prices started off remarkably in the early 1950s with a deflationary phase for five consecutive years from 1950–51. 1 This was accompanied and indeed caused by significant output growth in the same quinquennium, with growth rates in foodgrain production averaging more than 7 per cent. However, these trends in prices as well as output were not sustained thereafter. No wonder that the First Five-year Plan primarily focused on uplifting the agrarian economy of India. Notably, the then Prime Minister Jawaharlal Nehru said ‘everything else can wait but not agriculture’. Of the total plan outlay, 31 per cent was devoted to the agricultural sector. However, in subsequent Plans, the importance given to the agricultural sector was steadily reduced. As a result of reduced public investment, in addition to other factors like adverse weather conditions, increased pressure on land for non-agricultural purposes and increased input prices, foodgrain production in the recent decades more or less stagnated.
Inflation and Growth Rates for Foodgrains (Quinquennial Averages)
Inflation and Growth Rates for Foodgrains (Quinquennial Averages)
However, it is quite remarkable that from about 51 million tonnes in 1951–52 the production of foodgrains has increased to 258 million tonnes in 2011–12. But, we also see that this rate of growth in foodgrain production increased at a decreasing rate. The Five-year average growth rate in the production of foodgrains declined from 7.88 per cent in 1950–51 through 1954–55 to about 3 per cent for the next 15 years, and was almost stagnant over 1970–71 through 1974–75. Although the average growth rates picked up in the next ten years, the upward trend was once again not sustained. In the following years, it again fell consistently registering negative rates of growth from 2000–01 to 2004–05. From Table 1, it is quite clear that for those periods when the growth rate of production was low, the corresponding rate of inflation was high. This is important from a welfare point of view because rice, wheat, coarse cereals and pulses—the major components of the overall foodgrain basket—happen to be the staple food for the majority of the Indian population.
In 1956–57 when foodgrain production fell by about 7 per cent, prices immediately shot up to over 27 per cent. Fortunately for the economy, these explosive trends eased quickly and foodgrain prices became moderate subsequently. In the 1960s, foodgrain prices again exploded in 1964–65. Along with modest production, the Indo-Pak war had a significant impact on prices this time, and the growth rates in grain prices were closer to an average of 20 per cent for the years 1964–65 through 1967–68. To add to the problem, in 1965–66 annual rainfall was the lowest since Independence.
Foodgrain inflation in the early 1970s had causes similar to the inflation of the 1960s. Initially, the second Indo-Pak war over the creation of Bangladesh in 1971 pushed prices up. However, in the following years, the combined effect of the international oil shock and deficit monsoons exacerbated the problem by putting further upward pressure on prices. In 1971–72 the growth rate in production was negative at 3 per cent, which further declined to −7.73 per cent the following year. Although production increased marginally in 1973–74, it fell again by 4 per cent in 1974–75. Quite naturally in these years India had to import foodgrains to meet domestic demand, which further fuelled domestic prices as food prices were already high in the international markets. Owing to all these factors, foodgrain prices increased rapidly and the rate of increase almost touched 40 per cent by 1974–75.
During the first half of the 1980s, grain prices remained high at an average closer to 7 per cent despite excellent production. In fact, the growth rate in foodgrain production was above 6 per cent which had become a rarity since the early 1950s. However, the upward trend in prices was partly because of the adverse impact of the second oil crisis that fuelled relative prices. In the latter half of the 1980s, production was fairly good except for 1986–87 and 1987–88. In 1988–89 foodgrains marked record production with an annual increase of 20 per cent. Quite surprisingly, prices were high at 14 per cent this year, but this could be a result of the two preceding bad years. The next hike in foodgrain prices was in 1991–92 when prices increased at 20 per cent, again surprisingly, as the monsoons had been very good. During this period, cost-push factors were prominent and the economy was adjusting to structural changes. As part of the new policies, there was huge dose of devaluation which made imports costlier. Nevertheless, the overall foodgrain inflation in the 1990s was high with the decadal average almost touching 11 per cent. On the production front, the average decadal growth rate in production was just above 2 per cent with three years having negative signs.
In the first decade of the twenty-first century, contrary to earlier trends, foodgrain inflation was under control despite poor production growth. The Five-year annual average inflation during 2000–01 through 2004–05 was only 0.13 per cent, even with a decline in average production by 0.18 per cent. However, over the next five years, 2005–06 through 2009–10 annual rate of growth of foodgrain production was little more than 2 per cent despite a 7 per cent decline in production in 2009–10. Nevertheless, quinquennial foodgrain inflation during these five years increased to 10.76 per cent. This paradoxical trend in recent years has gained attention, as many academicians and policy makers consider it a clear case of inflation dynamics undergoing a paradigm shift with prices driven by demand factors and not merely by production trends.
A closer look reveals a more general problem, namely, a shortage of food supply along with increasing demand, because of increasing incomes, monetary growth, increasing population and urbanisation, have fuelled food inflation in recent years. The growth rate in GDP in the post-reform period has on average been of the order of 6.3 per cent compared to 4 per cent in the pre-reform era. Thus, the current episode of inflation needs to be understood as relating to a period in which GDP growth has been high. The fact that inequalities in the distribution of income and wealth influence the level and pattern of inflation also needs to be underlined in understanding inflation as well as in shaping government policy.
While these factors are fairly well understood and widely discussed for a freely functioning market economy, their specification is relatively harder for a developing economy where government intervention takes different forms in different situations. These interventions are broadly motivated by the desire to minimise the distress of the common people. Government interventions in the Indian food market are mainly through the Food Corporation of India (FCI) through which government procures and maintains stocks of foodgrains which are released through the public distribution system (PDS) and open market operations. 2 Keeping in mind the need for a clear understanding of the underlying process we specify a structural model with a conventional demand–supply framework, but incorporating government interventions to explain movements in foodgrain prices. This is important.
Policy makers, commentators and a wider class of researchers link inflation prominently with monetary phenomena, even though the discussions may not routinely be guided by the quantity theory of money or even by sophisticated monetarism or the new classical economics. A structuralist view of the problem does, however, make us look more closely at other factors underlying the price system even at the macroeconomic level. In an early study Pandit (1978) highlights the role of cost-push factors, which supports the structuralist explanation with considerable emphasis on the price of food and raw materials originating from the agricultural sector. Taylor (1983) clearly emphasises how lagging food supply leads to inflation. For further insights into the problem, we may also refer to other studies, such as Pandit (1984), Bhujangarao (1987), Sekhar (2003), Gaiha and Kulkarni (2005) and Kumar and Sharma (2006).
Studies relating to the movement of food prices in India have generally tried to analyse these under the simple demand–supply framework at equilibrium price and quantities. Many of these studies
3
have used either a single-equation reduced-form framework or partial adjustment models to explain the inflationary process. This may formally be expressed as:
Where price level P is estimated as a function of money supply M, net availability Q, and a dummy variable ZPF to account for other factors. With a year lagged dependent variable as well as other factors it gives us a dynamic system. The model is basically within the quantity theory framework with an adaptive expectation formulation, where current/future prices are estimated as functions of lagged and contemporary prices and money supply. However, such a specification may not be able to adequately capture the price dynamics in India, especially that for agricultural/food articles for two reasons. First, quantity cannot be considered purely exogenous as there can be possibilities of bi-directional causality between prices and quantities demanded/supplied in the market. Second, government interventions play a crucial role in the market.
Estimating a demand–supply framework at equilibrium price and quantity can considerably solve this problem by accounting for interactions between price and quantity phenomena. Further, the role of government interventions can be brought in either through demand or supply equations. However, this formulation will require solving the model simultaneously with appropriate estimation methods.
4
Alternatively, considering all the factors in each of these equations, a reduced-form equation can be estimated either with prices or quantity as the dependent variable. This model can be estimated using an OLS procedure. Such a reduced-form specification estimated by Bhujangarao (1987) is as follows:
Where, P is the price of foodgrains, estimated as a function of average costs C, yield YLD, one-year lagged stocks SK(–1), PDS, money supply M, real income GDP and credit to agriculture CRAG. The foregoing framework was highly relevant and well accepted for the estimated sample period till 1983. The analysis confirms the significance of the PDS, money supply, input costs and yields in determining price movements.
However, the structure of the Indian economy has changed significantly with its opening up in the early 1990s which threw the protected agricultural sector open to external challenges. Under the open economic framework, fluctuations in international prices are likely to affect relative prices in the domestic economy through exports and imports (Sekhar, 2003, 2011). Although an explanation of price movements under the reduced-form equations can to an extent be used to bring important variables into consideration under a single equation, the formulation will not allow the model to effectively capture the dynamics for determining prices, especially for variations in the rates of inflation. Moreover, given the macroeconomic set up, the strong role of structural factors in determining the rates of inflation cannot be sidelined.
A way out for estimating the model clearly lies in the use of a structural model (Pandit, 2000, 2004). The basic aim is to reflect the inter-dependence of endogenous variables which are jointly and simultaneously determined for equilibrium. Exogenous variables can either be policy instruments or lagged dependent variables or, even natural factors. Estimating the structural model will also have the advantage of specifying separate equations for each endogenous variable rather than bringing them under one reduced-form framework. Studies by Sekhar (2003) and Kumar and Sharma (2006) have used structural models in modelling some disaggregate food prices.
Irrespective of the analytical framework or the particular specifications, most studies have identified a few common variables that determine food prices: production measured in terms of actual production or even net availability; various measures of income, liquidity and population representing demand variables; various cost-push factors like area of production, fertiliser consumption, agricultural credit, wages etc.; policy variables such as procurement prices, support prices, government stocks; and more importantly natural factors like rainfall and area cultivated.
Keeping the foregoing issues in mind we specify our model as follows. Departing from the usual procedure of explaining the rate of inflation, we set up two market-oriented relationships—one for demand and one for supply. In addition, taking into account the fact that government intervenes in the market through its procurement policy, under which it announces the minimum support price (MSP) and accepts whatever quantity producers are willing to offer for procurement. Moreover, the government also runs PDS under which it offers a certain quantity (referred to as off-take) of foodgrains, mainly cereals. This framework gives us a structural model capable of a richer explanation of the price movements. Before we specify this, let us recapitulate the main features of what one is trying to capture.
Government runs PDS under which it maintains a stock of foodgrains acquired by procurement. A certain amount of this quantity is released in the market every year through the PDS.
An increase in the price of other food articles may tempt people to substitute the costly food with relatively cheaper grains, which may further increase the prices of foodgrains. This in short captures the effect of relative prices.
Farmers are free to offer part of their production for procurement at the MSP fixed by the government for each product for the year. Farmers have an alternative option to sell their produce in the market.
Supply is dependent on the capital stock in agriculture, MSP, area cultivated and weather conditions, as measured by the annual rainfall.
On the demand side we must include real disposable income and the liquidity condition as measured by real money stock.
The variables and the notation used are as follows.
The structural model, in keeping with the foregoing discussion, is as follows:
The model alone is referred to as a structural model which involves three endogenous variables P, Q and PR along with a series of exogenous variables. The basic point in this formulation is that it reflects the dependence of the exogenous variables which are jointly and simultaneously determined for equilibrium. The exogenous variables are either policy instruments or natural factors. These include: real income (Y), real money supply (M3), rainfall (R), prices of other food articles (P*), area under cultivation (A), MSP, off-take (OT) and KAG. Finally, considering the possibilities of bi-directional causality between the variables, the model is estimated using the 2SLS method rather than OLS.
The demand Equation (3) is normalised in terms of price P which is determined by both the current and one-year lagged value of quantity Q along with other independent variables like Y, M3, impact of relative prices through P*, government intervention in the market through PR and OT. Where, Y, M3 and P* are expected to have a positive impact on the dependent variable P. The lagged value of quantity Q (−1) is important as more than half the production in a given financial year is consumed in the next year (Chand, 2010). The coefficient of Q (−1) thus in a way measures the subsequent year’s inflation expectations given output. Nevertheless we expect a negative sign in both the cases. While considering government intervention, we have taken the difference between PR and OT as the relevant factor, as the effectiveness of government interventions depends on how procurement exceeds or falls short of OT.
Supply Equation (4) is specified with quantity produced, Q, as the dependent variable. Considering an elementary microeconomic framework, prices P enter the supply equation as an increase in price will encourage sellers to supply more. KAG is expected to have a positive impact on output as the former increases productivity and gives access to better technology and farm inputs. Government interventions enter the supply equation through MSP. Once again a higher MSP is an assurance to farmers, which encourages them to produce more. Other variables like rainfall, RF, and area cultivated, A, also enter the supply equation accounting for domestic supply inputs.
Apart from the standard demand–supply framework, an important feature of the model is the inclusion of Equation (5) for PR, in which we consider the difference of MSP and P, as an independent variable. This formulation is intended to take into account dual-selling windows available to the sellers. 5 In other words, sellers have an option to sell their product to government when the MSP is attractive. In an alternative scenario, if the MSP is below the market price, procurement will be low, as suppliers get a higher price in the local market. It can also be argued that higher procurement can lead to an increase in prices as it reduces availability in the common market. To this end, we have accounted for the link from the procurement equation to the demand equation by taking the difference between procurement and off-take as an independent variable, expecting a positive relation with the dependent variable, namely, price P. Furthermore, quantity, Q, enters the procurement equation with a year lagged value and is expected to have a positive impact on the dependent variable.
The model presented is in keeping with the basic theory involving several macroeconomic entities and facts on the ground. The purpose is to start with a micro-theoretic model which can eventually explain the overall macroeconomic phenomena under discussion. However, as mentioned earlier, these factors are not independently determined and have continuous interactions which make the system complex and dynamic. Since the empirical estimation of such a model using OLS may result in inconsistent estimates, we have used the 2SLS methodology for estimation as stated earlier.
Before going into the estimation of the model and discussion of the results we present details of the data used for the study. The data are obtained from official published sources or official websites of the relevant departments. The sample period for the analysis is from 1980–81 through 2011–12 and as mentioned earlier, variables are taken as growth rates. The following table presents the data and units, and their official sources. However, in the specified model, each variable is measured as percentage (y-o-y) rate of growth.
Data and Sources
Data and Sources
Augmented Dickey–Fuller (ADF) Stationarity Tests
The price of other food articles is taken as the weighted average of the WPI for fruits, vegetables, milk, eggs, meat and fish at 2004–05 prices. The MSP for total foodgrains is the weighted average of the MSP for coarse cereals, rice, wheat and pulses; where the weights are equal to their corresponding weights in the WPI with 2004–05 as the base year. Real personal income and real money supply are obtained by deflating the corresponding nominal magnitudes with the WPI for all commodities (WPI-AC) at 2004–05 = 100.
The assumption that the time-series used are stationary has to be carefully checked before estimation of the model is undertaken. This is checked using the Augmented Dickey–Fuller (ADF) test. Since the variables enter the model as growth rates, the stationarity tests are conducted for the converted series. The results are given in Table 3.
The three equations are then estimated using the 2SLS methodology considering the samples from 1980–81 through 2011–12. Each equation includes a dummy variable to take care of outliers caused by factors beyond consideration of the model. However, such outliers are typically very few. The ‘t’ statistics in parenthesis below each coefficients indicate the statistical significance of the coefficient. Throughout the estimation process we have used the following instrumental variables which satisfy the order condition 6 as required for the 2SLS estimation process.
Instruments (exogenous variables) include: P (−1), Q (−1), PR (−1), OT, Y, M3, MSP, KAG, A, R, DUM1, DUM2 and DUM3.
To begin with, we estimate the demand equation as follows.
In the demand equation, relative prices and income turn out to be highly significant along with the quantity available, money supply and PDS. Significant and positive coefficients for income and money supply indicate increased demand pressure on the rate of growth of foodgrain prices which is nothing but inflation. For every 1 per cent increase in the rate of growth in income and money supply, foodgrain inflation, as far as demand is concerned will increase by 1.16 per cent and 0.54 per cent, respectively. The impact of relative prices is also positive as expected; for each 1 per cent increase in the rate of inflation for other food articles, inflation for foodgrains increases by 0.51 per cent. Both the current and lagged coefficients for the production of foodgrains as expected have a negative and significant impact on inflation. For a 1 per cent increase in the rate of growth of quantity available, inflation decreases by 0.38 per cent, whereas the impact of the lagged coefficient will pull inflation down by 0.34 per cent. The difference of procurement and off-take, as expected, has a positive impact on prices and is fairly significant. DUM1 is assigned a value of +1 for the year 1985 and 2000 and −1 for 1987, 2001 and 2011. The model has got a good fit and the DW statistic rules out the possibility of serious auto correlation.
The supply of foodgrains is normalised in terms of quantity as a dependent variable. The estimated supply equation can be as follows:
In the estimated supply equation prices, rainfall, area of production, MSP and capital stock are found to be positive and significant in determining foodgrain production. For every 1 per cent increase in the rate of growth of minimum support prices, the output growth rate increases by 0.15 per cent and is significant. Another important variable from the policy perspective is capital stock—for each 1 per cent increase in the rate of growth of capital stock, output increases by 0.69 per cent and is significant. Similarly, increasing prices also encourage the producers and increase output by 0.26 per cent. The dummy variable DUM2 takes value +1 for 1985, 1988, 1997 and −1 for 2010 and 2011. The model has got high levels of goodness-of-fit and excellent D-W statistics.
The estimated equation for procurement can be as follows.
As explained earlier, in the procurement equation we have used the difference of MSP and prices instead of using them independently. The effect is found to be positive, implying that an increase in the MSP over prices will encourage producers to sell grains to the government. It is estimated that when the difference between the rates of growth of the MSP and prices increase by 1 per cent, the rate of growth in procurement increases by 1.41 per cent. The effect of production is also estimated to be positive as expected. For each per cent increase in the rate of growth of lagged quantity produced, the growth rate of the quantity procured will increase by 0.48 per cent. DUM3 is +1 for 1994, 2000, 2009; and −1 for 1992 and 1998. The goodness-of-fit for the model is satisfactory and the D-W statistics rule out the possibility of serially correlated error terms for the equation.
The overall fit for each of the three equations is reasonably high and has clearly brought out the dynamics in the foodgrain market. The estimated coefficients are significant and theoretically acceptable. First, we see that the growth of real income and real money supply have strong positive impacts on prices. This should be taken as a clear indication of increasing demand pressure on food prices. The impact of relative prices is also found to be positive and significant, where the substitution effect leads to a higher demand for grains and is thus inflationary. The difference between procurement and off-take is also found to be inflationary when the former is greater than the latter. The supply side of price movements is measured through rainfall, area of production capital stock and MSP. The impact of these factors was found to be positive and very strong. The positive sign of capital stock on output is important, as in India the share of public as well as private investment in agriculture is found to be declining and growth rates in production are found to be stagnant (Mani et al., 2011). In brief, we can argue that higher income and greater money supply are factors which significantly contribute to food inflation. And other factors like area, rainfall, capital stock, off-take, procurement and MSP also have the expected effects by boosting production and price stability.
Following the foregoing results we have the complete model as follows:
We subject the estimated model for counterfactual simulation experiments under alternative policy scenarios to understand the policy implications of the estimated structural model. However, to ensure validity of the exercise, we first need to check the accuracy of the estimated model. For this we first obtain the baseline solution.
7
For testing accuracy, two commonly used measures are:
Root Mean Square Percentage Error (RMSPE) for an endogenous variable y is given as follows:
Theil’s U statistics is given by:
where n is the size of the sample period, ys and ya are the forecasted or baseline solution and actual value of the endogenous variable y, respectively. RMSPE and the U are given in Table 4.
Model Accuracy: Theil’s U S tatistic and RMSPE, 1980–81 through 2011–12
Model Accuracy: Theil’s U S tatistic and RMSPE, 1980–81 through 2011–12
The forecasting performance of the model for the entire sample period can be obtained by taking 1 − Theil’s U, where a value between 0.80 and 1.00 indicates strong forecasting power. For the rate of inflation and quantity, the 1 − Theil’s U is 0.86 and 0.85, respectively, which falls well within this bracket. However, for procurement this is about 0.74, which indicates a fairly moderate forecasting performance. RMSPE values for the model is well below 1 per cent for all three endogenous variables. These clearly indicate that the predictive performance of the model is fairly good in capturing movements and turning points in the dependent variable. The corresponding graphs for the last 12 years are given in Figure 1.
Given these fairly satisfactory results we carry out simulation exercises under two broad categories: demand management and supply management.
Demand management can be seen as two independent exercises. In the first one, the growth rate of income is reduced by 3 per cent and, in the second one, the growth rate of the money supply is reduced by 5 per cent. The reductions in income and money supply are expected to imply a lower rate of inflation. This essentially can also be looked as a situation where some percentage of growth is sacrificed to contain inflation.
Under supply management, we first look at the impact of increased capital stock on the endogenous variables, namely output, inflation and procurement. For this purpose, the rate of growth of net capital stock in agriculture is increased by 5 per cent. This is expected to have a positive impact on output. This further implies lower inflation and higher procurement. Second, we look at the effectiveness of supply management through government intervention in the foodgrain market using the off-take and MSP. This is done by simultaneously raising the level of the annual off-take and support prices by 5 per cent. This policy mix is important as the government cannot set all its policy instruments independent of others. 8 It is expected that a higher MSP will encourage higher production and also provide a greater incentive to supply available foodgrain for procurement. The results from the simulation exercises are presented in Tables 5, 6 and 7. In the tables, we have presented the change in the rates of growth of output, inflation and procurement. For comparison with the actual figures refer to Appendix C, Tables C1 and C2.
Changes in Rate of Foodgrains Inflation under Demand Management
Changes in Rate of Foodgrains Inflation under Demand Management
Growth of Foodgrains Production, Inflation and Procurement (5 per cent increase in net capital stock in agriculture)
Growth of Foodgrains Production, Inflation and Procurement (5 per cent increase in MSP and off-take)
As expected from the structure of the model, demand management helps to reduce the rate of inflation. The data in Table 5 show that a 3 per cent lower growth rate for income would tend to reduce the rate of foodgrain inflation by a little less than 3 per cent. On the other hand, a 5 per cent lower growth rate of money supply reduces the rate of food inflation by a little more than 2 per cent. These results are not merely interesting but considerably useful in the policy context, where in recent times seemingly supply-driven food inflation was targeted through demand management.
Turning to supply management, we look at the impact of increased capital stock and government intervention on output, inflation and procurement. Under government intervention, both the MSP and off-take are hiked simultaneously. The way in which the model is structured, with an increase in capital stock and MSP we expect a direct positive impact on foodgrain production. Further, its impact on inflation and procurement will depend on how much each of these responds to an increase in quantity of production. However, apart from increased production leading to higher procurement, the latter also has a direct positive relationship with the MSP. The results are presented in Tables 6 and 7.
Comparing the two alternative scenarios under supply management (Tables 6 and 7) it is clear that increased capital stocks have a larger impact on foodgrain production and inflation. For a 5 per cent increase in the rate of growth of capital stock, the rate of growth of foodgrain production increased by an average closer to 3 per cent and this in turn reduces inflation by 2 per cent and adds a little more than 4 per cent to procurement. On the other hand, the impact of a 5 per cent increase in the MSP and off-take is found to be positive and stronger on procurement, with the latter increasing at an average close to 8 per cent. This is important from the point of view of food security and buffer stocks maintenance. However, the impact of government intervention is found to be marginal on production and inflation. Interestingly a higher MSP has not turned inflationary as many would have expected it to be.
A cross-comparison among the various policy alternatives clearly brings forth the role of both demand and supply measures in managing foodgrain inflation. However, when it comes to controlling foodgrain inflation, the role of demand management tends to be stronger. On the supply side, the impact of higher capital stock is more prominent compared to government intervention. Although higher capital stock helps reduce inflation, its impact is rather indirect through increasing productivity and output. This is generally believed to be more effective in the long run than in the short run. Finally, the role of government intervention is found to be more a tool to ensure price stability and food security than a measure to control inflation.
The basic motive for this article has been to identify the determinants of foodgrain inflation in India, under the given market conditions. When food prices started rising in the latter half of the last decade, many questions were asked as to what had caused a sudden resurgence in food prices after somewhat stable food price movements in the early years of the decade. The immediate response has usually been to relate this to perceived supply shocks in the preceding years. This explanation appeared to be fairly convincing till food prices shot up in 2009–10 despite favourable climatic conditions in 2008–09. Moreover, government policies and market interventions did not produce expected results as prices continued to increase. This motivated us to analyse the problem in a larger context. In particular, we have attempted to incorporate government intervention in the foodgrain market.
To this end we have specified a structural model constituting three equations relating demand, supply and procurement. The model is estimated using the 2SLS procedure covering the sample period, 1980–81 through 2011–12. The results confirm the strong and significant impact of demand as well as supply factors in determining foodgrain prices. On the demand side it is income, money supply and relative prices that turn out to be important whereas the significant supply factors are rainfall, area cultivated, capital stock and the nature and extent of government intervention. With regard to this we have specifically looked at the impact of the MSP, procurement and off-take on prices. The results are again significant as we find that the larger off-take with higher incentives for procurement has a desirable effect.
Since the model has an excellent measure of accuracy, it is legitimate to work out policy implications with confidence. The model is solved under two alternative policy scenarios. First, demand management (i) by reducing the rate of growth of income by 3 per cent and (ii) by lowering the rate of growth of money supply by 5 per cent. Second, supply management (i) by increasing the rate of growth of capital stock by 5 per cent and (ii) through government interventions in the market by increasing off-take and MSP by 5 per cent.
The simulation exercises confirm the substantial role of tighter demand conditions in controlling the increase in prices, where the impact of income is found to be stronger than that of money supply. Under supply-side management, the impact of higher capital stock is found to have a strong and positive impact on production, thereby bringing down inflation and adding to procurement. However, an increased MSP and off-take turns out to be more effective as a stabilising package by improving output as well as procurement, where the effect on the latter is found to be stronger. As mentioned earlier, this is important to ensure food security during hard times. An important point to be noted here is that for government policies to bring out the desired changes it is essential to have the right policy mix as highlighted by Rakshit (2003).
In conclusion, the current exercise clearly brings forth the role of demand and supply factors along with the government intervention in determining foodgrain prices. However, to have a complete understanding of the dynamics involved, it is important to study the problem at a disaggregate level separately for each foodgrain. Moreover, the problems of income and distributive inequalities need to be brought in, to have a complete analytical framework with more realistic policy implications. It is our intuition that income in the model is, in fact, a proxy for its increasingly inequitous distribution rather than its overall level. A formal testing of this ‘intuition’ is, however, not possible because of data problems.
Footnotes
APPENDIX A
Annual Foodgrain Inflation and Rates of Growth of Foodgrain Production
| Year | Foodgrain Inflation | Rates of Growth of Foodgrain Production | Year | Foodgrain Inflation | Rate of Growth of Foodgrain Production |
| 1951–52 | –0.78 | 2.28 | 1981–82 | 9.55 | 2.86 |
| 1952–53 | –5.49 | 13.87 | 1982–83 | 9.1 | –2.84 |
| 1953–54 | –3.53 | 17.94 | 1983–84 | 9.44 | 17.64 |
| 1954–55 | –21.51 | –2.56 | 1984–85 | –1.93 | –4.48 |
| 1955–56 | –3.56 | –1.73 | 1985–86 | 6.32 | 3.37 |
| 1956–57 | 27.84 | 4.50 | 1986–87 | 3.94 | –4.67 |
| 1957–58 | 4.44 | –7.94 | 1987–88 | 9.2 | –2.14 |
| 1958–59 | 8.94 | 19.95 | 1988–89 | 14.51 | 21.07 |
| 1959–60 | –3.91 | –0.61 | 1989–90 | 2.22 | 0.66 |
| 1960–61 | 0.20 | 6.98 | 1990–91 | 8.34 | 3.13 |
| 1961–62 | –1.83 | 0.84 | 1991–92 | 20.76 | –4.54 |
| 1962–63 | 5.37 | –3.10 | 1992–93 | 12.01 | 6.59 |
| 1963–64 | 9.22 | 0.61 | 1993–94 | 7.59 | 2.66 |
| 1964–65 | 26.39 | 10.81 | 1994–95 | 14.7 | 3.93 |
| 1965–66 | 5.97 | –19.04 | 1995–96 | 6.8 | –5.79 |
| 1966–67 | 18.50 | 2.60 | 1996–97 | 12.33 | 10.54 |
| 1967–68 | 24.89 | 28.05 | 1997–98 | 1.24 | –3.16 |
| 1968–69 | –11.96 | –1.09 | 1998–99 | 9.12 | 5.43 |
| 1969–70 | 3.60 | 5.84 | 1999–00 | 16.05 | 3.04 |
| 1970–71 | –0.70 | 8.96 | 2000–01 | –1.47 | –6.19 |
| 1971–72 | 3.40 | –3.00 | 2001–02 | –0.81 | 8.15 |
| 1972–73 | 15.57 | –7.74 | 2002–03 | 1.1 | –17.89 |
| 1973–74 | 18.74 | 7.87 | 2003–04 | 1.15 | 21.98 |
| 1974–75 | 37.98 | –4.62 | 2004–05 | 0.68 | –6.96 |
| 1975–76 | –11.08 | 21.24 | 2005–06 | 7.26 | 5.16 |
| 1976–77 | –12.29 | –8.15 | 2006–07 | 14.12 | 4.17 |
| 1977–78 | 11.59 | 13.70 | 2007–08 | 6.92 | 6.21 |
| 1978–79 | 1.29 | 4.34 | 2008–09 | 11.02 | 1.60 |
| 1979–80 | 7.42 | –16.83 | 2009–10 | 14.49 | –6.98 |
| 1980–81 | 16.88 | 18.13 | 2010–11 | 4.85 | 12.09 |
MODEL SOLUTIONS
Actual and Baseline Values for the Endogenous Variables
| Year | Inflation Actual | Inflation Baseline | Quantity Actual | Quantity Baseline | Procurement Actual | Procurement Baseline |
| 1983–84 | 9.36 | 8.68 | 17.64 | 16.76 | 6.87 | −2.45 |
| 1984–85 | −1.90 | −4.78 | −4.48 | −5.95 | 20.48 | 26.25 |
| 1985–86 | 6.32 | 3.67 | 3.37 | 2.32 | 4.45 | 0.48 |
| 1986–87 | 3.96 | 2.15 | −4.67 | −3.12 | −0.35 | 6.75 |
| 1987–88 | 9.17 | 9.63 | −2.14 | −3.23 | −25.03 | −7.38 |
| 1988–89 | 14.49 | 15.18 | 21.07 | 20.69 | −5.03 | −9.73 |
| 1989–90 | 2.27 | 4.80 | 0.66 | −2.53 | 42.34 | 30.63 |
| 1990–91 | 8.31 | 11.95 | 3.13 | 3.42 | 18.94 | −1.68 |
| 1991–92 | 20.77 | 18.96 | −4.54 | −7.01 | −28.47 | −19.69 |
| 1992–93 | 12.00 | 12.78 | 6.59 | 6.72 | 11.31 | 7.20 |
| 1993–94 | 7.58 | 9.97 | 2.66 | 3.50 | 38.22 | 28.09 |
| 1994–95 | 14.70 | 13.33 | 3.93 | 6.07 | −5.34 | −2.85 |
| 1995–96 | 6.80 | 7.64 | −5.79 | −4.91 | −11.32 | 3.93 |
| 1996–97 | 12.33 | 8.08 | 10.54 | 11.03 | −9.57 | 4.79 |
| 1997–98 | 1.24 | 2.70 | −3.16 | 1.70 | 18.96 | −2.03 |
| 1998–99 | 9.12 | 11.66 | 5.43 | 3.69 | 1.51 | −1.61 |
| 1999–00 | 16.05 | 10.89 | 3.04 | 0.77 | 27.11 | 24.64 |
| 2000–01 | −1.47 | 0.79 | −6.19 | −5.96 | 14.73 | 11.14 |
| 2001–02 | −0.81 | 2.05 | 8.15 | 7.38 | 18.31 | 5.08 |
| 2002–03 | 1.10 | 4.18 | −17.89 | −17.01 | −8.91 | 1.58 |
| 2003–04 | 1.15 | 2.56 | 21.98 | 21.45 | −3.81 | −3.17 |
| 2004–05 | 0.68 | 0.95 | −6.96 | −5.25 | 11.62 | 15.52 |
| 2005–06 | 7.26 | 9.36 | 5.16 | 4.80 | 1.59 | −9.47 |
| 2006–07 | 14.12 | 12.22 | 4.17 | 7.52 | −14.34 | −4.02 |
| 2007–08 | 6.92 | 6.29 | 6.21 | 5.60 | 5.32 | 22.85 |
| 2008–09 | 11.02 | 10.03 | 1.60 | 4.34 | 48.40 | 52.31 |
| 2009–10 | 14.49 | 11.55 | −6.98 | −7.03 | 4.41 | −2.83 |
| 2010–11 | 4.85 | 3.37 | 12.23 | 10.58 | −2.05 | 7.13 |
| 2011–12 | 3.61 | 5.23 | 5.17 | 3.52 | 16.83 | 15.38 |
PRODUCTION,INFLATION AND PROCUREMENT UNDER ALTERNATIVE SCENARIOS
Foodgrain Production and Procurement under Supply Management
| Year | Quantity Baseline | Capital Stock Higher by 5% | MSP and Off-take Higher by 5% | Procurement Baseline | Capital Stock Higher by 5% | MSP and Off-take Higher by 5% |
| 1983–84 | 16.76 | 19.98 | 17.57 | −2.45 | −0.43 | 5.12 |
| 1984–85 | −5.95 | −3.00 | −5.12 | 26.25 | 30.98 | 35.05 |
| 1985–86 | 2.32 | 5.39 | 3.05 | 0.48 | 4.59 | 8.33 |
| 1986–87 | −3.12 | −0.10 | −2.51 | 6.75 | 11.02 | 13.80 |
| 1987–88 | −3.23 | −0.28 | −2.48 | −7.38 | −3.29 | 0.92 |
| 1988–89 | 20.69 | 23.79 | 21.48 | −9.73 | −4.90 | −1.89 |
| 1989–90 | −2.53 | 0.46 | −1.78 | 30.63 | 35.70 | 38.35 |
| 1990–91 | 3.42 | 6.33 | 4.13 | −1.68 | 2.15 | 6.25 |
| 1991–92 | −7.01 | −3.97 | −6.41 | −19.69 | −15.07 | −11.78 |
| 1992–93 | 6.72 | 9.87 | 7.40 | 7.20 | 11.72 | 15.26 |
| 1993–94 | 3.50 | 6.61 | 4.23 | 28.09 | 32.52 | 35.88 |
| 1994–95 | 6.07 | 9.12 | 6.84 | −2.85 | 1.36 | 5.04 |
| 1995–96 | −4.91 | −2.02 | −4.33 | 3.93 | 8.44 | 12.19 |
| 1996–97 | 11.03 | 14.02 | 11.66 | 4.79 | 8.96 | 12.31 |
| 1997–98 | 1.70 | 4.56 | 2.38 | −2.03 | 2.63 | 5.87 |
| 1998–99 | 3.69 | 6.71 | 4.35 | −1.61 | 1.86 | 5.13 |
| 1999–2000 | 0.77 | 3.71 | 1.47 | 24.64 | 28.75 | 32.17 |
| 2000–01 | −5.96 | −2.84 | −5.29 | 11.14 | 15.41 | 18.77 |
| 2001–02 | 7.38 | 10.27 | 7.95 | 5.08 | 9.35 | 12.47 |
| 2002–03 | −17.01 | −14.10 | −16.40 | 1.58 | 5.99 | 9.48 |
| 2003–04 | 21.45 | 24.44 | 22.10 | −3.17 | 1.33 | 4.66 |
| 2004–05 | −5.25 | −2.24 | −4.49 | 15.52 | 20.11 | 22.66 |
| 2005–06 | 4.80 | 7.72 | 5.58 | −9.47 | −4.74 | −0.75 |
| 2006–07 | 7.52 | 10.52 | 8.19 | −4.02 | 0.23 | 4.85 |
| 2007–08 | 5.60 | 8.63 | 6.19 | 22.85 | 27.68 | 30.94 |
| 2008–09 | 4.34 | 7.18 | 4.99 | 52.31 | 56.96 | 60.52 |
| 2009–10 | −7.03 | −4.05 | −6.42 | −2.83 | 1.56 | 5.09 |
| 2010–11 | 10.58 | 13.50 | 11.24 | 7.13 | 10.55 | 14.23 |
| 2011–12 | 3.52 | 6.29 | 4.22 | 15.38 | 19.65 | 23.34 |
Acknowledgements
We are grateful to Professor Mihir Rakshit for his valuable suggestions. An earlier version of this article was presented at a seminar organised by the National Council of Applied Economic Research, New Delhi in July 2013. We thank the participants, particularly Dr Ganesh Kumar, for their useful comments.