“Strategies for Agricultural Liberalization: Consequences for Growth, Welfare and Distribution”

 

Citation: Parikh, K. S., N. S. S. Narayana, M. Panda and A. Ganesh-Kumar. 1995. “Strategies for Agricultural Liberalization: Consequences for Growth, Welfare and Distribution”. Report submitted to the World Bank, PP-16, Indira Gandhi Institute of Development Research, Mumbai.

 

Executive summary:

The Indian economy has not grown rapidly in the past in a command and control type of environment that offered little incentive for efficient use of resources. The economic reforms initiated in June 1991 aimed at putting the economy on a high growth path through improvements in efficiency in the industrial sector by bringing in competition.

Unlike industries, agriculture has functioned more or less, in a competitive environment all along. Nevertheless, inefficiencies in resource use in agriculture arose due to trade restrictions, inappropriate pricing of inputs and outputs, and the prevailing macro environment. Trade restrictions take various forms like, total ban on exports, export licensing/ceilings, canalization through governmental agencies, as well as restrictions on export prices. Distortions in input prices arise mainly due to subsidies on farm inputs such as fertilizer, irrigation, power and credit. A combination the trade restrictions and government intervention in commodity markets result in distortions in farm output prices.

It is time that the reforms address to these inefficiencies in agriculture so that gains from industrial reforms are enhanced further.

The purpose of this study is to examine a number of issues that arise in this connection. We have addressed the following specific questions:

1)    Should agricultural trade be liberalized? If so, at what speed i.e., over how many years? Should we liberalize all commodities or only selected ones?. In particular, should we liberalize rice as its global market is thin?

2)    What would be the impact of liberalization on the growth of the economy? On sectoral outputs? On trade? and on prices? Immediately and over 5-7 years?

3)    How would it affect welfare? in terms of calorie intake? in terms of average equivalent income? different expenditure classes in both rural and urban areas?

4)    How to offset welfare loss, if any? What would be the net impact of liberalization and such offsetting measures on growth, agricultural sector and welfare?

We have explored these and other related questions through policy simulations using an applied general equilibrium model.

We use the Agriculture, Growth and Redistribution of Incomes Model (AGRIM) of Narayana, Parikh and Srinivasan (1990). This is a sequential general equilibrium model that computes equilibrium price and output for each year in succession. Many of the supply and demand relations in the model have been econometrically estimated. It considers nine agricultural sectors and one composite nonagricultural sector. There are three sets of agents: producers, consumers (classified into five expenditure groups in each of rural and urban areas), and government. Producer behaviour determines commodity supplies and incomes. Consumer behaviour generates commodity demands and household savings. The  government sets policies such as investment targets, taxes,  tariffs, quotas, rations, price supports and ceilings. Finally, equilibrium is achieved through exchange in which  domestic  demands, together with export demand by the rest of the world for each sector's output, is equated to the sum of  domestic supply (emerging from previous year's production net of  changes in stocks) and imports.

The base/reference run is a business-as-usual scenario in which past policy regimes continue. Any policy changes relative to the reference scenario have been introduced beginning 1993 and their impacts studied over the period 1993-2000. Comparison of the outcomes of the reference run and the policy scenario for the indicators reflecting the various objectives of the society show the impact of the policy change.

In the reference run, it is assumed that the policies with respect to procurement and public distribution of foodgrains, public consumption and investment, foreign trade and aid etc. would correspond to those prevalent in the recent past. Trade quotas on different agricultural commodities range from 5% to 15% of domestic supplies. For rice, an export quota of half a million tonnes is imposed to reflect the thinness of the world market. Domestic price policy sets target prices in a way that provides the estimated degree of protection to that particular commodity.

An important assumption in the reference scenario and most policy scenarios is that total (public + private) investment follows a specified investment-GDP relationship. This is accomplished by requiring that government savings adjust through adjustment of tax rate to supplement private savings. This requirement is relaxed in some scenarios where tax rate is kept fixed and public investment adjusts.

Free trade implies removal of all price distortions including trade quotas. The model captures this by removing the prevailing protection/disprotection for the different sectors. Thus, domestic consumer price for an exported (imported) commodity is set as world market price minus (plus) international trade and transport margins in the policy scenarios. The prevailing protection rates used in the model are: wheat -0.12, rice -0.34, coarse grains 0.10, bovine & ovine meat -0.15, dairy products -0.15, other animal products -0.15, protein feed -0.30, other foods 0.40, non-food agriculture -0.10, tradeable nonagriculture 0.35. These (dis)protections are set to zero and trade quotas are removed in the policy scenarios. An exception is rice. Since world market for rice is thin, even when rice price is liberalized, an export quota of 3.5 million tonnes is imposed on rice to reflect the limitation of the world market to absorb it.

The following sets of policy scenarios have been examined in this study:

Set 1: The first set of scenarios study the impact of trade liberalization in agriculture and non-agriculture. In these scenarios, domestic subsidies on inputs continue and an export quota on rice is specified to reflect the thinness of world trade in rice.

Set 2: The effect of rice export quotas is studied here.

Set 3: Effects of alternative speeds of liberalization - viz. over 1 or 3 years - are studied here.

Set 4: The scenarios in this set study the impact of imposing import tariffs on non-agriculture and also the impact of foreign capital inflows.

Set 5: Here, the impact of removing domestic subsidies on agricultural inputs are studied.

Set 6: Effectiveness of providing additional irrigation as an alternative production incentive to subsidizing agricultural inputs is explored here.

Set 7: The role of welfare measures such as targeted and increased rationing in the context of removing subsidies on agricultural inputs is analyzed here.

Conclusions:

The conclusions emerging from the study are the following:

1)    Trade liberalization helps to accelerate economic growth in the medium run by

a)    increasing allocative efficiency within agricultural sectors and between agriculture and nonagriculture, and

b)    increasing real investment due to terms of trade effects. In fact, in seven years after liberalization in 2000, the GDP is larger by 4.5 per cent and agricultural GDP larger by 4.1 per cent compared to the reference scenario.

2)    The impact of increase in investment is much stronger than that of increases in allocative efficiency.

3)    This implies that investment goods liberalization has a greater impact on growth, even agricultural growth, than agricultural liberalization alone. Thus, a very important conclusion emerges. Nonagricultural trade liberalization is even more important for Indian agriculture than agricultural trade liberalization. This also means that the process of liberalization which has so far only reduced nonagricultural protection cannot be said to have bypassed agriculture.

4)    Agricultural liberalization beginning 1994 results in a growth of 1 per cent by the year 2000 in agricultural GDP, where­as when non‑agricultural sector is also liberalized, agricultural GDP grows by more than 3 per cent.

5)    Agricultural liberalization increases output of all agricul­tural commodities excepting coarse grains and `other foods', the sectors for which our estimates of the nominal protection rates were positive. However, when non‑agriculture is also liberalized, the outputs of these two sectors also increase.

6)    Liberalization leads to higher volume of exports of all agricultural goods, except coarse grains. Particularly, exports of wheat, rice, dairy products and non-food products could expand substantially.

7)    Prices of several agricultural sectors, which are disprotect­ed now, would rise with trade liberalization, while prices of industry and some agricultural sectors (coarse grain and other foods), which are protected now, would fall. If an export quota restriction is imposed in a sector and if the quota is binding, as in case of rice in our experiment, then its price could de­cline even when nominal disprotection is removed.

8)    If India were to export rice freely, the world market price of rice is likely to fall, given the thinness of the world mar­ket for rice. When we consider such effects, rice sector is not disprotected to the extent implied by the present world prices. In commodities where India has market power, the non-linearity in world prices in response to India's trade of the commodity should be accounted for. A correct measure of disprotection of rice would be the difference between present domestic price and the world market price that would prevail if India were to have free trade in rice.

9)    Given its market power in world rice trade, India could think of imposing tariff on rice exports. Our results indicate that this tariff would be socially optimal at a level such that rice exports are around 0.5 million tonnes.

10)  Agricultural liberalization by itself leads to decrease in poverty in both rural and urban areas. The marginal increase in terms of trade for agriculture and the  decline in prices of coarse grain and other food together result in different impacts on consumer price indices for different expenditure classes. The adverse income effect associated with it is more than adequately compensated by the gains to the poor as consumers due to fall in rice and coarse grain price.

11)  The rural rich are, however, worse off due to liberalization in terms of their equivalent income. As government loses tariff revenue due to liberalization it has to raise other taxes, which are stipulated to be only on non-agricultural income.

12)  If the speed of liberalization is slowed down, the beneficial impacts are postponed. However, by 2000 much of these are regained, though not fully.

13)  Retaining (imposition) of moderate protective tariffs as permissible under GATT have their own costs and should be imposed only after careful consideration of benefits and costs. It may be noted that our model does not account for potential benefits (if any), of dynamic improvement in efficiency due to (infant industry) protection. In any case, a case for protection has to be made before it is granted. This should be particularly so for capital goods.

14)  If liberalization results in additional inflow of foreign capital which is used for increasing investment then it helps in both accelerating growth and reducing adverse welfare effects.

15)  Removal of agricultural input subsidies along with liberalization, if the savings are used merely to provide tax relief, aggravates rural poverty though it does benefit urban population, as the tax rate goes down. Note that output prices are not affected by removal of subsidies when agriculture is liberalized (except rice price due to export quota).

16)  However, when the savings from input subsidy removal are used for increasing irrigation investment, the adverse effects on growth and distribution are more than offset resulting in higher growth and more equitable distribution. If agriculture is to be subsidized, it is better to subsidize investment in irrigation rather than use of current inputs.

17)  Targeted subsidies, either through targeted rationing or through employment schemes such as Employment Guarantee Scheme (EGS) or Jawahar Rozgar Yojana (JRY), financed through input subsidy removal can fully protect the poor, in fact, can substantially improve their welfare. Along with trade liberalization residual tariff as per GATT, subsidy removal, targeted safety nets, stepped up investment, if the expected additional foreign inflows materialize, produce a scenario which gives higher growth and substantial positive welfare gains for the poor, and this can be financed without raising taxes.