The economics of agricultural subsidies

PART ONE

1. Agricultural subsidies have been defined as a government induced change of relative prices of goods, services and factors of production in the agricultural sector. These agricultural price changes may result from a large number of different government measures varying from direct cash transfers to government trade policy. Workable definitions have been derived by a further classification into specific categories of agricultural subsidies.

2. Agricultural subsidies have been considered to be economically justified when under a system of laissez-faire public welfare decreases. They have been defended on three grounds, viz. the existence of domestic distortions, price instability, and the infant-industry argument.

3. Domestic distortions like monopoly power, pollution and other externalities, can theoretically be overcome by a subsidy policy in the direction opposite to the distortion.

4. Large price fluctuations of basic foodstuffs have a detrimental impact on the economy. Conversely, minor price variations are needed for profit maximization and for attuning supply to demand. The principal problem in designing an appropriate agricultural price policy is how to prevent large price fluctuations without undercutting the function of price as a market signal.

5. The infant-industry argument can not be adequately defended on theoretical grounds as developed by Mill, Bastable, Kemp and Meade. In accordance with modern equilibrium theory, protection has to be based on the existence of dynamic internal economics, i.e. the net positive diffuse impact on welfare stemming from the learning process of the industry under protection.

6. Subsidies do not only have the desired effects. Suboptimal behaviour by subsidized agents often leads to welfare losses. Also, welfare in non-subsidized sectors may decline because of a reallocation of resources towards the subsidized sectors.

7. In theory, the distorting and inter-sectoral impact of agricultural subsidies can best be ascertained by the use of macro-economic models. Total impact analysis, however, has not been very succesful so far, mainly because of the data required.

8. The estimation of welfare effects of agricultural subsidy programs, notably for developing countries, has mostly been based on partial equilibrium analysis. The estimation of elasticities and changes in economic surplus is of great analytical value, because it provides answers to ex-ante questions about who will benefit most from the subsidy, and how cost-effective the subsidy will be in raising farm output.

9. The inaccuracy in approximating welfare changes by changes in economic surplus is an increasing function of the proportion of individual income spent/earned with respect to the commodity or group of commodities in question, the magnitude of the (induced) price change, the interdependency of the goods that are added up in the surplus measure, and the proportion of the value of the commodities concerned as compared to total GNP.

10. Direct farm income subsidies are more geared to relieving the symptoms rather than the basic causes of income inequity. If, however, the major political objective is to guarantee farmers' incomes, direct farm income subsidies seem to be most appropriate since such a policy is less distorting and much cheaper than a policy that guarantees income by direct farm input or direct farm output subsidies.

11. Direct farm output subsidies, studied within a partial equilibrium framework, are either indirect money transfers (i.e. taxes) from consumers to farmers without additional welfare losses to consumers, or direct money transfers from consumers to farmers or government with additional welfare losses to consumers.

12. Generally, it will be uncertain whether farmers will benefit from farm output price stabilization, whereas it is probably true in most cases that consumers will suffer a welfare loss from such a policy (provided we are not dealing with major price fluctuations of basic foodstuffs).

13. It has been suggested that the establishment of active sales agencies and a thorough analysis of markets in developed countries by experts from developing countries may form a better means of boosting farm exports than the imposition of farm export subsidies.

14. When direct farm input subsidies are not accompanied by improved farm technology, the resulting shift in farm output supply will be nothing but parallel and likely to occur at net social cost.

15. Farm input utilization is determined by a large number of different factors affecting the farm input demand and supply side, whereas a farm input subsidy only affects a few of these factors, and to some degree in an unpredictable way.

PART TWO

16. The empirical estimation of welfare effects of agricultural subsidies is extremely difficult when a large number of subsidy programs is being implemented at the same time. Hence, an optimal determination of the subsidy mix in terms of cost-effectiveness in meeting the objectives is feasible only when the total number of support measures is limited. Another reason for starting on a small scale with agricultural subsidy programs is that most new subsidy programs suffer from administrative problems in their initial stages. These infant disease costs can be minimized by allowing subsidy programs to benefit as much as possible from the experience gained in earlier subsidy programs.

17. The determination of the optimal subsidy mix in a dynamic context requires adequate subsidy statistics. Therefore, it is advisable to allocate part of the subsidy udget to the collection and organisation of subsidy statistics.

18. The empirical study of farm labour distortions in Trinidad and Tobago has shown how a distinction between distorted and undistorted farm production could be made with the help of estimated production functions.

19. It has been argued that agricultural subsidy programs can be succesful only if farmers are maximizing their profits and hence respond logically to induced changes in relative prices. A practical method to test price-responsiveness, with the help of cross-section data only, is the normalized restricted profit function (NRPF) approach.

20. A serious limitation to the NRPF-approach is that this method cannot be applied when agricultural prices are fixed. It must be noted that if farm prices are fixed, an optimal response to subsidies is doubtful anyway.

21. Elasticities indirectly derived from the NRPF are more reliable than those that are directly obtained from single-equation production functions. Nevertheless, the signs and relative magnitudes of the various elasticities of either approach may match.

22. The empirical analysis shows that farm input subsidies had no impact on input utilization of crop farmers. Apart from monopoly power in farm input supply, four fundamental administrative defects causing this program failure were identified.

23. In the ex-ante analysis of a fertilizer subsidy it was estimated that the shadow price of saving foreign exchange by subsidizing locally produced fertilizer only, would have been relatively high.

24. The computed analysis of production value into a price and a volume component concerning temporary crops, has provided support for a price stabilization program beneficial to farmers.

25. The concept of guaranteed threshold prices has been introduced in the ex-ante analysis of farm output subsidies. A sensitivity analysis within a partial equilibrium framework has been employed to determine the relative cost-effectiveness of different anti-distortive guaranteed threshold price programs.