Ajay Chhibber and John Wilton
As the importance of structural adjustment programs has increased, more attention has been paid to the effects that macroeconomic policies have on the structure of incentives for production in different sectors of the economy. These effects need to be taken into account because distortions in relative prices created by inappropriate macroeconomic policies can often overwhelm the impact of sector- or commodity-specific policies. A narrow focus on project-level policies, such as crop-specific taxes or subsidies, may fail to take into account the broader policy-induced distortions that are often dominant in determining resource allocation and efficiency.
This article looks at how macroeconomic policy has affected the growth and the composition of trade in agriculture in selected developing countries. Macroeconomic policy is used here in a wide sense to denote those economic policies that play a predominant role in determining the intersectoral allocation of resources, though they are seldom viewed by policy makers as being directly concerned with agriculture. Typically, the net effect of macroeconomic policies on agriculture has been depressive, particularly for agricultural exports. In most cases this negative effect has compounded those of explicit measures to protect domestic manufacturing relative to agriculture. In some other cases recent reforms designed to reduce an explicit tax burden on agriculture have been offset by an increase in implicit taxation created by macroeconomic policies.
Trade policies have often been the primary source of policy-induced discrimination against agriculture. This has generally taken effect through tariffs and quotas that give heavy protection to import-competing manufacturing activities relative to agriculture. The table shows an index of the bias in protection to industry relative to agriculture for selected developing countries. A value of 1.0 indicates no discrimination against agriculture, and a value of less than one indicates that agriculture is disfavored relative to industry. Apart from the Republic of Korea, all countries in the table favored industry relative to agriculture—this bias was particularly marked in Colombia, Egypt, and Nigeria. While some domestic agricultural activities may also be protected, such as livestock products in Egypt or maize in Colombia, the general pattern of protection overwhelmingly favored nonagricultural activities.
Such policy bias often reflects a view that the protection of industry, in its early stages, is a necessary condition for economic development. There may be instances where protection of particular activities is socially profitable, for example where strategic or longer-run factors are perceived to be important. But these factors usually call for only temporary and moderate assistance. Many developing countries have imposed trade restrictions far beyond what is justified for these purposes.
How strongly trade policies can affect agriculture is illustrated in some recent case studies of trade regimes. In the Philippines, for example, a policy of heavy protection to industrial consumer goods was estimated to have reduced the prices of agricultural exports relative to those of protected goods by between 44 and 71 percent, depending on the category of import. Protection of import substitutes also affects the prices of protected goods relative to the prices of nontraded goods. The higher prices of protected goods encourage a shift in demand toward nontraded goods and less-protected traded goods. In the case of nontraded goods the higher demand generally causes prices to rise. The prices of less-protected traded goods are less likely to rise because they are determined by the prevailing international price and the exchange rate. This is particularly so for exports. In the Philippines, for example, the prices of agricultural exports (that is, nonprotected traded goods) were estimated to have fallen by 33 to 55 percent relative to the price of nontraded goods over the period 1950-80.
Similar policies in countries as diverse as Argentina and Nigeria, Chile and Zaire, and Colombia and Peru have to varying degrees reduced the incentives for nonprotected traded goods (usually agricultural exports) relative to those for protected and nontraded goods. Studies have estimated that, in these countries, a 10 percent increase in import tariffs reduced the prices of export commodities by 10 percent relative to those of protected commodities and by 4-9 percent relative to those of nontraded goods. Of course such policies discriminate against both traditional exports (generally agricultural commodities) and nontraditional exports (generally manufactured goods). But for nontraditional exports, at least part of the bias is often offset by export subsidies and special export promotion schemes.
Argentina provides another example of the impact that trade policy can have on agriculture. The growth rate of Argentina’s agricultural sector was significantly reduced through a combination of high import protection for manufactured goods, export taxation of agricultural commodities, and an inflexible exchange rate policy. Between 1965 and 1983 the annual agricultural growth rate averaged only 0.8 percent, compared with 1.9 percent in 1950-64 and about 2.6 percent before World War II. If this declining growth rate were the outcome of shifting comparative advantage in response to a fairly neutral incentive framework, it would not be of concern. But recent studies, designed to examine the combined effects of distortionary trade and exchange rate policies in Argentina’s agriculture, show that this was not the case. Over the period 1960-83, economic policies reduced real farm prices to an average of 38 percent below what they otherwise would have been. Had these policies not been followed, the studies indicate that, by 1983, agricultural output would have been 33 percent higher than it actually was, and that GDP growth would have been stronger.
|In the 1960s|
|Republic of Korea||1968||1.18|
|In the 1970s and1980s|
|Republic of Korea2||1982||1.36|
One plus the effective rate of protection for agriculture divided by one plus the effective rate of protection for manufacturing.
Refers to primary sector.
One plus the effective rate of protection for agriculture divided by one plus the effective rate of protection for manufacturing.
Refers to primary sector.
The bias against agriculture caused by trade policies has often been exacerbated in developing countries by expansionary monetary and fiscal policies and an inflexible exchange rate regime. In many developing countries excessive increases in the money supply derive from inappropriate fiscal policies. With restricted—or, in some cases, strictly rationed—access to foreign capital and limited domestic private savings, public sector deficits can only be financed by printing money. However, when domestic inflation increases, nominal exchange rates are often prevented from adjusting by the difference between domestic and international inflation. To manage the balance of payments, ad hoc foreign exchange and trade controls are introduced instead. As a result, nominal exchange rates become overvalued, and the index of the real exchange rate (defined as the foreign price index divided by the domestic price index) declines.
The resulting nominal overvaluation of the currency has important implications for producers of traded goods—and particularly for producers of agricultural goods. It lowers returns to agricultural exports and makes agricultural imports, generally food items, cheaper relative to domestic substitutes. Moreover, with expansionary fiscal and monetary policies the level of effective protection to domestic industry relative to agriculture increases. This happens for two reasons. First, as external imbalances worsen, ad hoc trade controls, particularly on industrial goods, are often introduced or intensified. Second, for goods subject to quotas or other nontariff restraints, the extent to which domestic prices exceed world prices is strongly influenced by fiscal and monetary policies. If excess credit is created, the subsequent increase in demand for the goods whose importation is restricted will be reflected primarily in their prices, particularly where domestic supply is inelastic. Unless there is a change in trade policy this will cause the domestic price to rise even further above the world price. The net result is lower relative prices for the unprotected agricultural sector.
In the 1970s and 1980s, overvaluation of the nominal exchange rate increased sharply in many African countries. In sub-Saharan African countries as a group, the real exchange rate declined by 31 percent between 1969-71 and 1981-83 (measured against the US dollar and using the US consumption deflator). In the rest of the developing world this real exchange rate index showed no significant change. African agricultural exporters would have received prices 31 percent higher had the nominal exchange rate been devalued by the difference between domestic and international inflation over that period. At the same time, consumers of agricultural imports would have paid 31 percent higher prices in local currency. This would have both reduced imported food consumption and raised output prices for domestic food producers.
Changes in real exchange rates have significantly altered the profitability of farming relative to other activities in Africa. For example, between 1969-71 and 1981-83 the price of cereals in Africa increased by 9 percent at the official exchange rate. Had the nominal exchange rate fully reflected domestic inflation, the prices for the same cereals would have increased by about 50 percent. Similarly, while the prices of export crops at the official exchange rate actually declined by 27 percent, they would have increased by 2 percent had the exchange rate been adjusted appropriately.
The changes in relative prices arising from overvaluation of the exchange rate have led to large losses in world export markets for many African countries, even in those export crops in which the continent is presumed to be an efficient producer. In palm oil, for instance, in the 1960s African producers exported as much as their Asian competitors, but by the 1980s Indonesia, Malaysia, and Singapore supplied over 90 percent of the world market. Cocoa, another traditional African commodity, witnessed the same reversal, with Brazil and Malaysia taking a larger share of the cocoa market. Côte d’lvoire, which followed more balanced macroeconomic and sectoral policies, also increased its relative share of the cocoa market by taking over many of Ghana’s markets.
Real exchange rates and agricultural exports, 1961-83
Source: World Bank. World Development fleporf 1988.
Another disturbing effect of Africa’s overvalued nominal exchange rates has been the rising dependence on food imports. The region’s food grain imports have quadrupled over the last 15 years, while foreign exchange earnings from agricultural exports have fallen. Import growth has been faster in wheat and rice than in other cereals, not because these two commodities are traditional African crops but because government policies have made them cheaper and because of their convenience and prestige value to consumers.
Within Africa, Ghana, Nigeria, and Tanzania show the most significant decline in the index of their real exchange rates between 1969-71 and 1981-83—about tenfold in Ghana, twofold to threefold in Nigeria, and twofold in Tanzania. These movements can be compared with those in Brazil and Chile, two countries where the real exchange rate appreciated (see charts). The charts show a marked positive correlation between real exchange rate movements and agricultural exports. A recent econometric study shows this relationship to be more generally true. It calculates that for every percentage point decline in the real exchange rate, agricultural exports declined on average by 0.6-0.8 percentage points in all developing countries and by 1 percentage point in Africa.
The heavy implicit taxation of African farmers through overvalued currencies, coupled with explicit crop-specific taxes and the high marketing margins of parastatal agencies, has led to the widespread emergence of parallel markets. Up until the recent reforms, Ghana provided a classic example with respect to its cocoa markets, especially when exchange controls began to cause a heavy overvaluation of its exchange rate. Ghanaian cocoa has been smuggled into Côte d’lvoire, causing heavy losses in official foreign exchange earnings and government revenue. Similarly, Sierra Leone incurred heavy losses when the overvaluation of its currency between 1982 and 1984 led to the smuggling of coffee, cocoa, shrimp, rice, and palm kernels into Liberia. In Tanzania, the parallel market in food crops became so profitable, relative to those export crops which had to be sold at controlled prices to parastatals, that farmers switched from cotton, tobacco, and pyrethrum into the more profitable maize.
The rise of these parallel markets provides at least two important lessons. First, if incentives are distorted enough, farmers will incur the costs of trading in an illegal market or plant crops which are taxed at a lower rate. Under either alternative, both the income of farmers and the growth of the agricultural economy will be lower than it would with less distorted prices. Second, official foreign exchange earnings and government revenue will fall significantly if the costs to farmers of trading in parallel markets are lower than the policy-induced direct and indirect costs of continuing to trade in formal markets. Some governments have reacted to this situation by attempting to increase their control over foreign exchange transactions and border trade, and by printing more money to make up for the lost tax revenue. Such actions have, however, distorted production incentives further and increased inflation, slowing growth even more.
Public expenditure, investment
Another aspect of macroeconomic policy that is central to the performance of the agricultural sector is the allocation of government expenditure. In many developing countries, public investment accounts for a major share of total investment, and influences the allocation of private investment. Countries that tax agriculture heavily through trade and exchange rate policies often discriminate against agriculture in the allocation of government expenditure as well. As a result, they further discourage resources from being invested in agriculture. Conversely, those countries which have followed balanced pricing policies, through more competitive exchange rates and lower direct taxation of agriculture, have channeled considerable public expenditure to necessary public goods and services such as research and extension, irrigation, electricity, transport, health services, and rural education. Not surprisingly, countries in the latter group have outperformed those in the former group.
The constructive interaction between incentives and expenditure programs, discussed in the accompanying box on Indonesia and Nigeria, has also been evident in China and India. In China recent reductions in the direct and indirect taxation of agriculture and the removal of controls on area and production have, in combination with previous public investments, produced spectacular productivity increases. In India, too, the last two decades have witnessed a marked turnaround in the food grain sector, brought about by a combination of large investments in irrigation, introduction of high-yielding grain varieties, and a steady increase in domestic prices in line with international prices.
The costs of adopting policies that discriminate against agriculture are not borne by that sector alone. Ironically it is those countries such as the Republic of Korea, which did not discriminate against agriculture relative to industry, that experienced very high industrial growth rates. With some exceptions, such as the oil and mineral exporters, countries with low agricultural growth have experienced low industrial growth, and countries with high agricultural growth have experienced high industrial growth. During the period 1973-84 Benin, China, Côte d’lvoire, and Thailand registered high growth in both sectors. The low-growth syndrome has been exemplified by many countries in sub-Saharan Africa and some in Latin America and South Asia. Countries such as Burkina Faso, India, and Turkey are in the middle, attaining moderate growth in both sectors.
Effects of policy on agricultural performance
Indonesia and Nigeria, two middle-income oil exporters, provide an interesting contrast: both economies benefited from the oil boom in the 1970s, but the effect on agriculture was markedly different. This was partly because of differences in macroeconomic policy, including public investment expenditure.
In Nigeria, the oil boom led to a severe disruption of the agricultural economy and a large exodus of labor to the cities. Between 1970 and 1982, annual production of Nigeria’s major cash crops—cocoa, rubber, cotton, and groundnuts—fell by 43, 29, 65, and 64 percent, respectively. The share of agricultural imports in total imports increased from about 3 percent in the late 1960s to about 7 percent in the early 1980s with per capita food imports increasing very rapidly.
Indonesia, perhaps alone among the oil exporting developing countries with large populations, succeeded in avoiding serious economic disruptions in its agricultural sector. Agricultural growth slowed in the mid-1970s but recovered to previous levels by the late 1970s. Rice production grew by 4.2 percent a year from 1968 to 1978 and by 6.7 percent a year from 1978 to 1984, largely because of rapid increases in rice yields. The share of agricultural imports in total imports remained unchanged at about 1 percent; and the share of Indonesia’s agricultural exports rose as a proportion of both developing country agricultural exports and world agricultural exports—at rates of 2 percent a year and 0.5 percent a year, respectively, during 1965-83. Nigeria’s corresponding export market shares over the same period declined at rates of – 5.7 percent and – 7.1 percent a year, respectively.
Differences in their monetary, fiscal, and exchange rate policies partly explain the divergent paths of these two economies. Initially both economies reacted to the first oil price increase in a similar way. The improvement in their terms of trade and the large inflow of foreign capital that was associated with oil earnings and an increased capacity to borrow led to a decline in the real exchange rate by about 30 percent between 1970-72 and 1974-78. To the extent that this change in their real exchange rates reflected a movement to a new and sustainable equilibrium, it was appropriate. After 1978, however, the macroeconomic management of the two economies differed significantly. Indonesia avoided a further marked decline in its real exchange rate through a combination of a less expansionary monetary and fiscal policy, restrained foreign borrowing, and a floating exchange rate. Ironically, the scare following the Pertamina oil crisis may have helped policy makers by providing a warning of the dangers of overborrowing on the basis of future oil earnings. Between November 1978 and March 1983, the nominal value of the rupiah was devalued against the dollar by over 100 percent to reflect differences in their respective inflation rates. Nigeria, by contrast, followed an expansionary fiscal and monetary policy and resisted any devaluation of the naira despite high domestic inflation rates. Nigeria also borrowed heavily against future oil earnings, increasing the pressure on the exchange rate caused by large inflows of foreign capital. As a result, by 1982 the real exchange rate had declined by over 100 percent from its 1970-72 value.
Another important difference between the two countries was in the allocation of public expenditure to agriculture. In Nigeria the government virtually neglected the agricultural sector. The bulk of its increased public expenditure went into providing universal primary education, transport, a major steel complex, and a new federal capital. As a result, labor and other resources were drawn away from agriculture into other sectors. Indonesia, on the other hand, pursued a more balanced expenditure policy between agriculture and other sectors. Within agriculture the government invested heavily in irrigation and other infrastructure, and provided support in the areas of research and extension and fertilizer. Domestic producer prices were also allowed to adjust via a flexible exchange rate to reflect movements in world prices. In Nigeria, however, the policy was to hold down food prices in an attempt to supply cheaper food for the rapidly expanding urban population.
Historical experience confirms the critical role that a dynamic agricultural sector plays in supporting overall growth. In the nineteenth and early twentieth century in England and Japan, considerable increases in farm productivity and output freed domestic resources and provided an important market for industrial goods. In Japan, substantial voluntary transfers of capital and labor from agriculture to the rest of the economy were a major contributor to industrial development. The complementarity is also evident from recent studies at the macro and micro level on some developing countries. In India, Malaysia, and the Philippines more rapid agricultural growth has strong positive effects on the rest of the economy by raising demand for consumer goods and by increasing household savings and government revenue.
The economy-wide distortions that result from the type of in appropriate macroeconomic policies discussed above can therefore lead to a persistent and pervasive misallocation of resources. In many developing countries the net effect of macroeconomic policies has been to create a bias against agriculture, particularly agricultural exports. Where policy makers have adopted an explicit policy objective to favor domestic manufacturing relative to agriculture, the discrimination against agriculture has been overwhelming, with heavy costs to farmers, rural labor, and ultimately to the whole economy.
This article draws on some of the background material prepared for Chapter 4 of the World Development Report 1986 by Domingo Cavallo, Ajay Chhibber, and Yair Mundlak, as well as published material by Alberto Valdez. A full bibliographical note on the topics covered in this article can be referred to in the WDR itself.