Chapter 14. Trade and related economic reforms in Latin America^[244]

14.1 Introduction

This chapter reviews the trade policy reform process in the last two decades relevant to agriculture in Latin America and the effects of those reforms on production, trade, household incomes, and food security^[245]. Although trade liberalization was a prominent element in the reorientation of the economic policies of many Latin American countries, it was only part of the story. Trade reforms were made in the context of larger structural reforms, macroeconomic adjustments, deregulation and privatization, and a general redefinition of the role of the state in the economy. The final outcome of the reforms not only reflects what happened on the trade side, but also reflects both the consequences of other reforms and the initial conditions in which countries found themselves when introducing broader reform programmes.

14.2 The specificity of Latin America

In general, Latin American countries reformed early. At the beginning of the shift in economic policy, most were suffering from both low growth and serious macroeconomic disequilibria - high inflation, fiscal deficits, current account deficits, and financial sectors in critical trouble associated in part with the foreign debt crisis of the early 1980s. Not only were economic reforms pushed by an ideological impulse toward smaller governments, but agricultural trade reforms were implemented in this context of fiscal deficits and large foreign debt. Such constraints induced significant budget cuts generally, and - more relevantly for agriculture - specific cuts in subsidized credit, marketing programmes, and infrastructure. As well as trade liberalization, the macroeconomic constraint on government budgets has affected agricultural performance and farmers, particularly lower-income farmers, over the last two decades.

As part of a general reorientation of governments toward freer markets in order to solve a broad set of problems, the impetus for trade reforms in most counties began before the signing of the Uruguay Round Agreement. The removal of quantitative restrictions and the process of tariffication, which was an integral part of the UR, had already been adopted as a general principle in most of Latin America^[246]. Even before the UR, Latin American countries were, relative to the rest of the world, among those with the most significant reductions in agricultural protection^[247].

Most LAC countries, therefore, emerged from the UR with no mandated policy changes^[248] and ironically they did not get any concessions for their early moves toward freer trade. Most notably, countries that had previously tariffied were left without the benefits of special safeguards. Nevertheless, the impulse to trade liberalization in many countries was reinforced by that Agreement. The later emergence of numerous regional trade agreements (notably MERCOSUR, the Andean Group, the Central American common market, the Caribbean Group, Mexico entering NAFTA) and bilateral accords (Chile with Mexico, Canada and several countries) also contributed to a reduction in the levels of protection in many sectors, including agriculture.

Most Latin American countries are not agrarian economies and are land abundant. They are predominantly middle-income, with the majority of their populations engaged in non-agricultural activities, and with most having relatively low rural population densities. Structural transformations in most countries have already reduced the proportion of the labour force in rural areas to less than one third. In their important study on agrarian economies, Tomich, Kilby and Johnston^[249] conclude that only three Latin American countries - Haiti, Honduras and Guatemala - can be considered agrarian countries or CARLs (countries with abundant rural labour). To put this in context, there are 58 CARLs worldwide.

The region can be characterized as having a small number of commercial farms - producing the bulk of agricultural production - coexisting with a much larger number of small farms. But the hacienda image is a thing of the past. Commercial farms are not particularly large compared to Canada, Australia and the western United States, although they are large compared to most of Asia, Africa and the Middle East. Another important characteristic of the region is that there is a high incidence of hired labour. For both farm and off-farm income, rural household incomes depend less on land ownership than on employment opportunities.

Differing from most other developing regions, Latin America has a smaller share of rural workers in the labour market and of agriculture in the total economy. The rural populations of some small economies (El Salvador, Guatemala, Honduras, Paraguay, Jamaica) still represent more than 45 percent of the total, but in the case of the larger economies (Argentina, Brazil, Chile, Colombia, Mexico, Peru, Venezuela) - representing at least two-thirds of the region’s total population - rural populations make up less than 30 percent of the total and agriculture less than a fourth of GDP^[250]. In general, the structural characteristics of Latin America should lead us to emphasize the importance of liberalization on economy-wide effects and consumers’ welfare, and not merely on the impacts of trade policy changes on agriculture and the rural population.

Important though it is, the impact of agricultural growth in middle-income countries is likely to be less than in low-income countries. In contrast to least developed, agrarian countries, in most of Latin America it is hard to defend the argument that it is agricultural growth that reduces poverty (as argued in Chapter 7). But agricultural growth will unquestionably contribute towards reducing rural poverty. (Although the incidence of poverty is greater in rural areas, the absolute number of poor is small compared to urban areas.)

Heterogeneity with respect to trade liberalization

Latin America is heterogeneous not only in the timing and depth of reforms undertaken in individual countries (Table 14.1), but in terms of net trade positions, which influence the consequences of those reforms. Three different groups can be identified within the region. The first is the South American Block, represented in the Cairns Group, which was the group most oriented towards liberalized trade. A second group is the Caribbean countries, which favoured a slower pace of trade liberalization. The third group is Central America and Mexico, which took a position somewhat between the first two, but closer to the Cairns Group.

Table 14.1 Degree and period of trade liberalization, selected Latin American countries, 1980s - 1990s

^*Colombia experienced a significant turn toward free markets in the early 1990s, but by 1995 had almost retreated to prior levels of protectionism for the agricultural sector. Source: p. 151, Spoor, 2001 op cit.

With respect to net trade positions, McCalla and Valdés^[251] conclude that 13 out of 23 countries were net importers of both food and agricultural products; and 17 were net importers of food. More revealingly in terms of trade policy, is that 10 of the 13 net food and agricultural importers were small Caribbean island economies. The analysis helps dispel the misperception that the region as a whole is oriented purely toward agricultural exports. It is true that export sectors are a large component of agriculture in many countries, but the most sensitive domestic trade policy debates that remain typically centre on import-competing sectors.

14.3 Trade reforms in Latin America

With respect to the timing of reforms, one can identify early reformers and late reformers (Table 14.2). The experience of the early reformers, especially Chile and then Mexico, provided lessons that influenced at least in part the reform efforts in other countries. The most important elements of policy reforms in the region relating to agricultural trade over the last two decade have been:

• the replacement of most quantitative restrictions on imports with tariff;

• the reduction in both the level and dispersion of tariffs;

• the removal of all export taxes, quotas and licenses;

• a reduction in the importance of food self-sufficiency as a policy objective;

• a reduction or elimination of state trading;

• the elimination of domestic price controls and the gradual removal of state procurement programmes.

Table 14.2 Agricultural performance of early and later reformers, Latin America, 1980-1999

¹ Agricultural GDP includes crops, livestock, forestry and fisheries.

² Growth rates of agricultural exports include only crops and livestock.

The averages for later reformers use the Honduran data excluding 1999.

Source: Elaborated by the authors from data of the Inter-American Development Bank (these may differ from FAO statistics.)

In broad terms, the primary objective of trade liberalization programmes in Latin America was to reverse the negative consequences of protectionism, especially its inherent anti-export bias^[252]. It was not merely a question of eliminating explicit export taxes, but also of reducing the implicit taxation resulting from distorted relative prices that favoured importables and, indirectly, non-tradables. International trade was viewed as a potential engine of growth, consistent with newly-popular models of endogenous growth.

The introduction of an open economy was expected to affect economic growth through technological progress and productivity growth. In turn, higher potential productivity would induce greater investments and lead to further growth. The impact of trade liberalization would of course be conditional on progress in reforms on other fronts: deregulation, privatization, macroeconomic stability. The disappointments with some trade liberalization programmes (Colombia is an instructive example) can be linked to failures on these fronts, such as exchange rate appreciation, and the lack of advances in deregulation and privatizations.

The principal policy mechanism adopted to achieve this objective was a reduction in average protection economy-wide, not merely in agriculture. Once exchange rates were adjusted and quantitative restrictions reduced, the next goal was to adjust tariffs so that their levels and range (or dispersion) decreased. In almost all Latin American countries, the non-tariff barriers (NTBs) for all goods that existed in the 1980s were cut or eliminated during the initial phases of reform. Of the 14 countries surveyed by Edwards^[253], six had effectively eliminated NTBs, and the highest coverage of NTBs was for Mexico (20 percent of import positions). Likewise, tariff reductions were indeed substantial by the early 1990s compared to the 1980s (Table 14.3). While no country’s tariff protection was below 20 percent in 1985, and the majority’s was above 30 percent, by the early 1990s only Brazil had tariff protection slightly above 20 percent, and the majority had protection rates of 16 percent or below. In the specific case of agriculture, although the coverage of non-tariff barriers did fall considerably, in some countries it nevertheless remained for a time more significant than in other sectors^[254]. This can be attributed in part due to the desire of governments to protect their farm sectors from world price fluctuations and to counteract export subsidies.

Table 14.3 Changes in tariff protection and non-tariff barrier coverage, Latin America, 1985-1992 (percent)

Note: Average tariff protection is average total charges (tariffs plus paratariffs), unweighted. Average non-tariff barrier coverage is also unweighted.

Source: Table 5.2 in Edwards 1993, based on World Bank data and others.

Prior to the reforms, the trade regime in the region was that of a strong import substitution and an anti-export bias^[255]. There was a sharp contrast between import-competing activities and export-oriented sectors. Importables were protected: the average nominal protection rate (NPR) for the decade 1985-1995 was 18.7 percent. By contrast, exportables were overall taxed: the average NPR during 1985-1995 was -7.7 percent. For some countries, there were significant policy-induced transfers of income out of the farm sector. For the period 1985-1990, prior to the reforms, transfers out of agriculture amounted to between 12 and 23 percent of agricultural GDP in Argentina, the Dominican Republic, Ecuador and Uruguay. Brazil and Paraguay extracted only small amounts from agriculture. Those input subsidies and non-price transfers that did exist compensated little or not at all for these transfers. During the same years, Chile (which had reformed much earlier, in the mid 1970s) and Colombia had positive transfers, subsidizing their agricultural sectors, from 5 percent to 8 percent of agricultural GDP.

14.4 The consequences of reforms for agriculture

Following trade reforms, the direct taxation of exportables declined in most countries. Between the mid-1980s and the early 1990s, the average level of taxation, measured as NPR on exportables, fell from 11.8 percent to 4.8 percent. Some countries, such as Ecuador, continued to tax agricultural exports, and others (Brazil and Colombia) continued minor subsidies on exportables.

From the perspective of incentives and resource allocation, however, the effective rate of protection (EPR) is more relevant than the nominal. In some countries, the patterns of effective protection in the initial years of reforms were still sizable and negative for exportables, averaging minus6.9 percent in 1993, compared to a positive 30 percent for importables. This can be explained in part by an asymmetry in the effects of reforms. Although direct export taxes were removed, intermediate inputs were often still protected. Furthermore, some countries continued to provide a wide and selective pattern of boarder protection. The Dominican Republic, for example, in 1993 had an EPR of minus7.1 percent for sugar and simultaneously had an EPR of 129 percent for rice, 270 percent for beans and 223 percent for maize^[256]. At least in the early phase of reform, the bias against exportables, although reduced, nevertheless continued. Although for the period 1995-2002 no hard evidence is available, the overall trend in the region has been towards the complete elimination of export taxes and a significant reduction of intermediate input tariffs. As a consequence, the effective rates of protection of export products are now more favourable. Moreover, the proliferation of bilateral and regional trade agreements (e.g. MERCOSUR, Andean Group and others) has perforated tariff barriers and reduced the operative average tariffs below their MFN rates.

Trade policy is only part of the story of incentives facing farmers. Other factors include the evolution of the exchange rate, interest rates and world prices. In the early phase of reforms in several countries, unfortunately for the political credibility of reformists, trade and agricultural policy reforms were taking place in an environment of declining profitability of farming. Real farm prices of tradables in domestic markets decreased, leading to strong pressures by farm lobbies for renewed protection and subsidies. In some cases, most notably in Colombia, the situation was such that there was political backsliding by the mid-1990s, a reversal of the reformist trend and a return to protectionism.

Between 1986 and 1995, in seven of the eight Latin American countries studied by Valdés all major domestic farm prices declined in real terms. Worse, during the early reform years of 1990-1993, real farm prices declined more than in the pre-reform period. Only in the case of Chile did the real price of exportables increase, due to a very different mix of export products. In this early phase of reform, real domestic farm prices for exportables fell between 22 percent and 47 percent for Ecuador, Dominican Republic and Colombia, and by over 50 percent for Argentina, Brazil and Uruguay. For the years 1994-1995, an increase in real farm prices took place in Brazil, Colombia, Dominican Republic and Uruguay; but a continued decline in Argentina and Ecuador. For these years Chile was also affected negatively. With respect to importables, for three of five countries,^[257] real farm prices also declined during 1990-1993, and continued to decline for 1994-1995.

The main factor that explains these price declines was the appreciation of exchange rates of the early 1990s, a phenomenon that was amplified by tariff reductions and in some cases a fall in border prices. The decomposition analysis for the sample of countries in the Valdés study shows a substantial decline in the real exchange rate in all countries, coinciding with the initiation of the liberalization programmes. In several cases, the real exchange rate appreciation reinforced the decline in domestic prices, particularly between 1990 and 1993. While border prices recovered somewhat between 1994 and 1995, the continued exchange rate appreciation offset the gains. The case of Argentina is instructive: the real domestic price of beef, an exportable, fell 60 percent between 1990 and 1993, despite the increase in border prices of 14 percent and the removal of export taxes.

In overall terms the principal results of trade liberalization were the change in relative incentives affecting the production of exportables and importables and the significant increase in the output of exportables. There is a positive correlation between reforms and the growth rates of the agricultural sector overall and of exports. Following reform, all countries except Brazil, Colombia, Ecuador and Honduras experienced an increase in growth rates of agricultural GDP during the 1990s^[258].

In the case of exports, the impact of the reforms is even more striking, most countries having experienced an increase in the growth rates of exports. More broadly, there was a consequent general growth in both exports and imports of agricultural products, although the net effect on the agricultural trade balance varied across countries. Early reformers, excluding the very early reformer Chile, expanded the value of agricultural exports at an average yearly rate of 11 percent during 1990s, compared to 3.4 percent in the previous decade. Later reformers expanded agricultural exports by 6 percent per annum in the 1990s, compared to 1.6 percent during the previous decade. Chile, having reformed in the mid-1970s, enjoyed rates of agricultural export growth of over 11 percent during the two decades following the initiation of its market-oriented policies.

Not only did the total value of exports expand, but trade liberalization appears to have encouraged a diversification of export products. There was a notable growth in the export of fruits in Chile, Costa Rica, Mexico, Saint Lucia and Dominica, Brazil and Panama. While the economic significance of traditional, tropical fruit exports (most obviously bananas in Ecuador, Costa Rica and Panama) continued, liberalization spurred an increase in non-traditional crops (e.g., soybeans in Brazil and Argentina), and induced an experimentation in products, many of which have been successful in export markets (e.g. asparagus in Peru, tomatoes and mangos in Mexico)^[259].

Moreover, the considerable expansion of perishable exports occurred in an environment of many simultaneous changes attributable to economy-wide reforms: modernization of ports, deregulation and privatization of telecommunications, domestic and foreign direct investments in processing and marketing channels, the adoption of technologies available elsewhere for processing, storage and shipment. One example is the adoption of technologies permitting the profitable storage of fresh fruit on board ships.^[260] It would be difficult to isolate the partial effects of trade liberalization alone from the myriad impacts on the supply chain of agricultural products that resulted from general economic reforms.

Nevertheless, reforms in Latin America generally led to an increase in the tradability of agriculture (Tables 14.4 and 14.5). The ratio of agricultural trade, both exports and imports, to total agricultural GDP averages 87 percent for 18 of the region’s countries in 1996, compared to 74 percent in 1990. Some notable cases of increases in tradability are Argentina, Costa Rica (reaching 130 percent), Chile, Ecuador, El Salvador, Mexico and Venezuela. For all countries except Brazil and Peru, tradability is around or above 50 percent. Brazil’s low tradability index is understandable given the size of its domestic market. Peru’s low tradability index suggests that the country’s agriculture is relatively more oriented toward non-tradables. With the increase in tradability, agricultural sectors become more exposed to world price and exchange rate changes, and to shifts in foreign market conditions generally.

Table 14.4 Tradability indices* in agriculture, Latin America and Caribbean, 1990 and 1996

*Value of agricultural exports plus agricultural imports divided by agricultural GDP.
Source: J. Quiroz (2000), based on FAO and World Bank data.

Table 14.5 Tradability indices* in agriculture in low-income countries, Latin America and Caribbean, 1980-1996

*Value of agricultural exports plus agricultural imports divided by agricultural GDP.
Source: Table C-2 in Mathews and Trueblood (1999), based on World Bank’s World Development Indicators.

Differential impacts within agriculture

Within the agricultural sector of each country, reforms affected sub-sectors differently. To assess the diverse effects of trade reforms within agriculture it is useful to distinguish between the effects on producers of exports, producers of import-competing goods, and producers of home or non-tradable goods (often predominantly small farmers). It is also necessary to analyse the effects according to farm size and geographic region, characteristics which are not independent of the trade-orientation of the products produced by farmers. At least three well-documented cases (for Brazil^[261], for Colombia^[262], and for Nicaragua^[263]) show that policy reforms affected producers differently according to these characteristics. Moreover, the effects varied according to time period.

Producers of exports gained generally, as well as wage-earners in agriculture and processing, because export products tended to be the most labour-intensive sub-sector. All else being equal, those farmers who could take advantage of the expanding export sector were those with the wherewithal to make investments and to face the price risks associated with export markets. These were mainly larger, commercial operations. In Chile, only a minority of small farmers could participate as producers in the expanding export growth of fruits and vegetables^[264], although as households they participated as wage earners, especially in post-harvest activities (e.g. in packing houses).

Import-competing producers probably lost in the short-run, but their long-run welfare depends on their capacity to increase productivity and/or change cropping patterns. In some regions, farmers had little flexibility to adjust their productivity and output mix, and as a consequence this subset of farmers was and remains critical of trade liberalization. The problem is not so much one of poverty but of potential profitability within the freer-trade environment. Worth noting is the case of ejidatarios in central Mexico who were harmed by the reduced protection of traditional grains resulting from NAFTA. In response, the Mexican Government instituted a compensation scheme (Pro Campo) that provides decoupled income transfers to producers. By contrast, milk producers in Chile, who traditionally competed with imports (and who were protected prior to liberalization), managed to take advantage of the modernization of the economy in general and of the availability of new dairy technologies and genetic lines in particular. The sector grew in efficiency; production expanded and has now become an export sector - but those who could take advantage of the new environment were generally larger farm operators.

The third group - non-tradable producers - is less directly affected by trade reform, although it may be harmed indirectly by consumers switching to lower-priced importables, or may be benefited indirectly from the higher price of exportables. Small farmers tend to be producers of non-tradables, and the members of the household tend to be relatively more involved in rural non-farm labour. As in the case of Brazil, in Colombia, where the relationship between reforms and the non-tradable sector has been analysed, small farm households benefited during the reform process mainly due to increased employment opportunities that became available in the rural non-farm economy^[265]. Whether or not greater employment was caused directly by trade liberalization is unclear.

Commercial farms, whatever their pre-reform product orientation, tended to be more technologically advanced and more flexible in terms of adapting to changing market conditions. They also tended to use relatively more purchased, imported inputs, the prices of which fell with trade liberalization. The across-the-board reduction and elimination of government subsidies on input use also increased the relative advantage of larger farms. With the change in the market environment, there was also a trend toward lessening the impacts of economic reforms through government-supported, targeted extension and credit programmes for small farmers, and through targeted social programmes for the poor in rural areas, including farmers (for example PRONAF in Brazil and INDAP in Chile for extension and credit, and PROGRESA in Mexico, FONCODES in Peru, IRD in Nicaragua, and Previdencia Rural in Brazil in the case of social programmes.)

Brazil: a case study of the effects of trade liberalization and other reforms

Evidence from Brazil, where recent good disaggregated data exist on the distribution of production across types of farms, clearly shows the manner in which reforms were to the relative disadvantage of smaller, low-technology farmers^[266]. Both changes in government policy and market conditions reinforced in three ways the advantages of technologically-advance commercial producers. First, technical assistance appears to have benefited to a greater degree larger farmers; especially those oriented towards import substitution.

Second, the commercial farmer used purchased inputs more intensively than did small farmers. Because input prices fell after reforms more rapidly than output prices, costs for larger commercial farms fell by a greater degree than revenues, and proportionally more than for smaller farms. In fact, per-hectare returns were negative for low-technology, small and subsistence farmers, who used no purchased modern inputs, and who experienced falling output prices without seeing any compensation in terms of lower costs.

Third, changes in the processing and market sector resulted in increased quality standards and the development of sophisticated systems of production financing and marketing. Chapter 10 presents the effects of supermarket development. To the extent that larger commercial enterprises were more apt to adapt to the new system of financing and the new emphasis on quality control, they had a relative advantage with respect to smaller farms. The two newly introduced factors of private financing and enhanced demand for quality each acted to compound the effect of the other. As the Government reduced subsidized credit, capital-intensive farmers were able to make greater use of the private sector - and were better prepared to use self-financing - for investments to meet the increased demand for higher quality products. Having entrance into marketing channels for higher quality products brought the benefit of having access to a more sophisticated production financing system.

The shifting concentration of production in Brazil is illustrated in Table 14.6. The share of the total value of production has been declining for smaller size farms (100 hectares or less) and increasing for larger size farms. In 1970 smaller size farms produced 57.8 percent of total production, and the largest size farms (over 1 000 hectares) produced 12.6 percent. By 1985 the contribution of smaller farmer had fallen below half of total production. By 1995 the largest farms had increased their proportion of total output by two-thirds, reaching a 21 percent share.

Table 14.6 Percentage share of production value by farm size, Brazil, 1970-1995

Source: World Bank, Rural Poverty Reduction in Brazil: Towards an Integrated Strategy, Volume II, April, 2001.

Some of the changes observed in Brazil were almost certainly experienced in other countries in the region. In Chile, especially as the second phase of reforms unfolded in the mid-1980s, the demand by commercial, export-oriented enterprises for new varieties and for more sophisticated production techniques and financing was met by both the public and private sectors. To some extent, this was balanced by the emphasis on the targeting of public technical and financial assistance toward smaller and poorer farmers. But in relative terms, because private sector participation increased rapidly, the technological and financial support from all sources for commercial farms grew faster for larger farmers than for smaller.

Exchange rates can also have different effects across farm sizes

A common feature in Latin America during the early phases of the reforms was the appreciation of exchanges rates. Changes in exchange rates can have differing effects across farm sizes, and therefore longer-term changes in exchange rates can affect the size distribution of the agricultural sector. Considering both a direct effect through the price of tradable goods, and an opposing but smaller, indirect effect through changes in the CPI, Lopez and Romano^[267] find that a simulated 40 percent nominal devaluation increases real revenues by 13.4 percent for large farmers (50-2 000 ha), by 5.4 percent for small farmers (2-10 ha), and by 2.6 percent for very small farmers (minifundia, less than 2 hectares). At least in the case of Brazil, smaller farmers gain less with devaluation because their revenues are skewed toward non-traded goods and traded goods make up a relatively larger share of their total consumption. Larger farmers produce more traded goods, and non-traded goods (such as services, housing, education and transportation) make up a relatively larger share of their total consumption.

Real income effects on urban and rural consumers

It is well known that low-income households, urban and rural, spend a large proportion of their incomes on food. To the extent that trade liberalization lowers food prices, household income of the poor will increase in real terms. Certainly low-income consumers (farmers are consumers too) benefited from trade liberalization as lower protection reduced the price of food relative to wage rates. There is clear evidence of this in Chile, Brazil and Nicaragua^[268]. In the particular case of low-income farmers, there are fewer studies, but one recent study shows that importables weight heavily the consumption baskets in the real income of small farmers’ households in the northeast of Brazil.

14.5 The effects of trade reforms on food security

Chapter 2 describes the different concepts of food security. Here, food security is taken to be the ability of food-deficit countries, regions, or households to meet target levels of consumption on a year-to-year basis. Reaching target levels, which could be trend levels, does not guarantee that chronic malnutrition is avoided. At least for most of Latin America, the issue of chronic malnutrition associated with persistent poverty is a long-term problem, the solution to which lies beyond meeting aggregate targets.

Latin American food supplies in the aggregate

National food self-sufficiency is not synonymous with consumption stability for large segments of the population, who may remain exposed to food insecurity if the country does not have the means to transfer food or income to regions or households. An obvious example is Brazil and Argentina, where there is no aggregate deficit in the food supply, but there are households suffering from hunger and malnutrition, the result principally of poverty and inadequate social programmes.

In middle-income countries, aggregate food supply is less relevant in the context of an open economy. With a few exceptions, in Latin America the question of attaining a target of food self-sufficiency is no longer an important consideration. Arguments over grain reserves or food security stocks have almost disappeared from informed policy debate. Practically all LAC countries for all food imports are price takers in world markets, with perhaps only two exceptions: white maize in Central America and some bean varieties in Brazil. Both products are traded within very “thin” markets. The question of food supply management has become primarily a matter of foreign exchange and the capacity of infrastructure (mainly ports) and logistics to deliver food imports to consumers. Tariffs and QRs certainly would do little to improve delivery systems.

Trade policy has been decoupled from the political discussion on food policy, because trade policy instruments are seen as being inadequate to deal with the goal of increasing household income and food consumption in the poorest households. There are two notable reasons for this. First, a large proportion of the poor are net buyers of food, and thus border protection on food imports would increase domestic food prices for all consumers, including the poor - there would be a decrease in real income. Second, trade in all products, including food, is practically everywhere in the hands of the private sector, and price controls at the wholesale and retail level have been eliminated. Although the situation with food aid, which only applies to a few countries in the region (some in Central America, Bolivia, Haiti), still gives governments flexibility and influence in selected food markets, politically, governments are increasingly free from food supply management. They would face severe political consequences if major food shortages were to occur - but such shortage do not occur, except in the case of contradictory signals sent to private producers and traders by the most inept governments.

Food security at the household level

In the context of Latin America, with some exceptions like Haiti, defining food security in terms of calorie intake is not always appropriate. For the region as a whole, calorie intake, stagnant prior to 1990 has risen steadily since, coincidental with the initiation of reforms in most countries^[269]. At the household level, particularly in middle-income countries, food security should be identified with nutritional status and health. This implies an emphasis on anthropometric indicators to measure nutritional status, and on specific indicators of health, rather than calorie intake, which is more relevant to the extremely poor. This is partly captured in the evidence of the evolution of the incidence of stunting (Tables 14.7 and 14.8) where regional data is presented on the incidence and prevalence of underweight, stunted and wasted children. Latin America has by far the lowest prevalence of all three measures in the developing world, measured as the percentage of the population falling below minus two standard deviations of medium as a threshold. Three percent would be the expected incidence in a well-nourished population^[270].

Table 14.7 Incidence of stunting, Latin America and the Caribbean, 1980-1995

Source: Table 1.14 in Wodon, 2000, based on data from the World Health Organization; FAO State of Food and Agriculture 2001,

Table 14.8 Prevalence of underweight, stunted and wasted children, developing regions and countries, 1995-2000

Percent of population falling below minus two standard deviations of reference value. Source: FAO, The State of Food and Agriculture 2001, Table 5, p. 63, based on data from UNICEF.

More generally, one could argue that, for middle-income countries in Latin America, the variable of concern is household welfare, an important component of which is the health status of its members. Health status depends in part on nutritional status. As with nutrition, the concept of a health production function could be relevant: it would be determined by nutritional status, privately provided inputs (clean water, boiling capability, refrigeration, time and health care in preparing food), publicly provided inputs (potable water, sewage, electricity, nutritional information, and others), and socio-economic variables. Families decide to allocate additional income to improve their infrastructure in their home (electricity, refrigeration, sewage, potable water) rather than to buy more calories, and this could improve their nutritional and health status. Income gains could thus reduce malnutrition even when it has a small or insignificant effect on a simple measure of nutrient intake at the household level. Agricultural growth in the context of more open and dynamic economies could be one source of income growth at the household level, and indirectly a source of government revenues for the provision of publicly provided inputs^[271].

It appears that the principal policy concerns following the reforms of the last two decades have centred on household income, health, and sanitation, in conjunction with targeted nutritional programmes as part of broader social safety net schemes. Social programmes include food stamp type programmes, school lunch programmes, and direct cash and in-kind transfers as part of an integrated policy of nutrition and health systems. With adequate domestic policies targeted to the poor, trade liberalization in agriculture has not been inconsistent with food security; on the contrary, it may have lowered the price of food, benefiting all consumers, including those with the greatest food insecurity.

The effect of price shocks

Does trade liberalization increase the risk of shocks that might precipitate food crises? At a national level, is agricultural liberalization associated with increased volatility in production and prices? The diet in Latin America has become more diversified and less dependent on one or two products. The removal of price controls and the deregulation of markets contributed to this diversification by removing the artificially low prices of only a few staples. This in turn has made the poor less vulnerable to food supply or price shocks in specific products.

Nevertheless, tariffication has increased the transmission of world price variability to domestic markets. The peaks in international prices, however, tend to be of short duration, unlike periods of low commodity prices which tend to endure, much to the dismay of producers. Tariffication did expose consumers to momentary price spikes, but also allowed them to benefit from extended periods of low prices. In Latin America, the enhanced price transmission resulting from trade reform has not appeared to have generated major consumer problems associated with price volatility.

But what of small farmers producing food and not net-buyers of food? The small farmers in non-tradables are not affected by trade policy, but those producing importables are harmed by reduction in the border protection of food. As documented for farmers of North Eastern Brazil, small farmers, as consumers, are often net-buyers of importables (not only food) which could compensate part of their loss on the output price side. In most cases small, farmers qualify as recipients of social safety net schemes. Distant rural areas, however, are often the problem, and much is yet to be done on improving the delivery of programmes to these areas.

14.6 Conclusion

Most Latin American countries reformed early and trade liberalization was part of economy-wide reforms involving deregulation, privatization, and macroeconomic adjustment. These reforms were for the most part unilaterally implemented before the Uruguay Round Agreement. Compared to other regions, Latin America as a whole implemented the most significant reductions in agricultural protection. Considering the economy-wide nature of reforms, any assessment of the effects of trade reform must take account of the many other reforms and factors.

Latin America is a heterogeneous region, both in terms of the timing and depth of reforms, and in terms of the net trade positions and initial conditions which influenced the consequences of the reform process. Nevertheless, common threads are discernable. In general, prior to the reforms, the trade regime in the region was characterized by strong import substitution and an anti-export bias. Most countries adopted an explicit trade reform policy, the central objective of which was to reverse the negative consequences of protectionism, especially its inherent anti-export bias. The principal policy mechanism in common was to be a reduction in average protection economy-wide. Once exchange rates were adjusted and quantitative restrictions reduced, the next common goal was to adjust tariffs in a manner that their levels and range were decreased.

Unfortunately for reformists, early in the reform process trade and agricultural policy changes were taking place in an unanticipated environment of declining profitability of farming. Between 1986 and 1995, in seven of the eight countries studied, all major domestic farm prices declined in real terms. The main factor explaining these real farm price declines was the appreciation of exchange rates observed in the early 1990s, a phenomenon amplified by tariff reductions and in some cases a fall in border prices.

In spite of these unfavourable conditions, there is a positive correlation between reforms and the growth rates of the agricultural sector overall. In the case of exports, the impact of the reforms is even more striking. Not only did the total value of exports expand, but trade liberalization appears to have encouraged a diversification of export products.

Within the agricultural sector of each country reforms affected subsectors differently. A useful way to assess the diverse effects within agriculture of trade reforms is to distinguish between the effects on the producers of exports, the producers of import-competing goods, and the - often small-scale - producers of home or non-tradable goods. Winners were the export-oriented sectors, commercial farmers and hired workers. Increased labour income as a result of the output mix has been one of the positive effects of the reform process from a welfare point of view. Import competing producers, however, probably lost in the short-run. Their long-run welfare depends on their capacity to increase productivity and/or change cropping patterns, which for some farmers and regions were very limited. Certainly low-income consumers (and farmers are consumers too) benefited from trade liberalization as lower protection reduced the price of food relative to wage rates.

Regarding trade liberalization and food security, the most relevant dimension is at the household level, considering that in middle-income countries, food self-sufficiency, and aggregate food supply policy are less relevant in the context of an open economy. Moreover, one observes that trade policy has been largely decoupled from the political discussion on food policy. Trade policy instruments are now seen as being inadequate to deal with the goal of increasing household income and food consumption in the poorest households.

With adequate domestic social policies targeted to the poor, trade liberalization in agriculture has been consistent with food security. It has lowered the price of food benefiting all consumers, including those with the greatest food insecurity. Diet has become more diversified making the poor less vulnerable to insecure food supply or price shocks in specific products. Although tariffication has increased the transmission of world price variability, the spikes in international prices tend to be of short duration, unlike periods of low commodity prices which tend to endure. Tariffication did expose consumers to momentary price spikes, but also allowed them to benefit from extend periods of low prices.

Looking ahead to the forthcoming Doha Round of negotiations, the most sensitive domestic trade policy debates in Latin America will centre on policy instruments to deal with import-competing sectors. By contrast the issue of food security will remain imbedded in the problems of poverty at the household level, to be addressed not by trade policy, but by integrated programmes of health, nutrition, sanitation and income transfers.