Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Roberto de A. S. Vellutini
Department of Agricultural Economics, Federal University of
Vicosa, Vicosa, Brazil
The Brazilian experience with the balance of payments and economic growth since the late 1960s can be summarized as follows [1]. During 1968-1970 there occurred a strong economic recovery that followed the 1963-1967 stabilization period. An emphasis on the promotion of exports was initiated, along with the adoption of mini-devaluations of the domestic currency. After this period, there followed the socalled "economic miracle" (1971-1973), when the rates of growth of the gross internal product (GIP) reached record levels. An improvement in the terms of trade occurred together with a massive inflow of foreign capital. Inflation rates for this period averaged 20 per cent a year. After the first oil shock (1974-1975) some adjustments were made in the economy - such as reduction in the level of imports - although an attempt was made to keep GIP growth rates relatively high. The average inflation rate for this period reached 40 per cent. After the second oil shock, the economy still followed an expansionist path, although inflation rates surpassed 100 per cent in 1980 and 1981. In January 1983 an agreement was established with the International Monetary Fund (IMP), imposing an array of recessionary economic measures as part of a stabilization programme.
The economic policies followed after 1974 can be characterized as an attempt to reduce the country's external dependence. Much of the policymaking during the years of "crisis" was aimed, implicitly or explicitly, towards that objective. In this paper I discuss the macroeconomic adjustments in Brazil over the last decade together with the stabilization policies followed during this period. The next section provides an overview of such adjustments at the macro level. In the following sections, the impact of these macroeconomic adjustments and policies on the agricultural sector and on income levels and distribution are analysed.
MACROECONOMIC ADJUSTMENTS AND POLICIES
The period that preceded the 1973-1975 "crisis'' was characterized by a vigorous expansion of the Brazilian economy, mostly between 1968 and 1973. This economic growth was based mainly on expanding output in the consumer-durables sector (22 per cent a year), while other sectors of the economy showed smaller but positive growth rates (15 per cent in the capital-goods sector, 13.4 per cent in the intermediate-inputs sector). Agricultural output showed an average growth of 4.7 per cent a year [5].
Besides the record growth rates of output, the 1968-1973 period presented two important features with respect to economic policy-making. First, there was a considerable emphasis on export promotion, through exchange, and fiscal and credit incentives. These permitted Brazil to accrue some of the benefits of international trade: the dollar value of its exports grew at an average rate of 24.3 per cent, with its share of world exports rising from 0.86 to 1.18 per cent during the period [5].
The second characteristic of the years 1968-1973 was a considerable inflow of foreign capital, mostly in the form of medium- and long-term loans. The magnitude of this inflow before 1968 was roughly US$500 million, while in 1973 it reached US$4.2 billion (US$4,200 million). It is important to note that the overall domestic economic scenario was a decisive factor in attracting foreign investment. The performance of the export sector provided a means to generate the foreign-exchange earnings for payment of the loans, while a growing economy was a favourable scenario for investment projects.
The first oil shock, in 1973, reversed the trend of the previous period. There was a deterioration in the trade balance, triggered by an increase in the value of imports as well as by a reduction in the level of exports. The latter effect was a consequence of the "world crisis," which induced many countries to adopt austerity measures in line with the new relative prices. In spite of this new and adverse scenario, the Brazilian government decided to continue its previous path of expansion. The policy goals of this period were to control inflation rates and to re-equilibrate the balance of payments, which were clearly two antagonistic objectives in the overall economic context.
Control of accelerating inflation rates and of increases in the value of imports was attempted through a restrictive monetary policy to control domestic demand. For annual inflation rates of 34.5 per cent in 1974, 29.2 per cent in 1975, 46.3 per cent in 1976, 38.8 per cent in 1977, and 40.8 per cent in 1978, there were corresponding growth rates of the monetary base of 33.1, 36.1, 49.9, 50.6, and 44.9 per cent. This relationship illustrates the classic "rule" of monetary policy that suggests a close relationship between the growth of money supply and that of inflation rates. The results of a restrictive policy to control domestic demand were reflected in lower rates of GIP growth during this period: from an annual rate of 14 per cent in 1973, the GIP growth rate declined to 9.5 per cent in 1974, 5.6 per cent in 1975, 9.7 per cent in 1976, 5.4 per cent in 1977, and 4.8 per cent in 1978. Nevertheless, the economy was still moving along an expansionary path, benefited by favourable credit conditions in international financial markets.
To equilibrate the balance of payments, the inflow of foreign capital- external debt- was the chosen policy alternative. These funds were used for financing domestic expansion, both in the capital-inputs sector (to provide substitutes for imports) and in the export sector, the latter including mostly manufactured, mineral, and agricultural products. As a result, total exports grew at an average annual rate of 14 per cent between 1974 and 1979.
The consequence of these measures was a gradual increase in inflation rates and in the level of external indebtedness. In 1973 the total net foreign-capital inflow to Brazil was approximately US $ 11.8 billion in real terms (1982 = 100). Of this total, US$9.5 billion (81 per cent) consisted of foreign loans. By 1978 the total inflow amounted to US$25.2 billion in real terms, an 85 per cent in crease over the period, with the share of foreign loans increasing to 87 per cent, i.e., US$21.8 billion. Such a rate of external indebtedness led to an increase in the value of the foreign-debt service. In 1973 the ratio of the debt service (interest plus amortization) to the current transaction receipts (exports of goods and services) was 35 per cent, with the former amounting to US$2.5 billion in current terms. By 1978 the debt service reached US$7.8 billion, with its ratio to foreign receipts rising to 54 per cent. Inflation rates for 1974-1978 averaged 38.0 per cent, compared to the 1973 level of 15.5 per cent.
As pointed out before, an important aspect of the policies followed in the 1974-1979 period was their import-substitution orientation, which operated through changes in relative prices. This was of particular relevance in the energy sector, with the national alcohol programme being initiated at the end of 1975.
With the second shock of oil prices, in 1979, the adverse conditions surrounding the Brazilian economy were considerably aggravated. The dollar price of OPEC oil increased 200 per cent between 1978 and 1980, triggering world-wide inflationary pressures with impacts on the international trade and financial systems. The policy choice, however, was to continue the expansionary path, rather than adjusting the economy to the new scenario through. austerity measures. A package of anti-inflationary policies was established, which attempted to deal with the problem from the supply, rather than from the demand, side. Its rationale was that inflation could be dealt with more efficiently through supply-expansion measures than through demand controls. In particular, the supply of agricultural products was considered a key policy component, and a massive dose of subsidies were directed to the agricultural sector.
Control of inflation was also attempted on two other fronts: the indexation level and the exchange rate. The former was established at 45 per cent, compared to inflation rates in the neighbour-hood of 100 per cent in the years immediately after 1979. A fixed indexation below inflation rates was aimed at reducing the expectations of the inflationary process. Adjustment in the balance of payments was attempted through successive devaluations of the domestic currency, culminating with a substantial devaluation in 1983.
Despite these measures, on the whole not very "orthodox," the inflation rate substantially in creased, reaching 38.0 per cent for the 1974-1978 period; the rate reached 77.2 per cent in 1979 and 110.2 per cent in 1980, declined to 95.2 per cent in 1981, but rose to 99.7 per cent in 1982 and finally to 211.0 per cent in 1983. This was the worst inflation record since 1946, if one excludes the politically turbulent years of 1963 and 1964.
Rosetti 120] has pointed out some of the factors underlying the Brazilian inflationary process. First, there are difficulties in controlling the expansion of the monetary base and hence of the money supply because of the continuation of government credit programmes for agricultural and export activities and the implementation of regional development programmes. Second, the institutionalization of the wage-indexing process constitutes a powerful feedback mechanism for maintaining high levels of "inertial" inflation. Third, the non-recessionary path chosen for the Brazilian economy after 1973, as pointed out above, led to a continuation of the government programmes in basic sectors (e.g. energy). Since these projects were not redimensioned under the new adverse scenario, both the monetary and fiscal budgets were overloaded. Financing these additional resources through monetary expansion was a decisive factor for increasing the price level. Finally, there is the participation level of the public sector in the economy, considered one of the most important factors underlying the Brazilian inflationary process. This factor is so important because the public-sector deficit requires substantial amounts of resources, which are generated through monetary expansion and open-market operations, whereby government bonds are traded for monetary assets.
Among the factors associated with the inflationary process in Brazil, the size of the public-sector participation in the economy deserves special attention. The size of the public deficit as a percentage of the GIP amounted to 8.3 per cent in 1979, 6.7 per cent in 1980, and 5.6 per cent in 1981 [28]. The open-market operations used for generating funds for the deficit financing put an upward pressure on domestic interest rates, as the government had to offer higher remuneration rates for its bonds. Eventually, given the size of the public sector deficit, a partial "crowding-out'' of financial markets occurred, as discussed in the macroeconomic literature (see, for example, Turnovsky [24]). This represents a "takeover" of these markets by the public sector, due to government expenditure, which raises interest rates and offsets private investment. In fact, by the end of 1980 the amount of monetary assets held by the public sector (currency plus demand deposits in government banks) was nearly 54 per cent of the total monetary assets, which include demand deposits in commercial banks. For non-monetary assets (savings accounts, bonds, and other assets), this percentage reached 72 per cent of the total in the same period 118]. The increasing financial needs of the public sector, in addition to forcing an expansion of the money supply, pull interest rates up, hence increasing financial costs for private firms. These, in turn, tend to increase mark-up over costs to maintain profit margins, putting additional pressures on the price level.
Hence, the public-sector deficit was an additional factor behind the expansion of the monetary base after 1979. In fact, growth rates of this indicator over the period 1979-1983 were considerably higher than in previous years. From an average annual growth of 43.0 per cent between 1974 and 1978, the monetary base showed an increase of 84.4 per cent in 1979, 56.9 per cent in 1980, 69.9 per cent in 1981, 86.8 per cent in 1982, and 89.1 per cent in 1983.
It is interesting to notice that after 1980 there occurred a dissociation between the expansion of the money supply and the observed inflation rates. More specifically, the increase in the price level after 1980 has been considerably higher than that of either the monetary base or the money supply (as measured by currency plus demand deposits). This phenomenon violates the mechanism described in monetarist theory whereby an expansion in the money supply would occur along with a similar increase in the price level. Similarly, the income velocity of money, defined as the ratio of nominal GIP to the money supply, has increased substantially since 1979. During 1970- 1979 it was rather stable, in the neighbourhood of 6.0 to 7. O. In 1980 it increased to 9.6, reaching 11.2 in 1981, 13.1 in 1982, and 20.6 in 1983. Again, the monetarist assumption of a constant relationship between nominal income and money supply was violated.
One explanation of the apparent inconsistency between the data and the theory is that the relevant concept of "money" in Brazil has been changed over the past years, due to innovation in domestic financial markets. Thus, the relevance of the concept of money supply as measured by M1 (currency plus demand deposits) in explaining monetary phenomena became quite limited. In fact, given the multitude of financial assets protected against inflation through indexation, many of them of high liquidity, the substitution between money and financial assets predominates over that between money and real goods. In addition, the high interest rates prevailing in financial markets have increased the interest-elasticity of the demand for money. Estimates by Martone [19] for 1974-1982 indicate an interest-elasticity between 0.55 and - 0.69 (depending on the monetary indicator used), which is higher than that obtained for the 1960s (-0.3). That is, people who have available a variety of financial assets protected against inflation have substituted these assets for money because of the higher costs of maintaining currency in an inflationary environment.
It should be noted that the process described above has occurred since 1981. In the period 1979-1980, with a fixed indexation level (applied to financial assets) below inflation rates, as pointed out before, the behaviour was quite different. Since these assets were not protected against inflation, there occurred a substitution of real assets (goods) for financial ones, leading to an expansion of aggregate demand, partially responsible for the three-digit inflation record observed in 1980 (110.2 per cent). In addition to inflationary pressures, such mechanisms led to a trade deficit of approximately US$2.8 billion in 1979 and 1980, as well as an increase in the exchange earnings gap (trade balance plus commercial services balance) to US$5.2 billion in 1979 and US$5.9 billion in 1980.
By the end of 1980 a series of contractionary measures was implemented, aimed at reducing the level of domestic economic activity. These included a reduction in the money supply, an increase in interest rates, and control of the public-sector deficit. The monetary contraction can be measured by the behaviour of the conventional indicators of money supply, i.e. currency plus demand deposits (the M, monetary aggregate). The index of currency availability showed a substantial decrease between 1970 and 1983. From an average index of 100 for the 1970 decade, total real currency declined to 97.8 in 1980, 89.3 in 1981, 86.5 in 1982, and 55.7 in 1983. Demand deposits fell from an index of 100 for the 1970 decade to indices of 85.8 in 1980, 75.7 in 1981, 61.8 in 1982, and 44.6 in 1983. The M1 indicator presented a similar performance, moving from an index of 100 for the 1970s to 88.1 in 1980, 78.4 in 1981, 66.5 in 1982, and 46.7 in 1983. That is, the level of liquidity in the economy declined drastically after 1980, especially in 1983, increasing interest rates and accelerating the process of a reduction in output and employment.
The contractionary measures adopted after 1980 succeeded in reversing the trade deficit observed in the previous period, with surpluses of US$1.2 billion and US$0.8 billion in 1981 and 1982 respectively. The exchange earnings gap as defined above, was reduced by more than US$3.0 billion in 1981.
The cost of attempting to equilibrate the balance of payments and control inflation was a drastic reduction in the level of economic activity and employment. The rate of growth of real GIP was still reasonably high in 1979 (6.8 per cent) and 1980 17.9 per cent) because of the overall economic conditions, as discussed above. The recessionary environment that characterized subsequent years caused drops in real GIP growth rates to - 1.9 per cent in 1981, 1.4 per cent in 1982, and -3.3 per cent in 1983. Estimates by Rosetti [20] for sectoral growth rates showed a similar pattern. Industry was still showing positive growth rates in 1979 (6.6 per cent) and 1980 (7.9 per cent), whereas in the following years, this pattern was reversed: -5.4 per cent in 1981, 2.0 per cent in 1982, and - 7.0 per cent in 1983. The growth rates for the agricultural sector over the same period were: 5.0 per cent in 1979, 6.3 per cent in 1980, 6.8 per cent in 1981, -2.5 per cent in 1982, and 2.2 per cent in 1983.
Another indicator of the level of activity is the GIP gap, defined as the difference between potential and actual GIP. This is a measure of the economy's "idle capacity." Estimates of the GIP gap for this period showed a substantial increase during the years 1981-1983. For 1973-1978 this indicator oscillated between zero (1976) and 3.25 per cent (1978). It was reasonably stable during 1979 (3.74 per cent) and 1980 (2.12 per cent), after which it moved drastically upward to 12.82 per cent in 1981, 19.42 per cent in 1982, and 29.70 per cent in 1983.
With respect to the external sector, the conditions were not very favourable. Interest rates in international markets were going up, following the restrictive trend of US monetary policy. As a consequence, Brazil's real debt was increasing, because of higher interest rates and the adverse effects of higher oil prices on the trade balance. The increasing costs of the external debt can be measured by the rise in the net average interest rate prevailing over the period: from 10.9 per cent in 1978, it rose to 13.7 per cent in 1979, 15.7 per cent in 1980, and 19.6 per cent in 1981 [5]. The gradual increase in interest rates meant interest payments of US$5.3 billion in 1979, US$7.5 billion in 1980, US$10.3 billion in 1981, US$12.2 billion in 1982, and US$9.7 billion in 1983, compared to an average value of US$2.1 billion over 19741978. Payments of interest plus principal associated with the external debt represented an increasing proportion of the current-account receipts: from an average of 44.0 per cent in 1974-1978, this ratio increased to 64.7 per cent in 1979, 53.8 per cent in 1980, 61.4 per cent in 1981, 82.5 per cent in 1982, and 82.1 per cent in 1983.
It is important to note that worldwide economic conditions after 1980 were extremely adverse, which aggravated the scenario described above. In fact, the annual average growth of the real GIP in the OECD countries declined from 5.06 per cent (1971-1973) and 3.97 per cent (1976-1979) to 1.22 per cent in 1980, 1.25 per cent in 1981, and -0.50 per cent in 1982. Simultaneously, unemployment rates rose from 3.4 per cent (1971 - 1973) and 5.2 per cent (1976- 1979) to 8.8 per cent in 1982.3 Increasing interest rates can be measured by the labour rate, which increased from 8.47 per cent in 19741978 to 14.36 per cent in 1980, 16.38 per cent in 1981, and 13.11 per cent in 1982.
Adverse international economic conditions had a major impact on the performance of Brazilian exports. Along with protectionist measures implemented by principal trade partners, this scenario provoked a virtual stagnation of export earnings over 1980-1983. Combined with the second oil price shock, these effects led to a deterioration of the external terms of trade. From an index of 100 for 1970-1972, there was a gradual decline to indices of 92 in 1979, 76 in 1980, 58 in 1981, and 63 in 1982. This decline corresponds to an increase in the index for export prices (1970- 1972 = 100) to 249 in 1982, and that for import prices (1970-1972 = 100) to 394 in the same year. That is, the price of Brazilian exports increased 2.5 times (with a concomitant increase of 2.7 times in the quantum), whereas the price of its imports increased almost four times (for a less than twofold increase in the quantum) during the last 15 years. To a great extent, the poor performance of exports after 1980 neutralized the contractionary measures imposed on the domestic market (discussed above), which were aimed at reducing the need for external loans.
As can be seen, the prospects for an eventual adjustment in the Brazilian internal and external sectors after 1979 were not very promising. The balance in the current-transactions account (trade balance plus services balance) deteriorated from -US$0.6 billion in 1978 to -US$10.8 billion in 1979, US$13.0 billion in 1980, -US$13.3 billion in 1981, and - US $ 14.5 billion in 1982. The percentage of this deficit accrued to payments of interest plus profits (factor services) amounted to 59 per cent in 1978, remained at this level until 1980, then increased to 87.2 per cent in 1981, and in 1982 reached a magnitude above unity. Attempts to equilibrate the successive deficits through foreign loans were becoming increasingly difficult. Mexico's insolvency in 1982 triggered a virtual suspension of new loans to debt-troubled nations. Considering the net capital inflow to Brazil, one can observe a 44 per cent reduction (in real 1982 terms) in the amount of foreign loans after 1979: from US$21.9 billion in 1978, loans declined to US$12.2 billion in 1982. Under these circumstances, the administration of the Brazilian external debt became a major problem.
The policy orientation for the subsequent period was to deed with the external crisis. A series of measures to control aggregate domestic demand were implemented, and a maxi-devaluation of the domestic currency, of approximately 30 per cent, took place in February 1983. In the same year a state of total insolvency was reached, and Brazil had to start a round of negotiations with the International Monetary Fund (IMF). The basic terms of the agreements with the IMF, contained in ''letters of intention," established guidelines for medium-run economic policy. These included, among others: a reduction in the current-transactions deficit, a wage policy oriented toward employment expansion, elimination of fiscal and credit subsidies, expansion of government revenue through a series of taxation policies and reduction of government expenditures, reduction in the interest rates to boost domestic demand, and exchange-rate policies aimed at maintaining parity between domestic and foreign currency. These stabilization measures, on the whole recessionary, have had a considerable impact on the economy. Besides the negative growth rate of the GIF in 1983 (-3.3 per cent), the economy's idle capacity rate in the same year was approximately 30 per cent, and unemployment figures oscillated between 15 and 18 per cent of the total available labour force [20].
Under the IMF's supervision, the Brazilian economy started 1984 with an array of restrictive monetary and fiscal policies and an inflation rate of 211 per cent.
Fortunately, the external environment showed some positive trends, led by a partial recuperation of the US economy. Brazilian exports in 1984 increased 23.3 per cent over 1983, yielding a trade surplus of US$13.0 billion. A current transactions surplus of US$165 million was obtained, and inflation stabilized at 223.8 per cent in 1984.
PERFORMANCE OF THE AGRICULTURAL SECTOR
The performance of Brazilian agriculture over the past decade has been characterized by four major factors 17]. First, a change occurred in the sources of growth of the sector, with higher rates of labour and land productivity becoming increasingly important. Nevertheless, the incorporation of new lands into the productive process (extensive growth) has remained the main source of increases in agricultural production. Second, there was significant growth in the openness of foreign markets to the agricultural sector, which led to export products' representing an increased share of the value of agricultural production, from 28 per cent in 1970 to 44 per cent in 1980 [based on data presented in Barros [2]). Greater openness makes the sector more dependent on commercial and exchange-rate policies and international economic conditions. Third, a considerable imbalance occurred between the performance of the food-crop sector and that of both the energy- and export-crop sectors, especially with respect to the growth of cultivated areas. Finally, there was the influence of adverse external factors, such as the oil-price shocks, restrictive US monetary policy, overvaluation of the dollar, and the drop in international commodity prices. Obviously, a higher vulnerability to external shocks was an immediate consequence of the greater openness of the agricultural sector, as mentioned above. The macroeconomic stabilization programmes that were implemented determined, among other things, a tight control over agricultural credit. The total credit for the agricultural sector increased from Cr$78.5 billion (in real terms) in 1971 to Cr$227.3 billion in 1977, after which it declined to Cr$152.8 billion in 1982, a 33 per cent drop over five years. As a result, agricultural terms of trade were reduced between 30 and 36 per cent in the major states during 1979-1983.7
Much of the performance of the agricultural sector over the last decade was a result of macroeconomic policies aimed at two basic targets: control of inflation and external equilibrium (balance of payments). In particular, recent efforts to increase the supply of export, energy, and food crops have had a considerable impact on the economic structure of agriculture. The national alcohol programme, initiated in 1975, promoted the expansion of the energy-crop sector (sugar cane) through pricing and credit policies. At the same time, export-crop expansion was being stimulated through commercial and exchange-rate policies. The food-crop sectors, on the other hand, were subject to a price-support programme in which minimum producer prices were consistently set below market prices [27].
The most readily noticeable consequence of this set of policies was the upward trend in the area devoted to energy and export crops, which have shown higher rates of growth than food crops. The area under export crops (including cotton, peanuts, cocoa, coffee, tobacco, oranges, castor beans, and soybeans), which was 13.78 million ha in 1976, the last year before the effects of the alcohol programme, grew to 16.9 million ha in 1984, a 22.6 per cent increase over the period. This may be seen, however, as relatively quite stable, as it represents an average annual growth rate of 2.58 per cent. The area devoted to the energy crop (sugar cane) increased 84 per cent over the same period, from 2.09 million ha in 1976 to 3.85 million ha in 1984- an average annual growth rate of 7.9 per cent. The area under food crops Rice, potatoes, onions, beans, cassava, corn, and tomatoes), on the other hand, increased from 24.75 million ha to 25.02 million ha over the period, or 1.1 per cent - an average annual growth rate of 0.14 per cent.
The behaviour of production levels among the different sectors showed a similar pattern. Food-crop production increased 2.0 per cent between 1976 and 1984, from 56.94 million tons to 58.06 million tons respectively- a 0.24 per cent average annual growth rate. Production levels of individual crops remained relatively stable or even declined during the period. The export-crop sector (excluding oranges), on the other hand, increased its production from 15.66 million tons in 1976 to 21.96 million tons in 1984, a 40 per cent increase, or an average annual growth rate of 4.3 per cent. (Oranges are excluded because their production levels are measured in 1,000 pieces of fruit rather than in tons. If they were included, the growth rate of the sector would be even higher.)
The energy-crop sector (sugar cane) showed a considerably higher expansion in production levels, increasing more than twofold, from 117.5 million tons in 1976 to 247.53 million tons in 1984, an average annual growth rate of 9.8 per cent.
The virtually stagnant production levels of the food-crop sector during the period considered had important implications for domestic food availability. Average population growth rates of 2.5 per cent per year between 1977 and 1980, and 2.3 per cent per year between 1980 and 1984, meant declining levels of per capita food availability. Estimates by Homem de Melo [16] for 1977-1984 indicated a decline in the index of per capita food availability from 100 (1977) to 91.7 (1984), corresponding to an annual rate of - 1.73 per cent over the period. This index refers to food availability, and hence includes domestic production, imports, exports, and stock variations. As such, it is likely to yield a more favourable picture than that in which only domestic production is considered.
The effects of declining levels of per capita food supply on nutrient availability were not very favourable. Indices of calories and protein per capita calculated by Homem de Melo [16] for 1977-1984 show a pattern similar to that of the food-supply indices. Indices for calories per capita available from domestic production of five major staples - rice, beans, potatoes, cassava, and corn - with the 1977 value taken as 100, were as follows: 1978 = 79.3; 1979 = 84.3; 1980= 94.0; 1981 = 92.7; 1982 = 95.9; 1983 = 78.2; and 1984=85.7. Indices calculated in terms of total domestic availability of these staples (i.e., including exports, imports, and stock variations) show a slightly better performance: 1977 = 100.0; 1978 = 97.0; 1979 = 98.5; 1980 = 96.8; 1981 = 91.5; 1982 = 95.0; 1983 = 84.6; 1984=91.7. Note that, as these indices refer only to the five staples, they do not reflect the total calorie availability. In 1977, according to Homem de Melo's estimates [16], the five staples provided 1,802 calories per capita per day, whereas all sources taken together- including sugar, wheat, vegetable fats, animal products, etc. - provided 3,394 calories, in both cases considering total domestically available quantities.)
The indices for per capita protein availability from domestic production of the five major staples were: 1977 = 100.0; 1978 = 77.0; 1979 = 83.4; 1980 = 94.4; 1981 = 94.2; 1982 = 100.5; 1983 = 77.9; 1984=90.8. The high index for 1982 was due to a large domestic production of beans that year, a product with a relatively high protein content. Again, indices for per capita protein availability from total domestically available quantities of the staples show a better performance: 1977 = 100.0; 1978 = 96.1; 1979 = 99.8; 1980 = 105.0; 1981 = 97.8; 1982 = 99.1; 1983 = 91.4; and 1984= 98.6. (The five staples provided 42.6 g of protein per capita per day in 1977, and the total protein availability per capita from all sources was 79.3 9, according to Homem de Milo s estimate. )
Declining per capita levels of nutrient availability, given a positive income-elasticity of demand for food crops, are likely to increase the implicit price (cost) of nutrients. Barros [2] estimated the real cost of calories from domestic production of major food crops over the 1964-1980 period." Between 1964 and 1972, the cost of 106 calories was quite stable, oscillating between Cr$5.3 and Cr$6.8. This situation changed rapidly after 1972, with the cost of 106 calories increasing to the following values: 1973 =Cr$9.7; 1974=Cr$12.5; 1975 = Cr$ 13.2; 1976 = Cr$ 13.8; 1977 = Cr$ 13.2; 1978 = Cr$ 12.8; 1979 = Cr$ 12.6; 1980 = Cr$ 11.5. That is, the average cost of calories in real terms over the period 1973-1980 was more than twice as much as in 19641972.
Upward movements in nutrient prices, as measured above, result from escalating food prices. These, in turn, can be determined by factors from the demand and/or supply side. Demand-related factors include higher levels of income and substitution effects, among others. Supply-related factors are usually associated with reductions in domestic availability caused by declines in cultivated land area and/or yields. Hence, it becomes important to identify the major determinants of food price behaviour over the last decade. As will be seen, much of this behaviour can be explained by economic conditions at the macro level, as well as by the policy measures that were followed during the years of crisis.
Food price behaviour can be analysed by comparing food price indices with other sectoral and aggregate price indices. For example, the wholesale price index for agricultural products (WPIA) was 20,070.2 in 1984 (1977 = 100), whereas the wholesale price index for industrial products (WPII) reached 14,196.5 and the general price index (GPI) - the index of inflation - reached 14,311.7. Thus, the increase in the WPIA was 41 per cent higher than the WPII, and 40 per cent higher than the GPI. The consumer cost-of-living index for food (CLIF) showed a similar pattern, with a value of 15,006.8 in 1984 (1977 = 100), 5 per cent higher than the inflation rate for that year. As can be seen, a change occurred in relative prices between 1977 and 1984, with consumer food prices increasing more than other prices in the economy. Food price indices showed higher percentage increases than other price indices not only over the 19771984 period as a whole but also generally on a year-to-year basis, except in the recessionary years 1981 and 1982, when the WPIA increased 104 per cent and 75 per cent respectively while the WPII increased 109.6 per cent and 100.8 per cent. The GPI increased 109.9 per cent and 95.4 per cent in the same years. At the consumer level, the CLIF increased 94.2 per cent in 1981, against a 98.0 per cent increase in the general cost-of-living index. Thus we see that there was a decline in relative prices of food during the recessionary period 1981 -1982.
The analysis above suggests that for low-income consumers, who spend a larger proportion of their income on food items, the behaviour of food prices over 1977-1984 was not very favourable. This situation was the result of the cost-of-living index for food experiencing higher upward movements than the general cost-of-living index, the latter being used for indexation of wage rates.
Among the factors explaining the upward movements of food prices, the evolution of both per capita GIP and domestic availability are the most readily noticeable. Homem de Melo 1161 compared the behaviour of these two variables with that of a producer price index for the five staple food crops referred to earlier over 1977-1984. His analysis shows the combined effect of demand (GIP) and supply (domestic availability) factors on the food price index. In fact, for a per capita GIP index of 106.5 and a domestic availability index of 98.5 in 1979 (1977 = 100 for both indices), the price index reached 106.1, i.e. a 6.1 per cent increase over 1977. In 1980, the last year of relatively high economic growth, the indices were GIP = 112.1, domestic availability = 96.8, and prices = 129.7. Thus, it is likely that demand pressures were the main sources of food price inflation in 1980. Between 1981 and 1983, the period characterized by a drastic contraction of the Brazilian economy, the declining levels of income (GIP) partially offset the upward pressure on food prices caused by reductions in domestic availability. The per capita GIP indices for this period were: 1981 = 105.7; 1982 = 104.5; 1983 = 98.2.
Domestic availability indices were 1981 = 91.5; 1982 = 95.0; 1983 = 84.6. Food price indices were: 1981 = 137.4; 107.2; 1983 = 111.0. Note that the drastic reduction in food supply in 1981 had a considerable impact on food prices, despite the reduction in per capita GIP from its previous level.
In addition to the levels of domestic availability (supply conditions) and per capita GIP (demand conditions), two other factors may have been partially responsible for the drastic reduction in producer prices for food crops between 1981(index = 137.4) and 1984 (index = 101.5). These were the reduced liquidity conditions in the economy and the substitution effects via relative prices. The first hypothesis, discussed in Sayad [21], suggests that a contraction in the money supply, which reduces the liquidity conditions in the economy, has a considerable downward impact on food prices, since this sector operates under competitive conditions and a closed economy perspective. As pointed out before, the level of liquidity in the economy was indeed drastically reduced after 1980, with the M, indicator declining 47 per cent between 1980 and 1983.
The second hypothesis for explaining the drop in producers' prices of food crops after 1981 suggests a substitution effect at the consumption level between food crops and those that either are exported land hence have prices exogenously determined in international markets) or have administered domestic prices. Such substitution in consumption would occur because of declining real prices of other crops, which increase the relative prices of food crops. The Laspeyres simple price indices for export crops (1977 = 100) computed by Homem de Melo [16] showed the following pattern: 1978 = 81.5; 1979 = 78.1; 1980 = 66.3; 1981 = 56.3; 1982 = 52.0; 1983 = 58.2; 1984 = 67.6.73 For sugar cane the indices were: 1978 = 97.2; 1979 = 96.9; 1980 = 104.0; 1981 = 103.0; 1982 = 94.4; 1983 = 77.7; 1984 = 76.9. Wheat, the prices of which are set by the government, had the following indices: 1978 = 102.3; 1979 = 89.3; 1980 = 75.8; 1981 = 87.4; 1982 =93.7; 1983 = 83.7. The drop in real prices of export crops occurred mainly after 1979 as a result of the world economic crisis that succeeded the second shock of oil prices. The recessionary US economic policy, which included a restrictive monetary policy, put an upward pressure on interest rates, which are crucial for commercial transactions, and depressed commodity prices in international markets. The resulting decline in real prices of export crops, coupled with reductions in wheat and milk prices (not shown), were important factors in limiting the upward movements of food prices.
Finally, it should be pointed out that, although there was an increase in the relative prices of food and export crops during 1977-1981 due to the fall in international prices, corresponding increases in per capita food production did not occur. This indicates that the expansion of the food supply in Brazil will depend also on factors other than prices, such as risk considerations (price and income instability) and quality of technology among others.
So far, the impact of economic conditions on the agricultural sector has been evaluated in terms of changes in the area of land allocated, production levels, nutrient availability, and food prices. However, it is likely that macroeconomic policies will have economy-wide effects, which tend to spill over across many sectors of the economy. In particular, there are additional variables in agriculture, beyond those mentioned above, that will be affected by economic conditions at both macro and sectoral levels. These include employment levels, factor payments, sectoral incomes, export earnings, and trade flows among others.
Most of the studies that have analysed agriculture in Brazil have been carried out in a partial-equilibrium framework. However, many of the effects beyond the production and price effects are not captured under such approach. Studies that used an analytical framework that incorporates some of the general equilibrium effects likely to be present in agriculture include those by Lopes [17] and, more recently, Vellutini [27]. Both studies analysed the impact of macroeconomic-oriented policies on the agricultural sector. Following the hypothesis put forth by Schuh [22], which states that an overvalued currency represents an implicit taxation of agriculture, Lopes [17] used Floyd's [8] model to analyse the effects of overvalued exchange rates on agricultural output prices, factor use, and factor payments. This was an aggregate model of the agricultural sector that provided links between product and factor prices. Lopes started with the assumption that the effect of the overvalued exchange rate and other restrictions on exports was to lower the price of agricultural products by 10 per cent relative to non-agricultural products. In the aggregate, his results showed that this shift in the internal terms of trade reduced agricultural employment by 18 per cent, capital use by 27 per cent, and land use by 6 per cent. Such policies constituted important factors in stimulating the out-migration of agricultural labour towards urban centres.
Vellutini [27] developed an econometric model of Sao Paula state's agricultural sector for evaluating the direct and indirect economic effects of policies that affect the production of energy and export crops. The policies analysed were a pricing policy for sugar cane and the devaluation of the domestic currency. These policies were evaluated in terms of the direct effects on the total area cultivated, the area and output for each major crop type, the levels of employment and capital use, and the amount of purchased inputs (e.g., fertilizer and fuel). The indirect effects were assessed under four major groups: crop export, fuel imports, food prices, and income levels in agriculture. The effects were captured in a model through macro-sectoral linkages which permit an assessment of the impact of macroeconomic policies on agricultural production decisions, as well as the impact of agricultural policies on macroeconomic components. Hence, the model explicitly incorporated market linkages at a more aggregated level, and a price-determination process for food crops that feeds back into the model, thus examining interactions not captured by previous studies.
The basic structural model comprises five major building blocks. The first determines production decisions in the agricultural sector, including area allocation, yields, and the resulting production at both the individual crop and the sectoral levels. The second structural component determines the demand for factors of production implied by production decisions in the first component. This system predicts employment levels in agriculture, capital use, and fertilizer and fuel consumption in the sector. The third component of the model provides the actual linkages between sectoral and macroeconomic variables. It determines labour and non-labour income in agriculture, total agricultural income, and state domestic income. The fourth component determines aggregate fuel demand and food prices. The final component of the model describes the trade flows in terms of fuel imports and the value of crop exports. Once food prices have been determined in this system, they are linked back to the production system as determinants of production decisions in the following year.
The policy simulation performed with the model consisted of increasing the real producer prices of sugar cane and export crops, i.e., an "improving the balance of payments'' scenario. The former reflects the policy guidelines of the ongoing national alcohol programme. Increases in the prices of export crops result from a devaluation of the domestic currency, a major macroeconomic policy in Brazil since the early 1970s. There are four major impacts of these two policies. First, there is a reduction in the area under food crops, ultimately reflected in higher food prices. Second, there is a transfer of income from labour to the non-labour component (returns to capital and land). The decline in income for agricultural labour results from the reduction in employment levels that occur under this policy scenario. Nevertheless, the net effect on total agricultural income is positive. Third, there is a decline in fuel imports due to the substitution of alcohol (from additional sugar cane production) for gasoline. Finally, these policies increase the value of crop exports as a result of higher production levels stimulated by higher prices for export crops which result from the devaluation.