Contents - Previous - Next

This is the old United Nations University website. Visit the new site at


Lance Taylor

What I would like to do is discuss, in general terms, the stylized facts about food price policy, and how they fit into planning and, in fact, make food price planning difficult.

The appropriate place to begin is with discussion of exactly how countries intervene in markets for food. By food, I mean the staple food or foods, typically a grain that is used to make bread, or that is eaten directly, as rice is in Asia. Beans also fit into this category if they form part of a staple diet as, combined with rice, they do in Brazil. In any case, there are one or two foods or food products that make up very large shares of the diet of poor people. Usually, the government tends to intervene rather drastically to control what happens in the markets for these food products, and this kind of intervention raises some problems.

To begin with, one has to think about three kinds of prices for food. Typically, there is a foreign price at which food is imported or exported; more often than not in an underdeveloped country, it is imported. There is a consumer price that is maintained at least for consumers in urban areas, and sometimes for consumers in rural areas as well if the government's writ extends as far as the rural market-as far as the bazaar, for example. Finally, there is the producer price. Governments almost always intervene to separate these prices. The form of the intervention is important in determining, in the final analysis, how much food different groups of people consume. For example, with the great increase in import prices of food that took place in 1973 - 1974, a number of food-importing countries protected their consumers by subsidies between the import price and the internal price of food. As a case in point, Egypt imports on the order of three million tons of wheat per year. When the external price tripled, the internal price was maintained constant by subsidy, and consequently the total food subsidy bill went up to about 30 per cent of the total Government expenditures-a very large share indeed.

Egypt is an extreme case, but nonetheless, a number of other countries have gone in this direction of subsidizing to keep internal food prices stable, and that sort of intervention in the market has substantial fiscal implications. There is an article by Jeffrey Davis (1 ) in the /International Monetary Fund Staff Papers that reviews some of these subsidy interventions, mostly in Africa and Asia.

On the other side of government accounts, taxes on essential food consumption can also be very important. There are studies in Sao Paulo, for example, showing that the indirect tax system is regressive and bears heavily on staple foods. Changing such a tax would have substantial fiscal impacts on the government, but on the other hand, such a tax strongly constrains the ability of poor people in Brazil to obtain adequate nutrition. Both subsidy and tax interventions affect consumer prices.

Another kind of intervention is in producer prices. For example, governments may maintain, by subsidy, high prices for farmers and producers. A classic case is in Europe under the Common Agricultural Policy.. Twenty million Frenchmen are kept on the farm by a high internal wheat price and a variable import tariff (that goes up and down) to separate completely the internal producer price in France and the rest of the Common Market from the rest of the world. That is a political reality in France that is unlikely to go away. There are similar political realities in the United States and Japan, where subsidies keep the prices that farmers receive well above what they might otherwise be.

Classically, that has not been the case in Latin America, although recent policies in Argentina and, to a lesser extent, in Chile, are certainly headed that way. On the other side of the coin, of course, producer prices can be forced down. If the government has some sort of monopolistic acquisition capacity, it sets the price and goes out into the market and buys, and it can hold the price low and use various kinds of measures to obtain the market surplus necessary for the city. That sort of intervention has been typical, for example, in Pakistan. As a consequence, a few hundred-thousand tons of wheat per year go across the border from Pakistan to India because prices there have been kept high. There are limits to how long low producer price policies can be maintained.

Finally, every country separates world prices from internal prices, or rather every country but the United States and Canada. When world prices shot up in 1973 - 1974, their consumer food prices went up almost as fast. The reason the World Food Conference was held, after all, was because there was an odd combination of US housewives on the one hand and underdeveloped importing countries on the other who were hurt by rising prices. Needless to say, that unusual coalition has dissolved and some of the current inaction about world food price stabilization has to do with that.

These price interventions can be found in most countries, and it is important to recognize that they are a permanent fixture. Given this fact, there are various questions to raise. First, how can price interventions be tailored to benefit poor people, or some groups of poor people, within a country? Unfortunately, some problems arise here, beginning with the general observation that in the short run, the amount of food that people seek to buy in the market and the amount of food that producers or sellers put on the market do not respond very much to changes in price.

Economists have their own special way of trying to quantify this lack of price response. The appropriate concept, of course, is an elasticity. What an elasticity reveals is the per cent change-for example, in demand for bread that would take place in response to a certain per cent change in the price of bread, rice, or some other commodity. When the economist says that the demand for bread is price-inelastic, he means that if the price of bread goes down by 10 per cent, then consumption of bread would be expected to go up by much less than 10 per cent.

Another way of putting it is that if the supply of bread goes down by 10 per cent, then the price has to rise by a lot more than 10 per cent to choke off demand enough so that the market will be in equilibrium. Therefore, the first stylized fact about food demand is that it is inelastic in the context described here. When the price of bread goes up by 10 per cent and consumption goes down by less than 10 per cent, the total value of consumption (price times quality) will go up. There will be less income left to consumers to spend on other things, and that will lead to less demand for products in the rest of the economy.

These generalizations about inelasticity hold almost universally, although among very poor people one does see some price responsiveness because they are spending essentially all of their income on food, so if the price goes up they simply have no choice but to cut their consumption. But at slightly higher income levels, such as are observed in most underdeveloped countries, the price elasticity is certainly less than one. In terms of macro-policy, that is one very important constraint that must be taken into account. Because staple food consumption also uses a substantial share of income, its inelastic demand and inelastic supply lead to large spill-over effects from food markets into other sectors that can be quite significant. It can become a full-fledged macro-economic problem. This problem is often avoided by desperate moves on the part of the government-for example, the Egyptians maintaining the internal food price constant when the external price tripled-but such maneuvers are not always possible.

Let me give some examples to help illustrate these macro-economic problems. I will use again this simple case of the food price going up by 10 per cent; then, as mentioned above, food intake will go down by less than 10 per cent. The share of income spent on food will go up, perhaps by 5 per cent, so there is a substantial shift in the share of income that the consumer has left for other things. When demand for these other things goes down, people are going to lose jobs or work fewer hours. This reduction in employment will lower incomes throughout the rest of the economy and there will be a series of successive rounds of this sort of effect. The ultimate impact can be quite substantial. It will show up, of course, in a poor country in reduced demands for textiles, clothing, shoes, and other commodities.

With the active urban proletariat that governments in many parts of the world face, the initial cut in the subsidy on food, or the initial increase in the price of food, coupled with the reduction in aggregate demand that it will create, can lead to food riots and extreme political unrest. Governments know this, and this considerable limitation on their freedom for policy maneuver arises precisely because of the large role of food in the economy.

There are similar examples about other interventions. For instance, most countries subsidize farmers' fertilizer purchases. If the subsidy is increased, a number of things will occur. In general, the reduced cost of fertilizer will lead farmers to produce more food, but it will also get passed along in lower food prices. If this is the case, farmers- especially small, poor farmers-may see their revenue and income go down. They will consume less food, and the intervention that is supposed to lead to more food and hence better nutrition may, in fact, lead to poorer nutrition in the countryside.

Flavio Machicado

I think the answers to that kind of question always depend on the conditions under which the question is posed. If the price of a basic product rises, its effect will depend heavily on the social group and income level where the price increase occurs. It is the same with the cost of fertilizers; it depends on the production structure, and on what kind of producer is affected. In Latin America, whether the price of fertilizer increases or decreases has not the slightest impact on the production of a large number of small producers, because they never even have access to fertilizers at any price. Consequently, I should like to ask Dr. Taylor to put his question into a well-defined context, otherwise no one can provide an answer.

Lance Taylor

That is a fair request; certainly one has to be more precise about the kind of people whom one is talking about to answer that kind of question.

Suppose that one distinguishes between large- and small-scale agricultural producers. Small scale agriculture may be largely detached from the market if it is the subsistence kind of agriculture that prevails in many Latin American countries. If the agriculture is attached to the market, and, in fact, some small-scale farmers and large-scale farmers buy fertilizer from the market, or sell it on the market, then my illustration holds.

To take that example farther-suppose the price of fertilizer is reduced and this does lead to lower food consumption by the rural poor. Are there policies to avoid that outcome; One way is for the government to buy up any increments in food pro auction for stock-building or export. But then the price of food will not fall, and groups within the city who are also poor will get no benefit from the increased crop. Any policy has various effects, and what these are depends on institutional circumstances in the country being considered. A move that appears on the surface to be a satisfactory solution-at least to economists who believe in getting the prices right-can actually have rather severe and deleterious impacts on groups within the population. That is another aspect of the difficulty of trying to manipulate prices for staple food in the market.

More examples can be given. Suppose there is an increase in food commercialization. There is a substantial difference between the value added on the farm and the total value added to the final price the consumer actually pays, because a good deal of the cost of food goes to the middleman. Also, commercialization and food-processing comprise a large share of the economic activity in poor economies, perhaps something on the order of 10 per cent.

It is, of course, traditional to haggle with the middleman and one may, in some cases, be successful in reducing food prices that way, but there are limits. For example, most food in Brazil is delivered in trucks. It doesn't take a lot of money to buy a truck in Brazil. You can get a loan from a bank and buy a truck. Many people are in the trucking business and they are probably fairly competitive. Food truckers are not making excess profits that would be easy to cut back. Moreover, if the truckers' productivity increased markedly, many of them might lose their jobs. That would lead to political problems and a need for policies to deal with the unemployment.

There is another example. Suppose food imports are reduced. The International Monetary Fund steps in and says that 200 million dollars a year is too much to spend on food imports, and lends only 100 million. How does a country deal with that? Typically, most countries have a government marketing corporation, and it would go to the world market and buy less food. The internal effect of that decrease in food importation would be to drive up the price of food sharply, at least for those who participate in the market, and shift income distribution toward agriculture.

Real income in the rural areas would go up, probably more than the increase in the food prices, and so food consumption there would go up. Food consumption in the urban areas would, of course, go down sharply as the price increase rations food away from poor people, and once again there would be political problems. Given the importance of food, the political price of import reduction will be high unless something is done to offset it. The obvious remedies like rationing are hard to enforce, and could lead to problems.

To look at the problem somewhat more dynamically, fluctuations in food prices and supply can clearly be a serious policy issue, and the distributional impacts and the unanticipated price changes that occur when there is a short crop or a poor season can be severe. In that case, policy options that are probably worth pursuing and that, in fact, many countries have pursued, come into consideration.

For example, stock-piling to maintain prices in consumer real income, a policy more or less followed by CONASUPO in Mexico, and by similar agencies in other countries, is presumably a good thing.

Making food markets "thicker" in the economists' sense, that is, with better-financed traders, large volumes transacted, etc., may make sense in a number of countries that follow that kind of policy. Or, the problem may be addressed by improving food storage conditions. Losses in some cases are of vital importance. In India, for example, I am told that very simple technical changes reduce the loss of grain in family storage bins by as much as a two-week supply. That is, some simple modifications of existing procedures can save two weeks' worth of grain for family consumption by preventing deterioration or predation by rats. If that were done, saving a two-week supply of food at the time just before the harvest comes in could be tremendously important for a poor family. It can keep them out of the hands of the money lender, for example. Technical changes of this type can be of considerable use.

Obviously, all these macro-problems really come from the instability of the food market and the fact that intervention on behalf of one group is almost bound to hurt another group. If both groups are in need in the nutritional sense, a fairly difficult situation emerges. If it is assumed that these are issues for economists to deal with, and health professionals are more interested in micro-problems, some of the solutions presented in the paper by Solimano and Jeria and elsewhere in these Proceedings must be considered.

To mention a few solutions from the economists' point of view, groups that benefit from a food price may or may not be those whom the physicians say are nutritionally at risk. If food distribution within the family is inequitable, it is not clear that any policy, at reasonable cost, would have much influence, and yet nutritionists put a great deal of emphasis on just this issue.

More generally, in looking at families as families who are at risk in some sense, the problem of identifying them can be rather difficult. A search procedure has to be instituted, and in underdeveloped countries this is always very costly. An alternative would be to use income as a measure of the position of the family that needs help. This causes problems with a means test-that is, actually finding out how much income people have. If this is done by means of the bureaucracy, it is almost always both inefficient and demoralizing.

In rich countries like the United States, there are vindictive welfare case-workers; another example is George Orwell's England before the Second World War when the means test narks were grossly unfair. With the bureaucracy in many underdeveloped countries, I would not be very optimistic about any means test they might run.

In fact, the approach most countries have used when they have actually tried to deliver food to the poor or the malnourished has almost always been a subsidy scheme of some kind.

Some of the macro problems that food subsidies can generate, or are subject to, have already been discussed. However, various other possibilities exist. For example, in Pakistan they happened to stumble on the policy of subsidizing a kind of wheat flour that is considered inferior by most people in the population. Ration cards were distributed that allowed people to go to a ration shop and buy five kilos a month per person of this flour, even though it was considered to taste bad and to spoil easily. The only advantage was that it was very cheap.

The characteristic of a system like that, of course, is that poor people select themselves to go into the ration shops, while rich people prefer, by and large, to buy their wheat elsewhere. From the nutritional point of view, because the cheap flour is, in fact, quite acceptable, that may be a good thing. From the social welfare point of view, it may or may not be a good thing; it is not clear that one should subsidize poor people's food and thus label the people who buy it as poor. Nevertheless, the policy has worked more or less effectively.

One can think of similar examples, such as inexpensive vegetable protein that rich people tend not to eat, though poor people do. If vegetable protein is subsidized and meat is taxed, the policy is likely to use market distribution channels in a way that can benefit poor consumers, at least in urban areas. Ideally, the food that poor people tend to consume could be subsidised intensively, because the price elasticity of the poor is high while that of the rich is low.

Shlomo Reutlinger

My first observation is that economists' models illustrate, for those outside the profession, that economics is really a dismal science. When all sides of a problem are looked at, there seems to be almost no solution. I hope that this is true only for interventions at the commodity or market levels. There are many indirect effects from policies, and many of the positive effects achieved are cancelled by the indirect ones.

My next observation is more critical. Taylor's model illustrates that these kinds of models are of very little use for explaining economic development, and they also are of little use for analyzing how to intervene to improve nutrition in a country. This is particularly frustrating, because hard as you try to capture some of the more indirect effects in the economy and trace them to their causes, there is still so much that is overlooked. On the other hand, if too many details of the entire economy are analyzed, i.e., what is the right model for the whole economy, then not much attention can be paid to direct effects of the model on the particular problem to be solved.

In many cases, some of the indirect effects of a policy that can be traced through this model are not addressing many of the indirect effects that arise from its application. For instance, to cite one example, there are three classes in Taylor's model, but when it comes to measuring the nutrition problem or the improvement hoped for in a population, concern is not just with all workers, or a/t farmers, or all of the rest of the people in the country. The first thing to do is to find some criteria for discovering who and where the undernourished people are. This requires a much finer breakdown than offered by these three classes, which cover up the interesting aspects of the food distribution problem as far as it affects particular groups of people. However, it is too complicated to consider individuals because policy cannot operate at the level of individuals. Also, some of these indirect effects can be studied without having a complete macro-economic model that covers everything. My own view is that much more needs to be known on a macro level in order to identify the problems and their ramifications and to learn their extent. It is also necessary to know how well interventions work at the micro-level.

First of all, in many Latin American countries, as much as 30 per cent of the nutrition problem is in the urban areas, at least in normal years. Particularly with populations migrating to the cities, there is a food problem that can be affected by changing prices of food and other commodities.

Second, within the rural areas, specifically when studying data from Brazil, most of the production is not in the hands of farmers who are undernourished, but rather in the hands of farmers who employ farm workers. It is not at all clear to me that higher farm prices do much to improve nutrition for most of the world's undernourished population.

My general conclusion is that I concur with Taylor's common-sense approach. There is very limited scope to policies working through commodities, i.e., changing prices or even improving the agricultural sector in general, even by improving the technology for production. I believe there is a much more limited scope for nutrition improvement than we often think. That does not mean that there is no hope at all, but it is limited because essentially the problem is one of low income and very uneven income distribution.

Therefore, many policies other than food policies must be considered. It is peculiar that people always think that a high price for beans would be a good thing for nutrition. They mean that there are many poor bean producers who would benefit from a higher price, but first of all, are so many bean producers poor? Second, if income is the problem, why raise the price only for beans? I am merely taking an example out of the air, but perhaps cotton or coffee production could create more employment and more income for people than growing beans generates. If more employment in rural areas is the goal, then a focus on beans or even on agriculture in general is too narrow. Policies that adversely affect income distribution outside the agricultural sector, or within industry, or between agriculture and industry, must be changed.

I agree that it is very difficult to identify interventions that are specifically target oriented. It is hard to implement them, to monitor them, and so on. All of the programmes I have seen-such as ration shops with lower food prices, feeding workers on the job, food programmes in schools-may not do anything more than transfer income to people. Politically, it may be more acceptable to give food to children or to give rationed food, but as far as nutrition goes, there is still the plague of this low elasticity discussed earlier.

Basically, what rationing does if 50 per cent of the food a family obtains costs nothing, or is sold at half price, is to save the family money on its food bill. It is not at all clear that the family is going to buy as much food as it bought before and use this additional food as such. In effect, the cheap food becomes an income transfer, but what is really needed are programmes that have more direct nutritional effects than just an income transfer. This is because perhaps 70 or 80 per cent of the income transfer is not spent for better nutrition; it may partly pay for food, but it may be spent on very high-cost, status food calories, or on many non-food items. Therefore, these programmes are likely to be the same in that respect unless ways can be found to make them work better. Perhaps there is some educational effect if, rather than income, foods or certain kinds of food are transferred, but this also needs to be studied carefully.

These special, or even target-oriented food programmes, are good to the extent that they produce income transfers; at least in that sense they are more efficient than are general food price policies or general food policies.

Eduardo Kertesz

I should like to begin where Taylor stopped, by saying that there is a general lack of opportunity and vitality in much of what is proposed in the policies and programmes being discussed. Solutions to the problem of malnutrition have evolved from an initial concern on the part of physicians and nutritionists with what kinds of food are eaten. If malnutrition resulted from an insufficient and unbalanced diet, the solution was to find more nourishing foods, but this often led to a rejection of the staple diet.

Next came the agricultural approach-to increase food production by using more fertilizer and other techniques that had brought about the Green Revolution This was an overemphasis on agriculture in a market economy, where people were suddenly expected to grow food for export instead of for local consumption. It did not transform the economic background of the people or the capacity for demand, much less income distribution.

Later, economic variables became the centre of attention. Problems of income generation and inequitable distribution of wealth began to be discussed. Unfortunately, this type of focus is still limited because the solutions proposed are presented as though they could be implemented in the existing social and political structure, accepting all the givens of the system as it is. If programmes are begun in this framework, there is the danger that the bureaucratic manipulations of the administration would be imitated; i.e., as one benefit is offered, another is taken away. In other words, the status quo would be maintained, and nothing suggesting any radical change in the political structure would be forthcoming.

The economic approach for solving malnutrition, while it represents an advance, is still very limited. Underlying the economic scene is the political structure, the question of class struggle, and of various sectors of the population that are excluded from social development. Most of the current analyses concentrate on manipulating existing methods without considering the political realities that support the status quo. The results of these analyses may prove useful to the multi nationals that produce food, helpful to us who are studying the problem, and to industrial research, but I doubt whether anything that has been proposed so far would have any real beneficial effect on the majority of the population.

To comment specifically on Taylor's paper, his is, first of all, an approach that does not consider changing the political system. What is emphasized is the search for the ideal intervention model, not better foods. Both economists and nutritionists recognize that under-nutrition is closely related to poverty and lack of buying power, which, in turn, reflects lack of any other kind of power.

Intervention models are often devised to be so abstract that they will apply to any country. In my opinion, the history and social conditions of a country at one point in time must always be considered in intervention programmes. Abstract models may make interesting mathematical exercises and theses, but I doubt that they help to explain the reality behind a country's food and nutrition problems.

I also think basic budgets should be emphasized. For example, fertilizer use by small producers requires money or credit. As Machicado pointed out, these producers use hardly any fertilizer; therefore, the importance of the latter in producing basic foods in Latin American and other developing countries is a fundamental issue that does not pose a problem in the United States. Budget is the determining factor in food consumption, and price instability obviously affects the budget.

I do not know whether the food-producing system in Brazil is similar to that in Jamaica. The latter depends wholly on food exports, but Brazil should be able to feed its own population and export food as well. However, in Brazil there is a modern, dynamic industrial sector that is entirely separate from basic food production, which has become isolated from the mainstream except for export crops. There is more than enough manpower; indeed, that is part of the problem. The poverty in the rural sector means high mortality rates, but this has no repercussion on the economy in the industrial sector. There is a staggering army of unemployed people in Brazil and throughout Latin America, and this is characteristic of all Third World countries undergoing rapid industrialization. Replacement of manpower by machines in these countries means that the labour force need not be large, therefore a large sector of the population is left out of the system.

Those who produce basic food crops have very little influence on the rest of the economy. These very low-income groups are locked into their poverty, and what is worse, their exposure to advertisements for cars, television sets, and processed convenience foods creates a desire for goods that are not affordable without going into debt. Food habits also change, to the detriment of nutritional status. This is apparently of no concern to the rest of the country.

Lance Taylor

I would like to avoid getting hung up with models. In response to both Reutlinger and Kertesz, it is fair to say that certain economists find it useful to communicate with each other in terms of strong, simple models that give straightforward conclusions. Everybody is perfectly aware that the world is much more complicated than the model. Nonetheless, it sometimes helps to clear away cobwebs in thinking and make a point strongly, which is the reason why I developed this one.

I think that the more interesting question, if we are really concerned about making policy in a given country, is how much of the model exercise will carry over. In particular, how would a model be useful in analyzing that set of issues discussed at this Workshop? The main point of my paper and my discussion is that staple food consumption has its own characteristic market features and comprises so large a share of economic activity that it represents an important restriction on thinking about policy in poor countries, especially if an active, as opposed to a passive, food policy is desired. If an active policy is to be pursued, I think that the only way to design it would be through use of a model like the one I suggested. Or better, through a model of the "general family" that takes into account your own perception of the reality of the country that concerns you.

It is fairly clear in the paper that I was not dealing with any specific country, but there are features in the model that would carry over into a given country exercise, while others would not. But in any country, it is unfortunately true that economic scarcity applies in the sense that, if a policy of redistribution is being considered, there is only a certain amount of surplus generated in the system that can be distributed. For example, Kertesz indicated that there are large numbers of workers and that food production is isolated from the mainstream of development, but he did not provide numbers. Unfortunately, the only way to make a sensible judgment about these issues is to try to quantify them, and in that case one has to work with the model, preferably a simple, understandable model. Certainly, a model that includes cotton growers, fruit growers, and bean growers would not do because the morass of details would be overwhelming and the macro-issues important for any real income redistribution would be blurred.

Kenneth Leslie

There is a hidden assumption in Taylor's paper. While it seems to imply that it is dealing with an open economy, much of the discussion is about a closed economy. Nowhere does it take special consideration of the fact that there is international trade. As Kertesz pointed out, in many of the developing countries a considerable proportion of the food consumed is imported. The price paid for this food is not directly set locally; external market determines the price. A closed economy model seems to be extremely inappropriate in light of this, particularly in the developing countries where, in many instances, they are highly dependent on imported food supplies. Because of this, they are left in a position where the only thing they can do is to subsidize food because they cannot control the prices of food coming into the country. Unless this problem of terms of trade for commodities is addressed (and this is not just a domestic eco nomic problem), a more deeply rooted problem is ignored. This is what most of the Third World has in mind when talking about the New Economic Order, which I would think should have caught on by now.

Lance Taylor

I agree that this is fundamental for the whole issue. It must be recognized that subsidy programmes are, in a sense, palliatives that governments are forced into because of the very limited power they have over the prices of commodities they import.

Flavio Machicado

I should like to return to the subject and scope of Taylor's model. I think it is good to have models, but sometimes they are dangerous if not used correctly. I think it is wrong to set up a model to explain a policy such as prices. I believe it is essential first to propose an interpretation of the rationale behind the economic systems in Third World countries, and it is my belief that here one can generalize somewhat.

It is my impression that the subject of price policy has not been touched, at least from my perspective, because I understand there will be a price policy only when it is possible to fix a price. The rest is simply the interplay of supply and demand, not a price policy. When Taylor considers a subsidy or taxes, in my opinion he is not talking about a price policy, but rather a tax policy that has to do with fixing prices. Prices can be lowered only by means of taxes.

Right now in the Dominican Republic, corn is being imported, which will affect a tremendous number of corn producers for whom the nutritional policy has to be applied. Consequently, I believe still another element must be added. A policy of price fixing is used in order to promote a massive increase in food consumption. In the Dominican Republic, 75 per cent of the population is undernourished, and 50 per cent of these are in a critical state of malnutrition. Their average daily intake is 1,500 calories. Where is this population located? In the agricultural sector, especially farmers and rural workers, who represent 50 per cent of the population. Other undernourished people belong to the marginal urban population, who are outside the market system. All this must be explained to understand what agricultural dynamic is needed to solve an apparently contradictory problem.

The dual economy is one further problem. For instance, in Chile, during the last two years under the Popular Unity, a price-fixing policy could not be applied. This was actually not a price-fixing policy, but consisted of estimation of costs and price levels based on the rate of inflation. The time came when fixing the price of wheat had to be considered at the planning office. This price fixing, with a 19 per cent rate of profit, was of the order of 500 escudos per 100 kg. (These may not be the exact figures, but they are close enough.) When the producers were asked if this seemed feasible, they agreed at first, but when the time came for them to sell their wheat at that price, they could not do it. Why? Because within the dual economy, they could get from 1,200 to 1,500 escudos per 100 kg due to speculation on the market. There could thus be no price-fixing policy because of the nature of the system for fixing prices.

In the case of the Dominican Republic, what has been called a price policy has consisted simply of the Government's intervening in order to stabilize prices through setting a price on certain agricultural products. But it is impossible to propose such a policy for another important reason. When prices are fixed, we are unfortunately still faced with a non-homogeneous production structure, which is by its nature inequitable. The fact is that, in the Dominican Republic, 60 per cent of the milk produced comes from small producers, and 40 per cent from large ones. Prices are set because of the pressure brought to bear by the large producers and are calculated on the basis of the cost of an agricultural economy that represents only 40 per cent of all milk production. The situation is further complicated because large producers want to recover their investments in only two or three years. Therefore, one would think that by fixing prices, the small producer might receive excess profits. However, the commercialization system is designed so that small producers cannot receive this excess. This is because of the absence of a basic commercial infrastructure for daily products. The Government is thus faced with the choice of being fair to the large producers and unfair to the whole of society; if it chooses the latter, it must fix a price at the behest of large producers that the economy cannot afford.

Alberto Carvalho da Silva

A point I would like to make is that it is important to consider the number of food commodities for which interventions should be made. Ideally, interventions should be restricted to one or two food commodities, but this is difficult for both biological and social reasons. The biological ones are that some foods are consumed at practically the highest levels the human organism can tolerate. This is the case, for example, with manioc in rural northeastern Brazil. The same is true for beans in certain areas where the bean is an important source of protein.

Another comment relates to the idea of switching food-producing farmers to other, more profitable, crops in the agricultural sector, and importing foods. This would be practically impossible because the income, and hence the purchasing capacity, of the food producers is so low that they could not possibly afford to buy food for their own consumption. Also, they could probably not establish themselves or be accepted in the capitalist agricultural sector. It would not be possible to raise food costs above certain limits because low food cost is the very mechanism by which the poor sectors of society survive at all. In fact, and I imagine this is true for all poor Latin American countries, it is the poor in the rural society who support the poor in the urban sector, another reason why it would not be possible to raise the prices of food.

David Lehmann

I would like to separate and identify some of the issues discussed because there is a whole series of different issues in different frameworks whose purposes are not always clear. First, Taylor did not kill the model question, and perhaps I can help to do so. It is impossible to make any kind of analysis of any social process without having an implicit model in mind, not to mention certain kinds of implicit value judgments.

Machicado's comments were right on target when he said that the question that must be asked is: "What is the rationale behind an economic system?" I do not know what kind of a model lies behind his concept-perhaps he will clarify it in the manner that Taylor enlightened us on his. The trouble is, of course, that in discussing models in a more technical sense, it was not sufficiently emphasized that what comes out of a model depends on what goes into it. It is not a question of right or wrong, but I think it should be recognized that we have been treading, epistemologically, in very muddy waters.

Second, I think trouble arises when models really try to deal with the feed-back effects mentioned earlier. The models then become so complex that it is hard to understand why they produce the results they do. Therefore, either a model is clear enough to discern what is happening inside it, but it turns out not to be realistic, or a model is so complex that it is impossible to understand the results it produces.

Moving on to the main issues, I think we are dealing with different areas. The first is a fairly technical area in which more information, perhaps just an account of existing results, would be very useful. I am asking that commentaries follow in the steps of Machicado and Carvalho da Silva and discuss the results they have had with first-hand experience. Otherwise, the discussion will be completely up in the air. (By this I do not mean that I am against either abstractions or generalizations, but it is good to have solid information to support them.) For example, there is the question Reutlinger mentioned briefly that seems to me very fundamental, namely, the relationship between incomes and food consumption. Specifically, John Wells showed that in Brazil urban incomes are going down and food consumption is also decreasing.

Lance Taylor

The as yet unpublished research by John Wells at Cambridge University complements similar work by Paulo Singer and others at CEBRAP in Sao Paulo. Both investigations are based on past consumer budget surveys in Brazil, and seem to show that, over time, real income and also the food consumption level of the urban poor have fallen. Also, additional "necessary" expenditures on items other than food, such as transport to and from work, have grown enormously as a share of family expenditure in places like Rio and Sao Paulo, and this has pushed down real food purchases.

David Lehmann

My impression from these papers is somewhat different. I infer that people were actually reducing their food intake but increasing their consumption of durables. This is the classic behaviour of the typical Third World man. Instead of being seduced by televisions and refrigerators, why aren't they buying more food? In fact, of course, if you think about it, it is not so stupid because, especially in inflationary situations, refrigerators, televisions, and motorcars keep their value. It is a way of saving, and if very unstable times are expected, or incomes are marginal, then actually this is a means to ensure future income by buying a television that can be resold later. It is not the individual who is irrational, but the economic system.

This relationship is a complicated one that should be dealt with in more depth. It is often assumed in this discussion, though Leslie pointed to the problem, that agricultural production and food production are the same, which they are not, and this is one of the main problems. Much agricultural production is devoted, not to food for domestic consumption, but to coffee or sugar for export. In connection with that, I would like to know if any work has been done on the relationship between agrarian structure and agriculture (and, as part of that, food production) and the distribution of income in terms of the different types of foods that different income groups consume.

For example, my guess is that large commercial farms often produce relatively expensive nutrients, especially proteins. Cattle and wheat are produced on large farms using capital-intensive technologies, and are consumed by middle and upper income groups in poor countries. Smaller farms use labour-intensive technologies to produce cheap proteins like beans, which are consumed, on the whole, by the poorest income groups in agriculture and industry. This kind of issue needs verification, as does, of course, the issue of whether producers and consumers are the same kind of people. Sometimes people who produce food also buy it for two reasons: 1. They may have to buy certain foods they don't produce in order to balance their diets, and 2.,they are forced to sell food at some time during the year because they are in debt or short on income, and then later in the year, have to buy it back. These are added complications in our implicit models of what is actually happening.

Another technical issue that needs to be addressed is the scope for what I call limited intervention, because the discussion has covered largely global types of intervention such as price policy, etc. For example, I have recently been reading about iodine deficiency and thyroid disease and, as I understand it, prevention does not involve large-scale price policies. This is really a specific deficiency that could be remedied by a simple, targeted intervention, i.e., iodizing salt. It seems that severe iodine deficiency in the very early years leads to irreversible deficiencies in neuromotor functioning.

There is a whole series of other issues that I call food systems issues because they are very interconnected. For instance, the role of food as a wage-good in the whole industrialization process is quite complex, especially under different economic and social systems with the different types of interventions they bring about. Also, of course, there are conflicts of interest among different social groups and the way they act politically over food price policy. Although we can tell, to some extent, what these phenomena do to different social groups, we cannot easily forecast how those social groups will organize to do something about them. The more powerful they are, the better they are at organizing, although history shows that this is not always the case.

Finally, there is the whole question of agrarian structures that I referred to just briefly when I discussed the different types and sizes of farms and the possibility of changing them to improve food production. The experience of land reform in Latin America has not been very heartening.

Bo Wickstrom

It seems we are all eager to make one comment about the model presented by Taylor. I think it is a beautiful model, with much internal logic in it, because of the data in it. What is a model like this used for? It can serve as a pedagogic model for a meeting like this to generate discussions, and it has been very successful in doing so at this Workshop. It can also be an instrument for decision-making, which is the real purpose of a model. I am not quite sure about the validity of this particular model, because it was not discussed in the paper. I think we should discuss the circumstances under which the model is valid. Perhaps it would be good for decision-makers in some countries but not in others. I feel that one should include individual food items with their costs, demand elasticities, etc., in a model. It would be much better to look at the whole assortment of foods and observe what happens in terms of price and consumption. If the price of only one food item is changed and people consume another in its place, the effect is nil if the food consumed is nutritionally adequate. That fact should be taken into account.

I do not know whether food price elasticities are always less than one in low-income countries. I recall seeing some data from a cross-sectional study in Tunisia where both calorie and protein demand were normally elastic in lowincome groups.

The delivery system should also be taken into account in determining the practicality of a model. If the delivery system for farm inputs like fertilizer is satisfactory, there is no need for a large extension service to reach small farmers. Similarly, a good delivery system for farm products will have a favourable effect on prices, and price policies can have more impact.

My field is marketing, and i have noticed a growing interest in what is called "social marketing." As Lehmann said, if you want to change people's consumption habits, price elasticities have to be changed to make people switch from one product to another. Industry has a great responsibility in this, because very active marketing techniques will influence people, sometimes adversely if they substitute a less nutritious item for a good food product. In this case, whatever they pay, they are not getting their money's worth.

Eduardo Kertesz

To return briefly to models, I would like to emphasize that my earlier criticism was not of all models, but to call attention to the fact that the model to seek should not be like the current search for a blanket cure for all kinds of cancer. No model can offer a magic solution, because behind all models and socio-economic questions are the political ones. A model developed for the purpose of intervention should first analyze the political variables.

Good intervention models are developed to solve nutrition problems on the presumption that strong political will exists to decide for improving nutritional status. It must be remembered that in underdeveloped countries, the proportion of the population requiring assistance amounts to 50 to 60 per cent of the total. Neither a subsidy policy nor a direct intervention policy can solve problems of this magnitude on a sustained basis. The case is very different in developed countries, where a smaller percentage of people needs this kind of assistance.

Taylor was surprised to learn that the food-producing sector has such a strong influence on other sectors that the latter must adjust to it. I do not believe this is, in fact, true in many countries or among individual families. A family's budget is based on consideration of their fixed expenses, such as maintaining their home, their means of transportation, etc. Money for food consumption is apt to be whatever is left after fixed expenses are met.

Lance Taylor

To summarize, I would first like to return to the points Leslie raised regarding the importance of foreign trade. It is certainly true that many food-importing countries -Jamaica is a very obvious example-have really been hurt by the change in the terms of trade against them, the rise in the cost of food. One can deal with this in a policy (and a policy model) in much the same way as described in the paper. Suppose the terms of trade go against a country's balance of payments At the same time, purchasing power is lost to the trade deficit forced upon the country. Purchasing power has to be generated internally by running at a bigger fiscal deficit. What better way to run a government deficit than by maintaining the price of food? That is precisely what many countries have done. But it is a difficult course to follow because of another kind of outside pressure, basically from the International Monetary Fund and other agencies that try to push countries into lower deficits. That can lead to serious political problems, as for example the food riots in Egypt in January, 1977. I suppose I can claim that part of the power of thinking in simple macro-economic terms is that I used a model similar to the one in my paper in June, 1976 to say that cutting the food subsidy in Egypt would be foolish. I was the only economist to say that, so maybe models can help one to be right at least some of the time.

On the issue of consumption patterns, it appeared from several of the discussions that very poor people, because they devote a very high proportion of their budget to food, will have a more elastic response to price changes than will people who are somewhat better off. If, as is typical at the very bottom of the income distribution, 75 or 80 per cent of a family's income is devoted to food, and the price of food goes up, the real income loss is so strong that food consumption necessarily has to fall substantially. Price inelasticities appear on average, however, across the whole population. But that makes food price policies even more important because the people who are really hurt by food price increases are not the ones whose price elasticity is .2 or .3, but the ones whose price elasticity is .7 or .8. If the average elasticity is .5, this raises macro problems. The .8 elasticity at the bottom of the income scale raises real social problems. A poorly used macro-modei may lead you to forget that sort of issue, but in fact, it is very important.

One last comment on Lehmann's point that in thinking about any real country, the class and political structures must be considered. That will vary from country to country, and certainly the historical dominance relationships will vary among the classes. In some sense that provides another set of variables requiring detailed information that has to be taken into account in the analysis. The fact that a political structure exists does not mean that asking macroquestions about macro-problems is wrong. What is needed is to ask the macro-questions in terms of the class structure of the country, and in a generalized paper, that is impossible to do.


1. J.M. Davis, "The Fiscal Role of Food Subsidy Programs." International Monetary Fund Staff Papers, 24: 100 127,1977.

Contents - Previous - Next