Contents - Previous - Next

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

Determining the beneficiaries of cheap-food policies in Trinidad and Tobago

Carlisle A. Pemberton and Emaline L. Harries



Solving the problem of undernutrition is one of the foremost challenges continuing to face developing countries. Difficulties of quantifying undernutrition precisely have resulted in widely varying estimates of the number of undernourished persons, ranging from fewer than half a billion to almost two billion [1]. Regardless of the differences in the figures, it is clear that hundreds of millions of people are nutritionally deficient. Moreover, the vast majority of the undernourished are in developing countries.

Nutrition strategies that have been devised over the years have had one principal aim: to increase the food intake of undernourished persons. These strategies have generally proved inadequate to the task, however, and therefore the theories of the nature and causes of undernutrition have been revised and new approaches are being suggested.

Undernutrition was once seen as wholly a problem of food scarcity. Consequently, the solutions presented were increases in food production and productivity. The valuable lesson gained from the experiences of countries that actually made progress in the Green Revolution was that increasing food supplies alone was not enough. Efforts must also be directed to giving the undernourished access to the food supplies. The consensus now being expressed is: "Rather than as a race between food and population, the food equation is to be viewed as a dynamic balance in the individual countries between food supply and demand that depends on complex relationships among a number of interacting variables" [2].

Current approaches to combating undernutrition fall into two main categories - food policy and nutrition policy. Food policy is the more comprehensive approach, which seeks to address the problems of the food system as a whole. The rationale is that the imbalances of food production, marketing, and demand are closely linked and require an integrated attack on all three fronts. Nutrition policy is limited to the concerns of directly increasing food consumption.

Among the experts, there is a preference for the broader rather than the narrower approach. Given that undernutrition is intrinsically tied to poverty, it is felt that a sustained solution of nutritional deprivation hinges on the elimination of poverty itself. As postwar development strategies have proved very slow in overcoming both poverty and undernutrition, some advocate an alternative development thrust that emphasizes the role of the food sector and agriculture as a whole [2]. An explicit incorporation of nutrition goals into overall development plans is also considered desirable. It is advisable that this path should be taken in order to achieve both rapid national development and the sustained solution of undernutrition.

So far, developing countries have more readily adopted direct nutrition interventions such as cheap-food policies, food stamps, and food-supplementation programmes. One reason for the preference for such policies is that they are more visible than nutrition schemes hidden in agricultural projects or development strategies. The rationale is two-sided. in that the recipients of such welfare-oriented policies may be more gratified at the visible concern for their wellbeing, and policy makers benefit in their efforts to mobilize financial and political support [3]. The Egyptian food riots of January 1977, which followed a reduction in consumer subsidies, showed clearly the political importance of these policies [4].

This paper focuses on means of determining who are the beneficiaries of cheap-food policies. The approach is applied to the case of Trinidad and Tobago, to identify the beneficiaries of the cheap-food policies implemented in this country from 1973 to 1985.


Cheap-food policies in Trinidad and Tobago

Two policies are widely used in an attempt to lower food prices to the consumer. First, there is the regulatory measure of price controls by which policy makers fix a maximum price that retailers cannot legally exceed. The second is the food subsidy. This method involves a financial flow from government to producers and/or retailers in order to cover the difference between the market price of the food item and the desired lower price that the consumers actually pay.

Trinidad and Tobago has implemented both price controls and food subsidies. These policies were first utilized as far back as 1948 [5]. By 1955, however, the subsidies were withdrawn and a number of items were deleted from the schedule of controlled prices. The present system of price controls comes under the Trade Ordinance of 1958 and subsequent amendments. In 1973 a food subsidy policy was again introduced.

The schedule of maximum food prices includes a wide range of basic food items: milk, bread, butter, margarine, cheese, cooking oil, fish, beef, poultry (local and imported), flour, rice, white potatoes, onions, peas and beans, canned sardines, and sugar. These controlled prices are reviewed periodically and amended. Local producers and retailers may request price increases on the basis of increases in costs of production or prices of imported items.

A two-tiered pricing system is in effect that permits retailers in specified areas outside the main urban areas to charge higher prices. The justification for this is to assist retailers in the higher-price districts to defray expenses for additional transportation costs.

Food subsidies were paid on many of the commodities under price control. The majority of these subsidies were administered by the Ministry of Industry, Commerce, and Consumer Affairs. The Ministry of Agriculture has been responsible for the subsidy on one item, milk. Table 1 shows the list of foods subsidized by both ministries and the payments made from 1973 to 1985.

TABLE 1. Food subsidy payments in Trinidad and Tobago (TT$ millions). 1973-1985

  1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985
Rice 3.3 8.8 9.2 18.2 7.1 10.3 11.9 19.4 18.5 17.8 4.1 7 8
Flour bakers 0.7 1.1 1.3 1.3 0.9 0.9 0.8 0.5 0.5 0 4
other 13.0 13.4 19.5 9.9 8.1 15.2 32.4 31.7 13.8 9 3    
Cooking oil 0.2 0.5 1.9 1.9 2.4 2.7 3.6 3.7 4.0 0.1
local 1.4 5.8 6.5 6.4 5.5 7.4 9.9 11.0 8.1 7.0
imported 1.6 1.5
Butter 0.3
local 55.0 79.1 87.4 201.8 126.0  
imported 2.4 2.4 17.1 17.0        
Onions 0.04
Milk 0.5 0.6 1 0 0.5 0.5 0.5 0.6 0.5 5.3 3.9 5.0 7.8 9.1
Total 3.8 22.6 24.1 42.24 26.3 31.5 97.4 160.0 171.9 252.0 156.0 24.2 16.5

Over this period. TT$2.40 = US$L
Sources: Ministry of Industry. Commerce and Consumer Affairs: for milk subsidies. Ministry of Agriculture. Lands and Food Production.

The figures in the table show the expansion of the subsidy programme between 1973 and 1981. Both the number of items and the size of payments increased steadily up to 1981. By 1981 subsidies were being paid on rice, flour, cooking oil, poultry, sugar, and milk. Onions and butter were each subsidized for one year only, in 1976 and 1979 respectively. Subsidy payments rose from TT$3.8 million in 1973 to TT$171.9 million in 1981.

The growth of the food-subsidy programme from 1973 to 1981 was paid for substantially by the greatly increased petroleum revenues of the government of Trinidad and Tobago. The sharp decline of subsidy payments after 1981 is also explained by the fall in international oil prices and a resulting drop in government revenues. In the space of four years, subsidies were reduced to less than one-sixth of their 1981 level. The only items that were still being subsidized in 1985 were flour, poultry, and milk.

From the government's viewpoint, cheap-food policies had the general aim of keeping the cost of living down, as well as the nutritional intent of ensuring that consumers could purchase food items. The price-control policy was designed as a means of directly containing food price increases. Subsidies were also used to keep down the prices of basic food items for low-income groups. The ultimate objective was to en able consumers, especially those in the lower-income groups, to buy their full nutritional requirements.

The foods studied in this paper - rice, flour, cooking oil, poultry, and milk - are all major items in the local diet and were all subject to price controls and subsidized from 1976 to 1983. Together, the food groups of cereals, oils and fats, meat, and milk accounted for 88% of the total calories excluding sugar consumed in 1981/82 [6].


Analytical framework

Cheap-food policies are frequently universal in coverage rather than restricted to certain consumers. One way of applying a discriminating policy is to select items that are more readily consumed by the target income group. Such a policy has a built-in targeting mechanism. The universal policy. on the other hand. is available to any consumer who enters the food market; hence its impact depends on the responses of consumers [7].

Universally applied cheap-food policies have several disadvantages. They are susceptible to leakages, benefits enjoyed by consumers who are not in need of public assistance. Since public cost is a direct function of the size of the group covered, broad-based policies tend to be very costly. There may also be conflicts between governments' efforts to maintain low food prices for consumers and to provide incentives to domestic food producers.

To consumers, the benefits that may by obtained from cheap-food policies are of two kinds: nutritional and financial. Therefore, in considering the effectiveness of policies, a relevant question is whether the beneficiaries are in nutritional need. Conversely. it may be asked whether the government is spending limited public funds on those who are able to buy all their nutritional requirements without policy incentives. To answer these questions, it is necessary to examine the market responses of consumers to price incentives on food items within the policy.

It is evident that those who are financially deprived are also those at greatest risk of being undernourished. Consequently, low-income consumers are the appropriate target of nutrition policies. Subsidizing the diets of high-income consumers cannot be justified on either financial or nutritional grounds.

The nutritional benefits of cheap-food policies go to those who are enabled by lower prices to increase their food consumption. Consumers who are sensitive to price changes and who function within the framework of rational decision making will increase their consumption as food prices decline. Since the consumers can determine the composition of their expenditures, the impact of a cheap-food policy rests on their choices between food and non-food items. If price does not enter into their decisions about food purchases, such policies will be ineffectual as nutritional measures.

The financial benefits of cheap-food policies accrue to everyone who buys food. The benefit is the difference between what consumers actually pay for their food basket and what they would have paid if there were no such policies. Hence, the financial benefits increase in direct proportion to the quantity of food purchased. Consumers who buy the most food get the greatest benefits.

It must be emphasized that the financial resources transferred by cheap-food policies may or may not be used to buy more food. Consumers allocate the financial benefits according to their own personal objectives, which may not be consistent with those of the policy makers. Once policy transfers are in excess of the amount of food the consumers are willing to buy, the financial benefits are treated simply as income transfers and may be used to buy non-food items.

Price elasticities that are income-specific are useful indicators of the food-purchasing behaviour of consumers within a particular income group. They show how the consumers determine their pattern of consumption to satisfy their own objectives. As a result, it is possible to predict the likely impact of cheap-food policies on the nutritional well-being of consumers in different income groups. The estimation of income-specific demand functions for food is, therefore, a vital means of identifying the beneficiaries of the policies.

The analysis in this paper is based on an income-specific empirical demand model. The test for nutritional beneficiaries was carried out on the basis of the significance of price elasticities. The rationale is that an income group with a price elasticity for a certain food that is not significantly less than zero is not price-responsive. The members of such a group would not benefit nutritionally from a cheap-food policy as they would not increase their purchase of the food. In contrast, a price elasticity that is significantly less than zero indicates an income group that is price-responsive. The members of this group would benefit from a cheap price for this food.

A further test was done to find out whether there was a significant difference in the overall pattern of consumption by consumers in different income groups. Such a test could be used to confirm whether it is possible to target certain foods to specific income groups. The existence of a significant difference between income groups would indicate that the groups respond differently to price factors as well as other variables. The Chow test was used for this purpose [8]. This test involves the calculation of an F ratio for the two regressions of the income groups being compared. The hypothesis that the two regressions - and by inference the responses of the two income groups - are the same is rejected if the computed F ratio exceeds the critical F value.

A corollary of the general analysis is the identification of those income groups that would incur loss from a reduction in subsidies. The test for these income groups is the same as for the beneficiaries of cheap-food policies. The groups that are price-sensitive are the beneficiaries of subsidies as well as the potential losers in the event of a reduction in subsidy payments.

As has been argued earlier, a simple indication of the extent of financial benefits or policy transfers can be obtained by analysing the income groups that buy the largest amounts of foods that are subject to cheap-food policies. However, there are other, more direct measures of the extent of such transfers.

One such measure can be obtained by comparing the domestic subsidized price with the equivalent international consumer price for the commodities at the domestic location. For example. Alderman and Von Braun [9] offer a computational model to determine the financial benefit for any individual consumer and the total benefit for all the consumers in a particular income group for a given commodity, taking into account its domestic consumer price at the subsidized or controlled level, the equivalent international consumer price for the same locality, and the quantity of the commodity consumed at the subsidized price. They note, however, that international prices are not a stable point of reference. In addition, in the present study, error would be introduced in imputing international prices for commodities that were not being imported (e.g.. poultry and fresh milk).

In this study, instead, the financial benefits to the income groups for each subsidized commodity were estimated by apportioning the total subsidy payments for that commodity among the income groups in proportion to their consumption of the commodity.

Thus, the financial benefit (TCgks) for consumer k belonging to income group g for each commodity s was estimated using the following model:

TCgks = (Vs gs) /Popg.


Vs is the value of the subsidy for commodity s,
gs is the proportion of commodity s consumed by income group g, and
Popg is the population of income group g.


Application of the analytical framework

Price elasticities were obtained from an econometric study with the use of data from the Household Budgetary Survey of Trinidad and Tobago carried out by the Central Statistical Office in 1981/82. The data on household expenditures gave details of quantities and prices of food items purchased by each household during a two-week period. The survey also provided the monthly income and other socioeconomic characteristics of each household.

A total of 490 households were selected randomly from the survey. Households from each of the country's nine counties were represented in the sample. The sample was then divided into the following four income groups, based on the income-tax groupings of the Board of Inland Revenue:

Given the general approach suggested by economic theory, the major variables that were considered for inclusion in the demand functions were income, own-price of a commodity, prices of related commodities, consumer tastes, and variables that modify the tastes and preferences of households in Trinidad and Tobago.

Since cheap-food policies are specific to food items, it would have been ideal to estimate demand curves for individual items. It was difficult in this case. however, to estimate functions for the diversity of items included in the survey data, and some level of aggregation into food groups was therefore necessary. For this analysis, demand estimations of the four subsidized food groups - cereals, oils and fats, meat. and milk - were used.

The specific functional form used for the demand functions was determined by empirical fit. Regressions were run for linear, log-linear, and semi-log models. The semi-log models gave the best fit. The general form of the demand function is:

Cih = f(Pih,Pjh,Mh,Hh,Ah,Kh,Sh,Eh), where

Cih is the amount of kilocalories of food group i consumed per fortnight by household h (the caloric measurement was used instead of quantity because the study was concerned with nutritional impact),
Pih is the own-price per unit calorie of food group i for household h,
Pjh is the price per unit calorie of food group j for household h,
Mh is the monthly income of household h,
Hh is the size of household h,
Ah is the mean age of household h,
Kh is the factor called food-kilometres of household h (a composite variable which attempts to capture the effects of home produce on food purchases; the observations were obtained by multiplying the distance from the main urban centres by the proportion of domestic food produced in the county of household h),
Sh is the level of education of the head of household h, and
Eh is a zero-one dummy variable representing the ethnic group of the head of household h (two such variables were used to capture differences between African, Indian, and other ethnic groups: Eih = 1 where the head of the household is Indian and 0 otherwise; Eah = 1 where the head of the household is African and 0 otherwise).



Nutritional benefits

The own-price elasticities for cereals and milk were significant for the middle- and low-income and poor groups. The last two had significant own-price elasticities for oils and fats. No income group had significant own-price elasticities for meat. In accordance with the nutritional test outlined, this means that the middleand low-income and poor households were the nutritional beneficiaries of the policies providing cheap rice, flour, and milk. Cheap oils and fats were beneficial nutritionally to the low-income and poor households, who were price-responsive in relation to this food group. As none of the income groups were responsive to meat prices, it appears that no households benefited nutritionally from price controls and subsidies on meats. Hence the subsidies on poultry were not effective as a nutritional measure.

For the poor group, as the price of oils and fats decreased, there was an increase in the consumption not only of oils and fats but also of cereals, since these food groups were complementary (i.e., had negative cross-price elasticities).

On the other hand, for the low-income households a drop in the price of oils and fats caused more of these items to be bought, but this increase may have been offset nutritionally by a decrease in the consumption of meat, since the cross-price elasticity in this case was significantly positive.

Similarly for the middle-income group, a drop in the price of milk caused more milk to be bought, but this increase in the consumption of milk may have been offset by a decrease in the consumption of cereals since the cross-price elasticity was significantly positive.

In general the price variables were the most important ones in explaining the food-consumption behaviour of the middle- and low-income and poor households. Only the household-size and ethnic variables showed any other consistent significance. which demonstrates the relative unimportance of the variables affecting consumer tastes in explaining their food-consumption behaviour.

Patterns of consumption by income groups

The poor households differed significantly from the high-income households in their overall pattern of demand for cereals. No other comparisons of two income groups showed significant differences in consumption behaviour towards this food group. The poor households were responsive not only to the price of cereals but also to household-size and ethnic-group variables. The elasticities of the high-income households were not significant with the exception of the cross price of milk.

The consumption patterns of the low-income and poor groups for milk were significantly different from those of high-income households. The high-income households were not responsive to any of the explanatory variables, while the poor were responsive to the price of milk and household size, and low-income households were responsive to the price of milk and both ethnic variables.

None of the income groups differed significantly from another in overall demand for oils and fats or for meat.

Financial benefits

To test for the financial benefits of food price policies, the levels of calories consumed by the different income groups were estimated as well as the subsidy received per capita within each income group for each subsidized food group.

The daily consumption of calories for each subsidized food group for each income group was obtained from the Household Budgetary Survey data. These figures are presented in table 2. It can be seen that the poor households consumed the lowest levels of calories in all food groups. Thus they would have received the least financial benefit from the cheap-food policies. The low-income group had the second lowest levels of consumption of all food groups except cereals. The middle-income group had the highest levels of consumption of cereals and oils and fats, while the high-income group had the highest consumption of meat and milk. These results suggest that the middle- and high-income groups were the greatest financial beneficiaries of food price policies in Trinidad and Tobago over the period 1973-1985.

TABLE 2. Mean calorie consumption by food group and income group (kilocalories per household per day)




Low income





4 571


Oils and fats















TABLE 3. Financial benefits per capita from food subsides, by income group and food group (TTS)




Low income







Oils and fats




















Index of total

(poor= 100)





TABLE 4. Aggregate food subsidy payments (TT$ millions), Trinidad and Tobago, 1973-1985



Oils and fats


Poultry meat




The results of the more direct test of the financial beneficiaries of cheap-food policies are illustrated in table 3, which gives the per capita subsidy received by income groups, calculated from the data in tables 4-6. The index in table 3 supports the implication of table 2 that the poor income group derived the least financial benefit from cheap-food policies. The middle-income group received the highest financial benefit overall (TT$490 per capita) and also the highest benefits for cereals and for oils and fats. For meat and milk, the high-income households received the greatest financial benefits, and overall they received the second highest total financial benefit (TT$439 per capita).



In selecting items for cheap-food policies, policy makers generally assume that the common list of basic food items is best suited to benefit the poor and low-income groups. The estimation of price responsiveness of households in different income groups in Trinidad and Tobago shows that this is not necessarily so. The price elasticities for cereals and milk were significant for the middle-income as well as the low-income and poor groups, showing that for the period 1973-1985 the nutritional benefits of the price policies on these items were enjoyed by all three groups. The price elasticities for oils and fats were significant only for the low-income and poor households, indicating that they were the nutritional beneficiaries of the cheap-food policy for cooking oil.

TABLE 5. Percentage consumption of food groups by income groups




Low income







Oils and fats














TABLE 6. Mean population of income groups, 1973-1985





High-income 7.3 77.879
Middle-income 43.5 464.072
Low-income 21.6 230.436
Poor 27.6 294.446
Total 100 1,066,833 b

a. These are the percentages of households in each income group in the sample used in the study.
b. Ref. 10.


Meat, however, although a basic food item, was not suited to the provision of nutritional benefits through subsidies since no income group showed significant price-responsiveness for this food group. The large sums spent by the government of Trinidad and Tobago on poultry subsidies therefore may have been of no nutritional benefit to households in the country. It may be that households had reached their peak in the consumption of meat and were unwilling to consume further amounts. The major financial beneficiaries of the cheap-food policies were the middle- and high-income households. Since the latter were not responsive to prices of any of the food groups, it appears that the policies were simply a source of income supplementation for them. Given the conclusions concerning the nutritional beneficiaries of cheap-food policies between 1973 and 1985, it is possible to determine which households have suffered from the reduction in subsidies since 1983. First, the reduction of the poultry subsidy and its subsequent withdrawal at the end of 1985 are not expected to have had any impact on the nutrient consumption of households. With respect to the reduction of subsidies on rice and flour, the middle- and low-income and poor households have been put at a disadvantage. Similarly, nutritionally the low-income and poor households would have responded negatively to the withdrawal of the subsidy on cooking oil. This study demonstrates the desirability of estimating income-specific price elasticities when designing nutritional policies. Together with the tests of significance, the Chow tests confirm that cereals and milk are the food groups that are potentially most effective for price policies in favour of the poor households.



1. Poleman T. Quantifying the nutrition situation in developing countries. Food Res Inst Stud 1981:18(1):45.

2. Mellor J, Johnson B. The world food equation: interrelations among development. employment and food consumption. J Econ Lit 1984;22(Jun).

3. Chafkin S. Nutrition policies, programmes and politics. Am J Agric Econ 1978;60(5):806-09

4. Scobie G. Government policy and food imports: the case of wheat in Egypt. Research report 29. Washington, DC: International Food Policy Research Institute. 1981:30.

5. Government of Trinidad and Tobago. Annual report of the colony of Trinidad and Tobago for the year 1953. 1953:43.

6. Harris EL. Food price sensitivities of income groups in Trinidad and Tobago. MSc thesis. St. Augustine. Trinidad: University of the West Indies. 1987:97-99.

7. Timmer CP. Falcon WP. Pearson SR. Food policy analysis. World Bank publication. Baltimore, Md, USA: Johns Hopkins University Press. 1983:65-72.

8. Gujarati D. Basic econometrics. New York: McGraw-Hill, 1978:306.

9. Alderman H, Von Braun J. The effects of the Egyptian food ration and subsidy system on income distribution and consumption. Research report 45. Washington. DC: International Food Policy Research Institute. 1984.

10. Trinidad and Tobago. Central Statistical Office. Annual statistical digest, no. 32. 1985.

Contents - Previous - Next