Program 6

Sustainability of Potato Consumption in Developing Countries: The Case of Bangladesh
G.J. Scott¹ and H.E. Bouis²

Materials and Methods
Results and Discussion
Conclusions
Selected Reading

Since the early 1960s, a number of horticultural crops have experienced very rapid increases in production in developing countries. Typically, they are high-value, short-duration, labor-intensive crops grown primarily for sale rather than for on-farm consumption. As a result, several countries, particularly in Asia, look increasingly to horticultural crops as a source of increased output, consumption, and income.

Potato (Solanum tuberosum) is particularly important, partly because production growth has been so noteworthy. The percentage increase in production for potatoes in developing countries from 1961 to 1996 was greater than that for any other major food crop except wheat. For example, potato production in India increased by 525% during the period to nearly 17 million t. In Pakistan, potato production increased from 28,400 t in 1947-48 to over 1 million t by 1994-96. Turkey, Iran, North Korea, and Bangladesh have also had substantial increases in potato output. The domestic market has absorbed virtually all of this increased production.

These trends have raised a series of questions about the potential for expanded potato consumption and production.

• Has the observed increase in consumption been due to changes in relative prices or changes in income?
• To what extent would even lower prices, through increased production, stimulate even higher demand?

We attempt to answer these questions by examining historical demand elasticity estimates for potatoes in the case of Bangladesh and then comparing them with observed trends in use.

Materials and Methods

Results presented in this paper draw upon three different methods and sets of data. Because of space limitations, only the results themselves and their interpretation are included here. The first method involves estimating the demand parameters for a number of key food groups based on household expenditure data and a simplified analytical technique designed especially to estimate a demand matrix for a highly disaggregated group of foods.

The second method consists of a more traditional procedure used to estimate the same parameters, but with household expenditure data collected some 15 yr later.

The third method includes the results of a rapid market appraisal and descriptive analysis of the latest secondary data on production and consumption gathered in field work and statistical monitoring since 1982. The cross-checking of results is intended to test the predictability of the estimated parameters as well as the reliability of such estimates for predicting longer-term trends in consumer behavior.

Results and Discussion

Baseline historical data for Bangladesh are taken from the Household Expenditure Survey for 1973-74, conducted by the Bureau of Statistics. Selected food consumption patterns using these data are shown for urban and rural, high- and low-income consumers (Table 1). Relative price data per 1,000 calories are expressed as a ratio of the price paid for the least expensive grain (in this case, for wheat) by the low-income, urban quartile. For example, in 1973-74, the lowest income urban consumer in Bangladesh paid 1.63 times the price of wheat for 1 kg of potatoes (Table 1). However, since potatoes contain roughly one-fifth the calories of wheat on a price per calorie basis, potatoes were 7.39 times more expensive than wheat, due partly to their higher price, but more importantly to the lower quantity of calories (Table 1).

Rice and wheat are the two main staple foods consumed in Bangladesh. In calorie equivalents, rice was more than twice as expensive as wheat in urban areas at the time of the 1973-74 survey. Based on these historical data, apparently both urban and rural populations would buy higher quality wheat and rice as incomes rose, although the tendency is more marked for rice. Wheat was more expensive in rural areas than in urban areas; the price of rice was about the same between urban and rural areas.

Per capita annual potato consumption in Bangladesh in 1973-74 was higher in urban areas than in rural areas (Table 1). More than 85% of the total population resided in rural areas at the time. Still, in urban areas, potatoes were more than seven times as expensive a source of calories as wheat for low-income groups and nearly nine times as expensive for high-income groups, which purchased higher quality potatoes. Also in urban areas, potatoes were more than three times as expensive a source of calories as rice. The price differential between wheat and potatoes and between rice and potatoes was not nearly so great in rural areas.

Historical information about food imports and real wages helps to put these figures in perspective. In 1973-74, Bangladesh had just recently achieved nationhood after a war of independence. The devastation of the war was aggravated by the famine of 1974. Hence, in 1973-74 Bangladesh was heavily dependent on food aid and commercial cereal imports.

Total wheat supply was in the neighborhood of 1.7 million t. According to FAO, nearly 95% was imported. Domestic wheat production at the time was slightly over 100,000 t. Domestic potato supply consisted of some 725,000 t, virtually all of which was produced locally. Local wheat prices were depressed by aid, imports, and food subsidies, particularly in urban areas. Hence, the large differences in relative prices per calorie for potatoes vs. wheat that prevailed at the time the household survey was carried out were at least partly due to these unusual circumstances.

Real wages in Bangladesh were also at a relatively low point in 1973-74. Reasons cited for this include the political unrest leading up to independence, destruction caused by the war for independence, and the famine of 1974.

The 1973-74 survey data reveal large percentage increases in per capita consumption of potatoes across income quartiles in both urban and rural areas (Table 1). Consumption was low, however, even for high-income groups when compared with industrialized countries.

Simulated demand changes
Using historical demand characteristics for potatoes in Bangladesh, we ran four sets of simulations to examine the effect changes in the prices of particular foods and income might have on potato demand. We applied the consumption level in 1973-74 to the food demand elasticities given in Table 2 to yield the simulations presented in Table 3.

The first set of simulations assumed a 25% increase in income for each income quartile. Not unexpectedly, consumption increases are large given the low initial levels of potato consumption and the high income elasticities.

The second set of simulations assumed a 25% increase in the prices of all foods. That is somewhat equivalent to a decrease in income (except that nonfood expenditures become relatively more attractive). Under this assumption, consumption decreases precipitously.

A third set of simulations assumed a 25% decrease in the price of rice, which effectively increases income. Even though rice becomes an even cheaper source of calories than potatoes, potato consumption increases because of the increased income.

The fourth simulation assumed a 25% decline in the price of potatoes. The increase in per capita consumption of potatoes based on that assumption is on the order of 25%.

   Actual demand changes
Given the rise in per capita consumption of potato and its growing importance in the lean season in Bangladesh since 1973-74, potato has evolved from a minor vegetable to the most important vegetable in the diet and an occasional partial substitute for rice. That is consistent with the historical demand parameter estimates reported in the previous section, and with an observed increase in the price of rice relative to potatoes. The price ratio rose 45% from 1973 to 1983 in Dhaka. The decrease in the relative cost of potatoes greatly contributed to the increase in rural consumption from around 5 kg per capita in 1973-74 to around 16 kg per capita in 1981-82.

More recent consumption patterns
Patterns in 1988-89. Several factors led to changes in potato consumption patterns in Bangladesh during the remainder of the 1980s. First, potato production increased markedly due to increases in yield and area planted. Yield increased an average 2.7% annually from 1961 to 1993, whereas area planted increased at an annual rate of 1.8%. By 1993, annual potato production was 1.3 million t.

Second, wheat production increases were also impressive during the period, although they have tended to level off in recent years. Wheat production was 1.08 million t in 1991-93, only slightly higher than the 1.03 million t produced in 1981-82. Area planted to wheat grew by about 7% in the 1980s, from 563,000 to 604,000 ha. Not only did local production of wheat increase, but imports in 1988-89 were higher in absolute terms than in 1973-74. However, population growth in Bangladesh virtually eliminated any increases in per capita availability of wheat.

Third, total rice supplies increased by more than 40%, from 12.9 to 18.8 million tons. The spread of high-yielding varieties was a principal factor. Rice import increases were minor compared with increases in domestic production.

Fourth, real income increased sharply during the 1980s (Figure 1). Among the possible reasons are the spread of improved rice varieties capable of high yields in the dry season.

 

Fifth, results from the Bangladesh household expenditure survey for 1988-89 show that per capita potato consumption rose sharply (Table 4). That appears consistent with the shift in relative prices for potatoes vs. wheat¾particularly in urban areas where relative prices fell by 50%.

Table 4.

Consumption of rice, wheat, and potatoes, and computed relative prices per 1,000 calories in Bangladesh, 1973-74 vs. 1988-89.

The relative price for potatoes vs. rice, however, remained virtually unchanged in urban areas and actually rose in the countryside (Table 4). As rice consumption increased, particularly among low-income consumers, this added sense of wealth engendered increased potato consumption as well. Increased potato consumption is consistent with the rise in real wages.

Potatoes are more than just an alternative source of calories, or a tasty alternative to a strictly cereal-based diet. In Bangladesh, potatoes are an important source of vitamin C and also provide essential amino acids that are low in rice.

   Patterns in 1995-96. Potato production in Bangladesh rose by nearly 250,000 t between 1988-89 and 1995-96 to 1.45 million t. Detailed consumption figures by place of residence and income quartile are harder to come by. But FAO Food Balance Sheet data indicate that average per capita consumption in 1992-94 was 9.8 kg/yr, more than double the 4.8 kg/yr reported for 1961-63. The International Food Policy Research Institute and CIP estimate continued increases in potato production and consumption well into the next century.

Conclusions

Potatoes are of increasing interest to policymakers and planners in developing countries because of their proven potential to raise farm incomes, rural employment, and food consumption. This last consideration is particularly important in South Asia where large segments of the rural population are low-income consumers.

To better understand past increases in potato consumption and to more accurately estimate the potential for further growth, basic information is required on the demand characteristics of the commodity. In North America and Europe, potatoes are regarded as a starchy staple. But in Asia they are valued for their dietary variety, taste, and the essential vitamins and amino acids they provide. Potatoes are still an expensive source of calories relative to wheat and rice. But as we have shown in Bangladesh, the relative price vs. wheat has fallen substantially. Estimates for Bangladesh indicate that the prospects for increased potato consumption are favorable, yet modest in per capita terms, if incomes increase and potato prices continue to decline relative to staples. Government policy can greatly influence the future demand for potatoes. Various measures might be adopted to help further lower production costs and the retail price. Among them are lowering the cost and improving the availability of planting material and storage facilities, and cutting subsidies on imported wheat. Given the estimated demand parameters, increased potato supplies could be readily consumed in the local market.

Selected Reading

Bouis, H.E. and G.J. Scott. 1996. Demand for high-value secondary crops in developing countries: The case of potatoes in Bangladesh and Pakistan. Food Consumption and Nutrition Division Discussion Paper No. 14. International Food Policy Research Institute (IFPRI), Washington, D.C. 42 p.

Goletti, F. 1993. Analysis of demand for potatoes in Bangladesh. IFPRI, Washington, D.C. 35 p. (In mimeograph.)

Scott, G. 1988. Marketing Bangladesh´s potatoes: Present patterns and future prospects. International Potato Center (CIP)/Australian Development Assistance Bureau (ADAB), Dhaka, Bangladesh. 107 p.

---

1 CIP, Lima, Peru.
2 International Food Policy Research Institute, Washington, D.C., USA.