Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu

Production of special foods

It is not possible to determine how much overall impact the project had on production of foods. Furthermore, it is clear that the NBPP influenced production in geographically adjacent areas, not just the project area. There were no baseline data, and as discussed below, of the various project activities undertaken by NBPP, production of special foods is perhaps the one most likely to have made substantial progress during the first two years of the project, 1990 and 1991.

Table I presents data on the percentage of households growing various foods, in relation to NBPP activities in promoting them, divided into five categories. Interpretation of the results is facilitated by keeping in mind that the table records the results of the third year of a three-year process of change, gradually increasing acceptance of the promoted foods during the course of the project. As could be expected, the universal distribution of seed (category 1) appeared to have the greatest effect on the proportion of households growing them. This probably was true also for the adjacent non-project subdistricts. Universal seed distribution appeared to have reached a saturation level by 1992. Thus the incremental increase during the third year was relatively smaller in the project than in the non-project area.

The foods in category 2 were distributed free to target groups and promoted to everyone. Production of these appeared to increase during 1992 in about equal increments in the two areas, with the non-project area being slightly behind in the process. Exotic or new varieties such as carrots and cabbage were exceptions. They appeared to be just catching on after two to three years of promotion, leading to greater incremental increases in the project than in the non-project area.

A rapid qualitative survey of vegetable and vegetable seed salesmen in five cities where NBPP activities could have been expected to have had a cumulative effect over several years found that NBPP-promoted foods were those that had increased most in sales in recent years. Carrots, which had been relatively unknown in the area a few years before, had become one of the most popular vegetables [4].

Production of foods in category 3, which NBPP only promoted, without providing any free seed, could be expected to increase more slowly. Examination of the data suggests that this was indeed the case. Only in this category were the incremental increases in production generally much greater in the project area than in the nonproject area. Again, these differences were especially pronounced for new or exotic varieties such as Chinese cabbage and sweet potato. There was also an increase in the number of people who grew traditional or wild foods for which seed need not be purchased, such as drumstick tree leaves and colocacia. Here the mass media may have enhanced the status of these easily available but underappreciated or "poor man's" varieties.

Foods for which NBPP provided free seed appeared to increase more in the non-project areas than those for which NBPP provided only promotional messages. This suggests that the free seed was widely sold to neighbouring areas, especially the seed that everyone received, which would rapidly have lost its market value in Gaibandah district. Information probe ably diffuses more slowly than something that can be sold. Among foods not promoted by NBPP (category 4), there was great variability in both the areas, with mainly increases in production, often of large magnitude, but also some decreases.

A statistically significant increase was seen in the proportion of households growing green vegetables and yellow fruits in the two areas, but for oil-rich foods only the increase in the project area was statistically significant. This is consistent with the fact that promotion of consumption of these foods started late, and seeds for them were not distributed. There was also a smaller but statistically significant increase in the proportion of households growing non-special foods in both areas (category 5).

Respondents were asked where the family was growing special foods and the size of the areas cultivated. As expected, respondents in the project area had fewer and smaller separate gardens, since more of them were landless. On the other hand, more of them were growing more special foods in strips beside their homes and on rooftops, presumably in response to NBPP encouragement and assistance. An increase in rooftop gardening was probably a year behind in the non-project area, as cheap seeds for vines were presumably coming there from Gaibandah.

Separate gardens are larger and are more likely than the other types to be intended to generate income. They actually seemed to be declining in prevalence and size more in the project area than in the non-project area. Three factors may have contributed to this rather anomalous finding.

First, it might have been partly due to disappointment among the larger vegetable farmers caused by the low prices their produce had obtained the year before. Apparently, NBPP efforts to increase demand had not kept pace with increases in production. Unlike the seed itself, fresh produce is expensive and difficult to transport; the market was not used to handling large quantities of surplus produce; and it cannot be stored. The Bogra portion of the non-project area has a long-established tradition of vegetable production and marketing and may have been better able to find markets for its increased production, especially since the pace of increase was slower.

Second, an unusual storm, including heavy rain and hail, occurred in early 1993 and destroyed over 70% of the NBPP gardens in Gaibandah. This may have had a greater effect on the more exposed separate gardens than on those against walls or on rooftops. Some of the affected gardens may not have been restarted by the time of the March 1993 survey.

Finally, a substitution effect may have been caused by distributing free seed for vegetables suitable for growing on rooftops. Gardens could then be used for other foods instead.

The proportion of sample households not growing any special foods declined in the project area from 6.8% to 2.7% and in the non-project area from 9.6% to 4.3%. In both areas, about 97% of these households gave as the reason that they had little or no land to cultivate. Other reasons mentioned by 5% to 10% were no money poor Yield.

Consumption of special foods

Respondents were asked to recall the types of foods the household and children age one to six years had consumed during the past 24 hours (tables 2 and 3). No information was requested regarding amounts consumed. Table 4 compares consumption of special foods by households and young children in the two areas. In table 5 the differences in consumption of special foods by children are compared between 1992 and 1993 and tested for statistical significance. The following four factors must be taken into account in interpreting these data:

>> The population in the non-project area had a higher socio-economic status than that in the project area. Therefore they were able to consume more expensive foods and less of the cheaper, less desirable foods such as wheat and green leafy vegetables.
>> The evaluation was conducted in the third year of the NBPP's field activity. Some of NBPP's communication efforts in the mass media probably had an effect on the non-project area as well as on the project area.

TABLE 2. Foods consumed by households (24-hour recall)

a. Foods probably consumed more than usual during Ramadan.

TABLE 3. Foods consumed by children age one to six years (24-hour recall)

a. Foods probably consumed more than usual during Ramadan.

TABLE 4. Consumption of special foods by households and young children

These figures represent the percentage of households and children, respectively, who ate at least one item within each respective category (as shown in tables 2 and 3).

>> Bangladesh had an especially large rice harvest a few months before the 1993 survey. In the early months of 1993, rice was selling for 5 take per kilogram. In the months before the 1992 survey, it had been selling for 9 take per kilogram. This meant that in early 1993 rice was actually cheaper than wheat (usually eaten in rural Bangladesh only by those who cannot afford rice), an unprecedented occurrence.

TABLE 5. Percentage of differences consumption of special foods among children from 1992 to 1993

a. Differences are statistically significant at the 5% level, using the two-tailed t test. Figures in parentheses are the estimated standard error of the difference NA = not applicable. Since the difference is nut statistically significant' the confidence limit is not given.

>> Poorer people tend to spend a larger proportion of their income on food. When the major source of calories declines this much in price, much of the money that is freed is spent on more desirable and more expensive foods. Through a substitution effect, consumption of cheaper, less desirable foods will concurrently decline.

About one-fourth of the households in Gaibandah are totally landless and another one-fourth are effectively so, since they own far too little land to be self sufficient in food production. For these families, lower rice prices have an income-increasing effect. For the few households that produce substantial surplus rice, the lower prices may have had an income-reducing effect. This would tend to negate or dilute the impact of lower rice prices on consumption patterns in the sample as a whole. However, the wealthier households are few in number, they were already spending a lower proportion of their income on food, and reduced income after one harvest probe ably reduced the variety of foods they consumed less than the quantities of expensive foods. The larger the farm, the greater the negative impact of lower prices, but the greater the cushion of wealth that could absorb the temporary shock and thus reduce any temporary impact on consumption patterns.

Due to the requirements of the donor agency, the timing of this evaluation was inflexible. This meant that it had to be conducted during the Islamic fasting month of Ramadan. Nearly everyone avoids consuming foods and fluids during the daylight hours during this month (exceptions are small children, sick people, and some pregnant women). This probably does not result in overall decreases in consumption, however. After sunset and before sunrise, people eat two or three meals, including certain distinctive foods, often resulting in higher than normal food expenditures for this month.

In general, the data in table 3 do behave in expected ways once the factors as explained are taken into account. In 1992 most of the expensive, desirable foods, including meat, fish, snacks, drinks (lemon and other concentrates), milk, and sugar, were consumed by a greater proportion of households in the non-project area than in the project area. The opposite, as expected, was true for wheat and vegetables, which are low-status, cheaper foods.

Although it was clear that NBPP seed had spread to the non-project area, it is more difficult to estimate what impact this and the mass media messages may have had on consumption patterns before and during the evaluation year. Since Bogra has a tradition of commercial production of vegetables, increased production there may not have resulted in an equivalent increased consumption. Increases in production caused by cheap seed from Gaibandah may also have had the effect mainly of reducing the production of similar vegetables. The NBPP messages that reach outside its districts were restricted to radio and television spots. These are likely to have only a small effect when not backed up by the other mass media and interpersonal channels used by NBPP in Gaibandah. Any such effect is also likely to be limited to the wealthier classes.

The remarkable decline in rice prices resulted in only small increases in the proportion of households consuming rice, since nearly all did so in 1992. The major effect was a drop in consumption of wheat. A higher proportion of the non-project sample continued to consume wheat. These may he more modernized or Westernized families who have become accustomed to eating bread and have an income that makes a shift to exclusive consumption of cheaper rice unnecessary. There was a slight drop in consumption of sweet potato as well, since it also is relatively inexpensive and can serve as a replacement for rice.

The money saved by having to spend less on the staple food allowed people to purchase more expensive foods. Greater proportions of people in both areas consumed meat, milk, and snacks, the use of which NBPP did not promote. Other desirable foods that increased in consumption and were also promoted by NBPP included yellow fruits, mustard seed, and oil on foods other than curry. (Oil is also added to rice to make it more tasty when people can afford it.) Some vegetables are fairly expensive and desirable, and this may explain why consumption of "other vegetables" went up slightly in both areas.

Green leafy vegetables, however, are widely agreed to have a reputation as "poor man's food" in Bangladesh [5]. In the non-project area, their consumption went down by 7.9%, a statistically significant decline. In contrast to this expected trend, consumption of these items increased in the project area by 12.5%, also statistically significant. Thus the net change due to the project was its 12.5% increase minus the 7.9% decrease in the non-project area: 12.5% - ( - 7.9%) = 20.4%. Whereas households in the two areas consumed green leafy vegetables about equally often in 1992, nearly twice as many did so in Gaibandah in 1993.

Young children are universally thought to resist eating green leafy vegetables, and studies in Manikganj [5] and Comilla [6] supported this belief. Parents in Bangladesh are also said to worry that green leaves cause diarrhoea among very young children. The NBPP encourages production and consumption of some softer and less bitter varieties that children liked.

If consumption levels were this low in the two areas at baseline, substantial increases may already have occurred by 1992, when 23% of one-year-olds in the non-project area and 33% in the project area had consumed leafy vegetables the day before the interview. The figures for the older children were roughly 35% and 40%, respectively. Consumption did generally increase by age in both areas except in the non-project area in 1993, where it was fairly evenly low in all age groups.

In the non-project area, leafy vegetable consumption among children declined from 34% to 26% (SE of the difference, 3.0, t = 2.6). In the project area, it increased from 40% to 52% (SE of the difference, 3.3; t = 3.8). Both these changes were statistically significant at the 5% level. Similar to the case for the households, consumption of leafy vegetables by 1993 was twice as high among children in the project as in the non-project area.

Consumption of these foods by young children in Gaibandah followed the pattern of consumption in the households in general. Whereas this method of analysis reveals which foods are eaten less by children than by the household as a whole, it cannot reveal which foods may be eaten exclusively or mainly by children. Thus we can only conclude that no special foods are avoided by children in this area. The biggest gap is among "other vegetables." Over 10% of children in all four cells avoided eating them the day before the interview, even though others in the household did so.

continue

Contents - Previous - Next