The intra-household economic and nutritional impacts of market-oriented dairy production: Evidence from the Ethiopian highlands

B.I. Shapiro¹, Jemal Haider², Alemu G/Wold³ and Abebe Misgina¹

1. International Livestock Research Institute, Addis Ababa, Ethiopia

2. Ethiopian Health and Nutrition Research Institute, Addis Ababa, Ethiopia

3. Ethiopian Agricultural Research Organization, Addis Ababa, Ethiopia

Abstract

Previous farm-level studies have shown that adopting crossbred cows and improved feeding and management strategies increases milk production and household income of resource-poor smallholder mixed crop–livestock farmers. This study shows how those changes affect intra-household decision making, including the control of new income by women, and the resulting impacts on the nutrition and health status of household members, especially those most vulnerable, the mothers and children. Members of households with crossbred cows on average consumed 17% more calories, 24% more fat and 13% more protein than members of households with no improved cows. The households with crossbred cows spent about 7% more on food purchases, and in addition, they consumed about 30% as much pulses. In spite of increased sales of milk, households with crossbred cows consumed more than twice as much milk as households with local cows, especially benefiting children.

Other studies in other areas have reported that as cash crops are introduced in smallholder production systems, with greater integration into the market, women may lose control over cash income to men, who tend to spend less on food for the household. Results of these studies indicate that women in households with crossbred cows maintain control over income allocated for food purchases and make over 80% of household expenditures on food. Furthermore, the men in these households spend some 28% more on food than men in households with no crossbred cows. It also proved that this food-based intervention can have a significant positive impact on human nutrition and health status. These results should be interpreted with some caution because they are based on a cross-sectional comparison of owners and non-owners of crossbred cows without controlling for the potential effects of other factors. Ideally, a comparison of situations before and after, and with and without crossbred cow ownership should be made to assess the net effect of crossbred cows.

Background and justification

Improving human nutrition, including micronutrition, are critical elements of achieving food security and are compelling and priority concerns of development efforts in sub-Saharan Africa (SSA). Interest is mounting in the donor and nutrition communities in food-based approaches (dietary change) to combating nutrient- and micronutrient-deficiencies. The essential role of livestock, and more particularly the potential for market-oriented dairy production (MODP) to combat these problems, may have been overlooked or understated.

Food-based approaches (dietary change) to combating nutrient- and micronutrient- deficiencies are likely to be more sustainable than chemical supplementation, such as vitamin A tablets (Neumann et al 1993). Not only the nutrition community, but also agriculturalists, have an interest in showing that food-based approaches can improve the well-being of the members of poor rural households. This includes establishing that positive nutrition and health impacts can occur when the incomes of household members are raised through market integration, making food more accessible through markets.

Livestock products, and especially dairy products, can make unique contributions to nutrition, and especially the micronurient and health status of smallholder household members. Contrary to the developed world where many people eat too much animal products, most people throughout SSA eat too little animal products for good nutrition. Under these circumstances, increased milk consumption has a significant positive effect on the nutritional status of household members, especially children. Besides containing important nutrients, especially carbohydrates, protein, and calcium, dairy products contain micronutrients in bioavailable form such as vitamin A (retinol), the B-complexes and zinc, essential for growth and proper physical and mental development. Evidence is mounting that the bioavailability of Vitamin A in carotene-rich leafy-green vegetables is much lower than previously believed (Gillespie and Mason 1994; de Pee et al 1995). Studies done by the Human Nutrition CRSP in Kenya have shown that children deprived of vitamin B12, which comes only from animal products, suffer impaired learning abilities and do not develop their full human potential (Neumann et al 1993).

The introduction of MODP technologies results in substantial changes in terms of allocation of labour and other resources. The consequences on individual family members, however, are not clear. A study conducted in the early 1980s indicated that the introduction of crossbred dairy cattle to the Debre Berhan region of the Ethiopian highlands provided a number of advantages to the household as a whole, e.g. higher milk output and incomes (Wagenaar-Brouwer 1986). At the same time, however, the new technologies increased the labour demands placed on women (Whalen 1983; Whalen 1984). Later studies in the Selale region showed that while women's labour time was increased, it was the overall labour time of men that increased the most (MoA/FINNIDA 1991).

There has also been conflicting evidence regarding the consequences of crossbred cows on intra-household decision making and control of the benefits of dairying, thus affecting individual household member's incomes and expenditures, and consumption. In the Debre Berhan region, women were reported to process most of the milk from local cattle breeds into butter before the introduction of crossbred cows and to spend most of this income on food for household consumption. Introduction of these higher productivity animals resulted in more direct sales of fresh milk by men. Men were reported to control this income from fresh milk sales, but it is not known what became of this income (Whalen 1983; Whalen 1984). Two years after the end of a development project in Selale in Ethiopia, the International Livestock Research Institute (ILRI) began evaluating the effects of these technologies on farm households. The results there showed that although the income from fresh milk sales was more likely to be controlled by men, it was used to buy food items and agricultural inputs, and per capita food consumption increased. The men did not spend the additional income on other non-food items such as drinking.

In Debre Berhan, average daily energy intake was slightly higher for the non-adopting households (Wagenaar-Brouwer 1986). There were large seasonal differences in the intake of livestock products. However, 8% of calories of the adopting households on average were provided by livestock products and only 4% for non-adopting households. Per capita milk consumption (whole milk plus buttermilk) was twice as high among adopting households as compared with non-adopting households. No analysis was undertaken of differences in micronutrient intakes between adopting and non-adopting households or between seasons. Furthermore, the nutritional analysis was not done at the individual household member level.

Thus, the impacts of MODP on nutrition and health, especially as they affect the more vulnerable household members, women and children, are not clear. Most studies have evaluated the impacts of new technologies only at the household, rather than at the individual household member level, and have not included explicit indicators of the consequences for human nutrition and health. The intra-household consequences of crossbred cow technologies need to be assessed to ensure that all household members are reaping the potential benefits.

Objectives

This study investigates the intra-household resource allocation, income, expenditures, food consumption, nutrition and health consequences of intensive dairying. The objective is to document the impacts on household members and identify policy options that could help ensure that the benefits of the new technologies are shared equitably by all household members.

To fully understand the impacts of the technologies, intra-household decision-making must be understood, and the complex intra-household relationships and consequences have to be investigated. A number of sub-issues thus need to be addressed that together provide a picture of the impacts on all household members.

Conceptual framework

The conceptual framework depicted in Figure 1 visualises the linkages to be investigated in this study. It provides an understanding of the effects on intra-household decision-making and outcomes brought about by the introduction of MODP with crossbred cows. These effects are those occurring in the short run.

Figure 1. The linkages between the introduction of MODP and intra-household impacts.

There are other factors, e.g. the prices for food and non-food items, the education level of parents, especially mothers, mother's nutritional knowledge and the overall health environment related to the introduction of MODP, influencing resource allocation, incomes, expenditures, consumption and nutrition. These factors could also be expressed as variables to provide a full picture of the impacts of MODP. Most of these factors not addressed in the figure are not specific, however, to whether MODP technologies are adopted or not.

Ideally a comparison of situations before and after, and with and without crossbred cow ownership should be made to assess the net effect of crossbred cows. Moreover, rigorous econometric analysis should be done to control the effects of factors other than cow type. However, the results presented in this paper are based on a partial analysis of a cross-sectional comparison of household with and without crossbred cows. Therefore, the implication of any significant difference between the two groups should be interpreted with some degree of caution.

The farming systems in the area

The farming system in the study area is classified as a mixed crop–livestock system. Farmers owning crossbred cows (CBC) have larger landholdings of 2.8 ha compared to 2.1 ha for farmers owning local breeds of cows (LBC) and keep less livestock per ha. The average land area cultivated by all households comprises about 1.5–2.0 ha and livestock holdings are equivalent to 8.1 TLUs¹. Main crops grown are teff, barley, wheat, and various types of pulses.

 

1.    Tropical livestock unit, taking an animal of 250 kg live weight as the basis for conversion factors for different sizes and species (Gryseels 1988; Jahnke 1988).

Farmers with CBC keep fewer, but more productive animals (the crossbred cows). For MODP households TLU/ha are 3.2, for non-adopters it is 3.5. This difference is not significant at the 0.1% level indicating the wide variation in livestock holdings. The average animal herd consists of 2 cows, 2 oxen, a horse or a donkey, and a number of sheep and chicken. While the control farmers (LBC) exclusively own indigenous cows, the group of technology adopters (CBC) own indigenous and crossbred cows.

Households at Holetta comprise a family unit of husband, wife and children. Often they live together with other relatives and accommodate a farm employee who usually is related to the family, as well. The average household size of the sample farmers is about 6.2 people. Households have an average of 4 children. This equals an agricultural labour force of 3.5 adult male equivalents with a very small and insignificant difference between the two groups under study.

To obtain more meaningful labour, food consumption, expenditure, and nutrition assessment, household size can be converted into respective adult male equivalents. For agricultural labour, man equivalents were used (Norman et al 1981). There are 3.1 men equivalents in CBC households and 3.2 in LBC households. For food consumption, expenditure and nutrition assessment, adult consumption equivalents were used. There are 4.7 adult equivalents in CBC households and 4.9 in LBC households. Adult consumption equivalents were calculated according to Food and Agriculture Organization of the United Nations/World Health Organization (FAO/WHO) conversion factors (FAO/WHO 1985).

Data

The data for this study was provided by a collaborative dairy technology project involving the Ethiopian Agricultural Research Organization (EARO), and the ILRI and the Ethiopian Health and Nutrition Research Institute (EHNRI). One major objective of the project is to develop technologies to enable resource-poor smallholder mixed crop–livestock farmers to participate in market-oriented dairying. Another major objective is testing the use of crossbred dairy cows for traction, as well as milk production.

The study site is located in the Ethiopian highlands, about 40 km west of Addis Ababa in the vicinity of a small town called Holetta. The altitude of the research area is about 2600 m asl. Average annual rainfall is 1100 mm with mean daily temperature of less than 20°C. The main rainy period starts in mid June and continues up to September and is called the Meher season. The short rains or Belg are from February to April and are mainly used to break the soil. Farmers in Holleta mainly depend on the Meher rains to plant crops.

Pairs of crossbred dairy cows were introduced on 14 farms in Holetta in 1993, half for milk production only, and half for traction as well as milk production. On-farm mean milk yield and average lactation days were not affected very much by traction. Thus, the use of the crossbred cows for traction by households, in addition to dairy production, is not considered of relevance for the purposes of this consultation. In 1995 and early 1996, 120 more crossbred cows were introduced into an additional 60 households that were all using the cows for traction, in addition to milk production and reproduction. Sixty control households using traditional practices of local Zebu cows for milk production and oxen for traction are also being followed since mid-1996. There were then 134 households who were surveyed for one year from October 1996 through September of 1997. The 134 households included two groups of 74 with crossbred cows and 60 without.

Intensive farm-level data collection was carried out on a daily basis for this aspect of the study on 84 farm households during 1997. Special attention was given to the nutrition and health status of children and women since they are at most risk. The target population of the study thus included all members of households with pre-school children 6 months to 6 years old. These households include both those that were using the new technologies and those following traditional practices.

Data collection included land use, labour allocation and activity times, draft power use and source, input use, output use and price data to enable a whole-farm analysis to evaluate the potential of dairy-draft, intra-household data on resource allocation, income distribution, and expenditure by household members. Food availability and food intake by household member were also collected.

Anthropometric measurements (weight for height, weight for age and height for age) were taken on women and pre-school children in all the households. Anthropometric measurements were focused on pre-school children six months to six years old since they have relatively high risks of developing nutrient-deficiency related health problems. Furthermore, due to their key position in ensuring food availability for all household members, child health care measures and the nutritional knowledge of mothers was assessed. Environmental health factors, child care and health care measures, child morbidity, and mothers' nutritional knowledge were also assessed.

Results and discussion

Production and labour allocation

Milk yields of crossbred cows in intensified smallholder production systems can exceed those of local cows by up to 6, even 8 times. In a conducive peri-urban marketing environment this leads to doubling, even tripling of farm cash income (Gryseels and Whalen 1984; Mengistu Buta 1997). In the study region, milk production per crossbred cow is about 5 times higher than for local cows.

Members of adopting and non-adopting households spend slightly more time in animal husbandry, but CBC households hire more labour, both men and children, than do LBC households (Table 1). There was not much seasonal variation in time spent in animal husbandry, although there is considerable seasonal variation in labour devoted to crop production. Children (as herders), and to a lesser extent men, were found to be primarily responsible for animal husbandry.

Table 1. Weekly average family and hired labour in person equivalent (hrs) for all dairy activities.

Farm group	Family labour (hrs)			Hired labour (hrs)
	Men	Women	Children	Men	Children
CBC	10.9	2.7	24.7	5.4	10.4
LBC	4.8	1.8	25.1	0.5	0.5

The introduction of MODP technologies does not increase the time women spend on animal husbandry and would not be expected to change women's time allocation patterns for other activities provided the higher milk output is sold fresh rather than processed. Thus, the minimal impact on time spent on household activities such as childcare and food preparation would not be expected to negatively impact on child nutrition and health.

Impact on income

Previous studies at various sites in the Ethiopian highlands, as well as in Kenya, Uganda and India, have shown that cash incomes of CBC households increase significantly as a consequence of higher milk production and increased commercialisation of dairy products (Gryseels and Whalen 1984; Kennedy 1987; Walshe et al 1991; Sanders et al 1996; Mengistu Buta 1997).

Cash income from dairying at the study site was found to be significantly higher in households that keep CBC (25 Ethiopian Birr per adult equivalent per month) than in households that keep LBC (3.6 Birr per adult equivalent per month). Moreover, both men and women were observed to benefit, with men making an average income of 15.2 Birr per month and women 9.6 Birr (Table 2). Cash incomes of farmers with intensified dairy production are about 75% higher than that of those farmers using traditional production methods. CBC owners earn almost 7 times more income from the sale of dairy products than sales by LBC owners. This difference is attributable mainly to liquid milk sales. CBC owners have only slightly more income (about 20%) from animal sales, primarily from sales of cattle and particularly crossbred cattle.

Table 2 also documents that CBC owners have more off-farm income. This is mainly due to wage labour and farm related businesses, primarily production of tej, a local beer, and renting out of draft animals. Perhaps due to the ability to hire labour, CBC households can allocate more time to off-farm income earning activities. The share of off-farm to total cash income is about 39% of total cash income for CBC households and about 17% for LBC households.

Table 2. Monthly cash income from sales of farm outputs (Birr/adult equivalent/month).

⁺, ^*, ^** Indicate significant at the 0.10, 0.05 and 0.01 levels, respectively.

An important issue is whether the introduction of MODP with crossbred cows makes women better or worse off and who controls the income from the new commercialised activity. It has been reported that under traditional dairy production practices in Ethiopia, almost the entire milk yield is usually processed into butter, an activity exclusively undertaken by women, who also sell the butter at the local market and retain the money for the purchase of household consumption items (Whalen 1983).

As indicated in Table 3, husbands in Holetta apparently gain more of the additional income through milk sales. Their cash income from fresh milk sales is about 13 times that of LBC owners. However, the anticipated negative effect on female income does not take place. The incomes of women in CBC households are almost 4 times that for women in LBC households. In addition to increases mainly from the marketing of fresh milk by husbands, wives gain additional cash income through more sales of butter and cheese, products traditionally produced by women.

Whereas women in LBC households realise more income in relative terms, in absolute terms it is the women in CBC households who gain most. Women in CBC households still realise about 63% of total dairy income. It can be observed thus that the distinction between male and female responsibilities in marketing of dairy products is not as rigid as often assumed. Wives of adopting households market large quantities of fresh milk under the changed production system. Income from milk sales by wives is about twice the amount of the increased cash income from cheese and butter sales combined. Thus, the hypothesis that women lose control over dairy income is rejected. The effect on food expenditure needs also to be considered and is in the next section.

Table 3. Dairy income from milk, butter and cheese sales of wife and husband by technology group (Birr/adult equivalent/month).

Product	Income earner	CBC households		LBC households
		Average	Coef. of variation	Average	Coef. of variation
Milk	Husband	15.1**	1.2	1.0**	6.4
	Wife	5.2**	2.4	0.0**	5.0
Butter	Husband	0.1	4.6	0.0	6.0
	Wife	2.8	1.6	1.6	2.8
Cheese	Husband	0.0	4.8	0.1	6.2
	Wife	1.6	1.8	0.8	2.5
Total	Husband	15.2**	1.2	1.1**	6.3
	Wife	9.6**	1.8	2.4**	2.7

** Indicate significant at the 0.01 level.

Expenditures documented in this section are cash expenditures on farm inputs, food, and other non-food commodities. Expenditure levels and patterns can be expected to change as a result of higher and more readily available cash income from the dairy operations. This has been documented for other examples of agricultural commercialisation (Kennedy 1987; von Braun 1995).

Impact on expenditure

The surveyed households differ in levels and patterns of cash expenditures between CBC and LBC owners, although not to the extent experienced at other sites under commercialisation of production. These differences mainly occur within the non-food categories of commodities. Households with CBCs have higher cash expenditures for farm inputs, transportation, tax and debt repayment, milling and clothing, which add up to higher but not significantly different total cash expenditures.

Households with CBCs do purchase significantly more animal feed and animal health and artificial insemination (AI) services, as reflected in Table 4. These households exclusively purchase seeds for pulses and vegetables, protein and vitamin-rich crops that can significantly affect consumption and nutrition.

Research elsewhere has shown that wife and husband have different responsibilities to provide for family needs utilising the respective income they control. Expenditure patterns usually depend upon the amount of money they have at their disposal and on social and cultural norms. It is widely believed that women spend more income on food (Quisumbing et al 1995). A comprehensive analysis therefore has to carefully investigate if changes in income control occur between husband and wife and if these changes adversely affect food consumption and nutrition.

As shown in Table 5, women in both CBC and LBC households spend more income on food items than men, over 80% more in both groups. However, with the introduction of the MODP technologies both men and women increased food expenditure by the same absolute amount (about 2 Birr/month) but men in CBC households increased their purchases of food by about 28% while women did so by only about 4%. Thus, women in CBC households do not lose control of cash income and expenditures on food do not decline with the introduction of the MODP technologies.

*,** Indicate significant at the 0.05 and 0.10 level.

Table 5. Monthly average expenditure per household on food and non-food items (Birr).                              

Husbands, meanwhile, expend more on farm inputs, taxes and debt repayments and durable household goods. Wives of CBC households increase their expenditures on non-food items, using some additional income on purchases of cloth and transportation. Tax, trade and debt repayments on the other hand, which are almost double for CBC households, remain the responsibility of husbands.

Impact on consumption

It has been documented in several studies that technological change and commercialisation of smallholder agricultural production improves the level of food consumption of participating households (von Braun and Kennedy 1994). Changes in food consumption are generally associated with more readily available cash income, but increased commercialisation can lead to increased self sufficiency through increased land and labour inputs into subsistence food production and changes in cropping patterns (von Braun et al 1989). With higher incomes, a substitution of cheap calories by more expensive calories often takes place and diets gain in quality and diversity. Consumption and nutrition affected by technological change and commercialisation have been attributed to income increases rather than higher milk availability (Alderman 1987; Binswanger and von Braun 1991). Increased household calorie consumption associated with cash crop production schemes has been identified as one of the important determinants for improved child nutrition, although the scope of improvements often falls short of expectations (Kennedy et al 1994).

The introduction of higher yielding crossbred cows is expected to improve consumption of dairy products, and possibly other foods because of increased income through milk sales. This is evident in differences in per capita average weekly milk consumption between the two groups of households (Table 6). All family members of CBC households profit from the higher milk yields. While priority in domestic milk consumption is clearly given to children, women and men gain as well. Although these households increase their milk consumption to such an extent, they still market a far higher amount of the produce. Households with LBCs consume about one-fourth the amount consumed by households with CBCs but the former group consume nearly 40% of their milk output compared to 13.5% by the CBC group.

Table 6. Average weekly milk consumption by technology group disaggregated for family members.                                                                   

 The numbers in parenthesis are the coefficients of variation.

 ⁺, *, **, *** , Indicate significant at the 0.10, 0.05 and 0.01 level.

Impact on nutrition

Does introduction of MODP technologies improve macro and micronutrient status of specific household members, measured by anthropometric measures and clinical signs? Children at the study site suffer exceptionally from eye problems closely related to vitamin A deficiency. In pre-study eye examinations a high incidence of Bitot's spots and conjunctival xerosis, a pre-cursor of Bitot's spots, were found in a representative sample of 122 pre-school children (ENI 1996). The World Health Organization has set a critical value for the prevalence rates of Bitot's spots, above which vitamin A deficiency should be considered a public health problem (West 1994). This value is exceeded by a multiple of twenty times in the children at the study site.

Most of these eye problems, however, are usually reversible when sufficient vitamin A is supplied. So long as children do not suffer from certain health problems, especially intestinal parasites, any increase of the micronutrient vitamin A in the diet will enhance their nutritional well-being appreciably. It is therefore promising to observe that the retinal content of foods consumed by CBC households is increased significantly. Retinal is a precursor of vitamin A exclusively derived from animal products.

In clinical assessment during the study, a higher number of eye problems were identified in LBC children. These eye problems are directly related to vitamin A deficiency. It is therefore encouraging to observe that children of CBC households, who have a comparably higher share of dairy products in their diets, suffer less from eye problems. It is possible that these children directly benefit from a high dietary content of retinal, a precursor of vitamin A, which animal products provide in a high bioavailability.

Ethiopia is one of the countries with the highest prevalence of stunting² among pre-school children in the world. The immediate factors attributed to this high incidence of chronic malnutrition are insufficient food consumption in terms of quantity and quality, as well as poor health. The underlying factors are poverty-related and encompass a wide spectrum from inadequate income, lack of education, prevalence of infectious diseases, harmful infant feeding practices, to inadequate sanitation and hygiene (ENI 1993). Linkages between improvements in income and changes in the farming system related to intensification and commercialisation of dairy production are complex and can alter immediate and underlying factors of child nutrition in a number of ways.

2.    Stunting is height for age measurement indicating the prevalence of chronic malnutrition by standard deviations from the median of the reference population; standard deviations ≤–2 indicates stunting (Gibson 1994).

The assessment of changes in the nutritional status of 152 pre-school children of the surveyed households age 6 months to 6 years was conducted by means of anthropometric measurements. By the end of the one year study period, stunting, a parameter used to indicate chronic malnutrition, was found in 33% of all children (Table 7). This is far lower than the average of 64% for all children of this age group in rural Ethiopia (ENI 1993). There was a significantly lower prevalence of stunting in children of CBC households compared to LBC households; 20% compared to 43%, respectively. Wasting³, a measure to indicate acute food deficit, was found in only 2 of the children of the CBC and 2 in the LBC households in the sample survey. Furthermore, there were no significant differences in morbidity between children of CBC and LBC households (also see the paper by Jemal Haider et al in this volume).

3.    Wasting is a weight for height measurement indicating the prevalence of acute malnutrition by standard deviations from the median of the reference population; standard deviations ≤–2 indicates wasting (Gibson 1994).

Table 7. Prevalence of chronic malnutrition among pre-school children of CBC and LBC households in quarter IV of 1997.                                                                                 

* Significant at the 0.05 level.

Summary and conclusions

Members of households with crossbred cows on average consumed 17% more calories, 24% more fat and 13% more protein than members of households with no improved cows. The households with crossbred cows spent about 7% more on food purchases, and in addition, they consumed about 30% more pulses.

The hypothesis that increased marketing activity would lead to less milk available for home consumption is rejected. Milk consumption in households with crossbred cows was found to be more than four times that in households with unimproved cows, even immediately after the crop harvest, when food availability is greatest. Children consumed most of the milk, men the next most and then women, but all family members benefited significantly.

Other studies in other areas have reported that as cash crops are introduced in smallholder production systems, with greater integration into the market women may lose control over cash income to men, who tend to spend less on food for the household. A key concern is that milk may be seen as a 'cash crop' as milk yields increase. When this happens, men may take over marketing the milk and women, who typically spend their income on food and household essentials, may lose control of the income. Contrary to the findings of other studies, these results indicate that women in households adopting improved crossbred cows maintain control over income allocated for food purchases and make over 80% of household expenditures on food. Furthermore, the men in these households spend some 28% more on food than men in households with no crossbred cows.

The anthropometric data point to the fact that the introduction of crossbred cows could significantly improve human nutrition and health. By the end of the study, stunting of children (height for age) was found to be less than half as prevalent in households with crossbred cows (20%) as in those households with local cows (43%). Stunting, a measure of chronic malnutrition, is related to poverty and chronic illness as well as inadequate food. Chronic illness did not vary between households keeping crossbred animals and those not, but income was greater and more and better food was available from both food purchases and production, leading to the reduction in stunting.

The increased production and incomes from improved dairying make important contributions to three dimensions of food security-availability, access, and stability. Beyond increasing availability of and access to nutritious food directly through both increased production of dairy products and higher income, the large increases in both dairy products and income made possible by improved dairying can make vital contributions to the stability of smallholder household consumption. The far higher production response to improved feeding of dairy cattle made possible by the use of exotic breeds rather than local animals results in a more steady flow of milk production and income.

Results from this study of the human nutrition and health impacts of market-oriented dairy production with crossbred cows show that this food-based intervention has the potential to make a significant positive impact on human nutrition and health status. The roles and importance of the factors other than crossbred cow ownership involved in this potentially dramatic change in production practices will be investigated in the next stage of the analysis.

References

Alderman H. 1987. Cooperative dairy development in Karnataka, India: An assessment. Research Report 64. IFPRI (International Food Policy Research Institute), Washington, D.C., USA. 60 pp.

Binswanger H.P. and von Braun J. 1991. Technological change and commercialization in agriculture. The World Bank research observer 57–80.

von Braun J. 1995. Agricultural commercialization: Impacts on income and nutrition and implications for policy. Food policy 20(3):187–202.

von Braun J. and Kennedy E. 1994. The effects of agricultural commercialization on nutrition. Johns Hopkins University Press, Baltimore, Maryland, USA.

von Braun J., Hotchkiss D. and Maarten I. 1989. Non-traditional export crops in Guatemala: Effects on production, income, and nutrition. IFPRI Research Report 73. IFPRI (International Food Policy Research Institute), Washington, D.C., USA. 99 pp.

ENI (Ethiopian Nutrition Institute). 1984. Nutritional tables for Ethiopia. ENI, Addis Ababa, Ethiopia.

ENI (Ethiopian Nutrition Institute). 1993. Nutritional survey, 1990–1991. ENI, Addis Ababa, Ethiopia.

ENI (Ethiopian Nutrition Institute). 1996. Nutritional survey. ENI, Addis Ababa, Ethiopia.

FAO/WHO (Food and Agriculture Organization of the United Nations/World Health Organization). 1985. Coefficients for converting family size into standardized household size. FAO, Rome, Italy. 87 pp.

Gibson R.D. 1994. Principles of nutritional assessment. Oxford University Press, Oxford, UK. 691 pp.

Gillespie S. and Mason J. 1994. Controlling Vitamin A deficiency. State-of-the-art Report 14. CC/SCN, UN (United Nations), Geneva, Switzerland.

Gryseels G. 1988. Role of livestock on mixed smallholder farms in the Ethiopian highlands. PhD thesis, Wageningen Agricultural University, Wageningen, The Netherlands. 249 pp.

Gryseels G. and Whalen I.T. 1984. Incorporating nutritional goals into international agricultural research-An ILCA perspective. In: Pinstrup-Andersen P., Berg A. and Forman M. (eds), International agricultural research and human nutrition. IFPRI (International Food Policy Research Institute), Washington, D.C., USA. pp. 141–152.

Jahnke H. 1988. Livestock production and livestock development in tropical Africa. Weltforum Verlag, Munich, Germany.

Jemal Haider, Shapiro B.I., Tsegaye Demisse, Alemu G/Wold. 2000. The nutritional and health status of women and children in households with and without crossbred cows in Holetta Wereda, Ethiopia. In: Jabbar M.A., Peden D.G., Mohamed Saleem M.A. and Li Pun H. (eds), Agro-ecosystem, natural resources management and human health related research in East Africa. Proceedings of an IDRC–ILRI international workshop held at ILRI, Addis Ababa, Ethiopia, 11–15 May 1998. ILRI (International Livestock Research Institute), Nairobi, Kenya. pp. 124–136.

Kennedy T.E. 1987. Income and nutritional effects of the commercialization of agriculture in Southwestern Kenya. Research Report 63. IFPRI (International Food Policy Research Institute), Washington, D.C., USA. 60 pp.

Kennedy E., Haddad L. and Jacinto E. 1994. The role of women's status in determining marital fertility and nutrition: A case study from rural Kenya. Report to the Rockefeller Foundation. IFPRI (International Food Policy Research Institute), Washington, D.C., USA.

Mengistu Buta. 1997. Use of crossbred cows for milk and traction in the highlands ecoregion: A whole-farm evaluation. MSc thesis, School of Graduate Studies, Alemaya University of Agriculture, Alemaya, Ethiopia. 87 pp.

MoA/FINNIDA (Mininstry of Agriculture/Finnish International Development Agency). 1991. Study on women in dairy production. Selale peasant dairy development pilot project. Report 6. MoA and FINNIDA, Addis Ababa, Ethiopia.

Neumann C., Bwibo N.O. and Sigman M. 1993. Diet quantity and quality: Functional effects on rural Kenyan families. Kenya Project Final Report, Phase II-1989–1992 Human Nutrition Collaborative Research Support Program. Office of nutrition, USAID (United States Agency for International Development), Washington, D.C., USA. Various pages.

Norman D.W., Newman M.D. and Ouedraogo I. 1981. Farm and village production systems in the semi-arid tropics of West Africa: An interpretive review of research. ICRISAT Research Bulletin Volume 1(4). ICRISAT (International Crops Research Institute for the Semi-Arid Tropics), Patancheru, Andhra Pradesh, India. 94 pp.

de Pee S., West C.E. and Haustvast J. 1995. Lack of improvement in Vitamin A status with increased consumption of dark-green leafy vegetables. The Lancet 346:75–81.

Quisumbing A., Brown L.R., Feldstein H.S., Haddad L. and Pena C. 1995. Women: The key to food security. IFPRI (International Food Policy Research Institute), Washington, D.C., USA. 22 pp.

Sanders J.H., Shapiro B.I. and Ramaswamy S. 1996. The economics of agricultural technology in semi-arid sub-Saharan Africa. Johns Hopkins University Press, Baltimore, Maryland, USA. 303 pp.

Wagenaar-Brouwer M. 1986. Nutrition and health of Ethiopian mixed smallholder farmers in relation to the use of crossbred cows. Draft Report. ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia.

Walshe M.J., Grindle J., Nell A. and Bachmann M. 1991. Dairy development in sub-Saharan Africa: A study of issues and options. World Bank Technical Paper 135. World Bank, Washington, D.C., USA. 94 pp.

West K. 1994. Vitamin A deficiency: Its epidemiology and relation to child mortality and morbidity. In: Blomhoff R. (ed), Vitamin A in health and disease. Marcel Dekker, Inc., New York, USA. 585–614.

Whalen I.T. 1983. Household decision-making and household objectives of participating farmers in Debre Zeit. Preliminary Report. ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia.

Whalen I.T. 1984. ILCA's Ethiopian highlands programme: Problems and perspectives in expanding the participation of women. Paper presented at ILCA/IITA, Ford Foundation workshop on women in agriculture in West Africa, Ibadan, Nigeria, 7–9 May 1984. 24 pp.