4.1. Introduction
4.2. The survey of acquisition and consumption patterns for dairy products
4.3. The efficiency of the alternative dairy marketing systems
4.4. Implications for government policy
The dairy subsector in Sub-Saharan Africa has performed badly over the last two to three decades. Demand for dairy products has increased, but per caput milk production has fallen. The consequence has been a rapid rise in imports of dairy commodities, either on commercial terms or as food aid. Dependence on food aid is not a sensible economic strategy. Most countries in sub-Saharan Africa suffer acute shortages of foreign exchange and can not afford to spend the little foreign exchange that they have on importation of agricultural commodities which they could produce themselves. Increased emphasis on domestic dairy production is thus necessary.
Increased domestic dairy production requires adequate market outlets to handle the increased output, to provide the right incentives to producers and market intermediaries, and to provide the right kind of signals to encourage the production of those commodities and services that consumers want at prices they can afford. An appropriate marketing system is an essential accompaniment of increased domestic dairy production.
There is no single type of dairy marketing system that is unique. Alternative marketing systems can (and do) differ in the services offered, in the size and ownership (e.g. government, cooperative or private) of the organisations that constitute them, and in their degree of vertical integration. Different types of marketing systems almost certainly differ in their effectiveness in serving different policy goals and in different circumstances. However, in sub-Saharan Africa very few studies analysed experiences with different dairy marketing systems, particularly with reference to marketing efficiency.
Efficiency of alternative dairy marketing systems is an important policy issue in livestock development. Appropriate systems should be developed to accompany dairy development projects under different sets of circumstances. This report presents the results and policy implications of the study of the relative efficiency of alternative dairy marketing systems in Addis Ababa, Ethiopia. This study has identified the major alternative marketing systems for fresh liquid milk and butter that are available to consuming households in Addis Ababa and examined the use made of these systems by different income classes in different zones of the city. The study has also examined the reasons cited by consumers for their decision to use particular marketing systems. The performance of the alternative marketing systems is evaluated in terms of four major partial efficiency criteria when objective measurements of performance are available, i.e. price levels, butterfat content, total solids and reliability of supply. Convenience of a particular source when collecting milk appeared to be the most important reason for choosing one marketing system rather than another. However, it was not possible to measure performance objectively in terms of the "consumers convenience" criterion. Therefore, "consumers convenience" does not form part of the efficiency criteria used in the study, except by looking at it from the "systems popularity" point of view. A summary of the major findings and their policy implications are presented hereafter.
The households survey covered a random sample of 3,888 households, each of which was visited and interviewed only once, in the exploratory survey, to ascertain its monthly cash income, its consumption patterns for dairy products (especially for fresh milk), and whether such products were supplied by its own production, by purchase, or as gifts. A sub-sample of 482 households was then selected and each was visited and interviewed again once (initial survey), and then daily for seven days to elicit greater detail on the general patterns of purchases and consumption of dairy products (subsequent survey). Individual households in the sub-sample were randomly selected from amongst those with a high probability of being milk purchasers, as determined through discriminant analysis of the results of the exploratory survey.
The exploratory survey indicated that over 80% of all households fall within the "low income group", each having an average monthly cash income of less than 250 Ethiopian birr per month. Over 70% of all households were found to acquire either no milk at all, or to do so very seldom. Of the consumers, nearly 90% buy their dairy products, while most of the rest keep their own cows. Gifts of dairy products from relatives, friends or employers are insignificant. Similarly, purchases of milk in powder, condensed, or evaporated form are negligible. The Annex Tables A.1 - A.10 give a summary of the details of the results of the exploratory survey and the initial survey.
An initial survey of butter trading indicated that, of the three types marketed, the acquisition of table butter was not widespread, with only 18 out of the 482 sample households reporting positively. About 2.5% of the households' outlay on butter was found to be spent on cosmetic butter? but the term "cosmetic" is descriptive rather than substantive in that most households bought cooking butter and then converted it for use for cosmetic purposes. Cooking butter, however, was bought by 115 households out of the sample of 482 during the week prior to the interview day during the subsequent survey. Thus the present study necessarily discusses the acquisition and purchase patterns of cooking butter. As with fresh milk, most butter is bought. About 98% of acquisition occasions were through purchase, and only 2% involved situations where households processed it either from own-produced or purchased raw milk or received it as gifts.
The study found that other dairy products are also consumed in Addis Ababa, notably: hard cheese and Ayib (cottage cheese), yoghurt and Ergo (a sour milk which is the Ethiopian equivalent of yoghurt). It was not possible to separate yoghurt and Ergo as distinct products in the survey, since the Ethiopians treat Ergo as if it is the same product as yoghurt.
About 92% of the households who consume cheese/Ayib buy it, while 3% receive it as gifts and the rest process it from either own-produced or purchased raw milk. The acquisition patterns for yoghurt/Ergo are strikingly different from those for fresh milk and butter, with the majority of the households (i.e. 52%) processing it from purchased raw milk and 8% processing it from their own-produced raw milk. Only 35% of the acquisition occasions are purchases, while 5% are gifts.
Introduction
Milk marketing systems
Butter marketing systems
A major objective of the survey was to identify alternative marketing systems for dairy products in Addis Ababa, to quantify the amounts of the different products flowing through them and to evaluate their efficiency. Such marketing systems are here defined on the basis of the chains linking producers to consumers, and identified in terms of the final retail outlets from which the households purchase their supplies. Identification and analysis of the alternative marketing systems was only carried out for the liquid milk and butter trades, as far fewer households were found to purchase and consume the other dairy products in any significant amounts.
Five major marketing systems for fresh milk were identified:
(i) DPR: sales of liquid milk by the actual producers.
(ii) KEB: sales by Kebele shops and other DDE-designated sales outlets.
(iii) ITR: sales by itinerant traders.
(iv) KIO: sales by small private shops and kiosks.
(v) GRO: sales by grocer stores and supermarkets.
All transactions were found to be direct sales, implying that the relevant outlet was dealing directly with the purchasing household or its agent. Three subsystems make up the Producers-Consumer chain (i.e. the DPR system), depending on where the producers' cows are kept, which may be local to the consumer, elsewhere within the city, or outside Addis Ababa. This latter source was found to be relatively insignificant, however, in terms of the volume of milk sold through it and the number of households using the sub-system. Direct Producer-to-Consumer sales represent about 71% by volume of all the fresh milk sold, and it is striking that the majority of the producers keep their cows within the city boundaries.
An examination of the main reasons which influence households to purchase milk frequently from particular sources indicated that both higher and lower income group consumers consider (i) convenience in collection (ii) regularity and reliability of milk supply and (iii) the association of hygiene and cleanliness with particular sources to be the critical factors which influence their choice. The level of milk prices was not ranked highly by any class of consumers as a critical factor. A major objective of the government marketing policies in Ethiopia is to provide wholesome fresh milk at reasonably low prices, particularly for lower income families. The results of this study suggest, however, that this goal is only partially met, for there are sources other than the official outlets (represented by the system KEB) in every zone of the city which appear to be used more extensively and regularly (in terms of purchase occasions) even by the small proportion of the low income group who consume milk at all. This occurs despite the fact that the KEB outlets are one of the two cheapest sources of milk in every city zone.
An important concern of the case study was to assess the relative merits of alternative marketing systems. Because of both practical and theoretical difficulties, no attempt was made to calculate a single all-embracing index of economic efficiency. Instead a number of partial efficiency criteria were established. Systems were evaluated primarily in terms of the extent to which they met certain consumer demands or priorities or why they were chosen as the regular sources of household supply. The results formed the basis on which the relative efficiency of the different marketing systems was assessed.
Based on the level of prices charged to consumers in relation to the quality of the milk sold, the DPR system of Producer - Consumer sales was judged to be the most efficient, with the S13 subsystem (i.e. sales by producers whose cows are kept outside Addis Ababa) recording the best level of efficiency. Milk quality was judged on the basis of the interactive index between butter-fat and solids-non-fat content. On the basis of the level of prices not adjusted for quality, the S13 subsystem was again judged to be the most efficient, but this is a marketing chain which is very little used (less than 1% of market volume). On the basis of prices not adjusted for quality, the KEB system of sales outlets was the cheapest source of purchased milk that is significantly patronised in all the three zones.
Another possible means of judging the relative efficiency of different marketing systems is by their relative popularity, or market share, which is assumed to reflect accurately their relative efficiency in responding to consumer demands. Accordingly, direct Producer - to Consumer sales (system DPR) are the most efficient, when market share is measured in volume terms and when households are not classified by income group. It is also the most efficient for all income classes, in all zones, except among low-income households in the inner city, where the system KEB (Kebele shops/DDE sales outlets) is the most popular.
The butter marketing systems were identified and defined as:
(i) MER: direct sales of butter by traders in the Mercato;
(ii) OMK: sales of butter by traders in markets other than the Mercato;
(iii) GRO: sales of butter by grocery stores and supermarkets;
(iv) IAP: sales of butter by producers who keep their cows within Addis Ababa;
(v) XAP: sales of butter by producers who keep their cows outside Addis Ababa;
(vi) ITR: sales of butter by itinerant traders;
(vii) KEB: sales of butter by Kebele shops or other DDE-designated sales outlets;
(viii) KIO: sales of butter by small private shops and kiosks.
About 87% of the cooking butter sold is handled through the Mercato, showing that the market for cooking butter is highly centralised in space.
An evaluation of the main reasons which influence the choice of regular butter sources highlighted the following factors:
(i) regularity and reliability of supply;(ii) convenience, primarily in terms of the accessibility of the source, either from the domicile or from a place where a household member goes on business;
(iii) association of hygiene and cleanliness with the source.
Both lower and higher income groups purchase their cooking butter from the Mercato (system MER) on a regular basis, irrespective of the zone of the city in which their households are located. The Mercato is in the inner zone of the city and cannot be regarded as being convenient to every consuming household. The choice of the Mercato as the source, therefore, must primarily rest on the regularity and reliability of supply. Since butter can be purchased in bulk and stored, the proximity of the butter retail source to the household location is not as important as in the case of the milk retail source.
System MER handles over 85% of the total sales in any given week, so that only tentative general evaluations and comments on the merits of the other systems can be made. There is a high degree of regularity in the use of the Mercato by both the lower and the higher income groups in all zones. Only 10% of the higher income households in the middle zone of the city were found to have bought butter from sources other than the usual source. Further, only minor variations in mean consumer price levels were observed across the zones of the city, where alternative marketing systems for cooking butter were in use by different income classes of consumers.
The lower income consumers in Addis Ababa were found to be using the following butter marketing systems:
· system MER in the inner zone of the city;
· systems MER, OMK, IAP and ITR in the middle zone of the city;
· systems MER, OMK, IAP and XAP in the peripheral zone
For the lower income consumers in the ICZ, only the MER system is used and so no comparison of prices with different marketing systems is feasible. The system ITR, however, is the cheapest source for the low income group in the MCZ, with the MER system, the dominant system, being second. In the PCZ, the system XAP is the cheapest source, with the system OMK being second and the MER being third. It appears that the Mercato is always more costly in comparison with the other systems present in a given zone of the city, but there is no quality standard for butter, so that quality-adjusted comparison is difficult.
For the higher income consumers, the following systems were found to be in use:
· systems MER and IAP in the inner zone of the city;
· all systems, except XAP and KEB in the middle zone;
· all systems, except GRO, XAP and KEB, in the peripheral zone.
On the basis of mean consumer price levels: (i) the system IAP is the cheapest source of butter in the ICZ; (ii) the system KIO is the cheapest source in the MCZ, followed by the system OMK; (iii) the systems KIO and ITR are the cheapest sources of butter in the PCZ. Hence the dominant MER system does not appear to be the cheapest source of butter in any zone of the city.
The difference between the prices being charged in the different systems are relatively small and, either for this reason or, perhaps, independently of it, price level did not rank very highly among the reasons cited by consumers for choosing to use one marketing system rather than another. In terms of "popularity" (i.e. as judged from the market share), the MER system is overwhelmingly the most popular in all zones and for both income classes, accounting for 87% of supply.
An important concern of the study was to assess the relative merits of alternative dairy marketing systems. As noted, no attempt was made to calculate a single all-embracing index of economic efficiency as a measure of such merits, because of both practical and theoretical difficulties. Instead, a number of partial efficiency criteria were established as discussed. Evaluations based on these partial efficiency criteria formed the basis from which the relative efficiency of the different dairy marketing systems in Addis Ababa is assessed.
The analysis has been consumer-oriented, evaluating different marketing systems in terms of specific criteria of interest to different classes of consumers, e.g. least cost and reliability. The results of the study, however, have implications for government policy, and there may be interesting conclusions to be drawn from them by several governments in sub-Saharan Africa.
An important observation from this study is that the Addis Ababa households did not rank prices highly as one of the main factors which influence their choice of the marketing system from which to purchase their dairy products. This factor suggests that Addis Ababa consumers do not consider dairy price levels as being high, despite the large variation of prices between different sources of dairy products. This further suggests that Addis Ababa consumers could still absorb some increases in consumer prices of dairy products without choking off their demand. This factor has important policy implications.
The Ethiopian government's policy objectives for dairy development or marketing systems in Ethiopia are not explicitly stated in any single document, but the main ones may be deduced from various statements in government documents as being:
(i) Increased domestic milk output, in order to improve the farmers' incomes and to reduce the parastatal DDE's reliance on imported raw materials. These imports were being provided substantially on concessional terms at the time of the survey (1985), but this situation may not last.(ii) Increased consumption of nutritive milk products by the lower income rural and urban groups.
(iii) Attainment of a reasonable level of hygiene for the dairy products that are consumed.
Within those sections of the different marketing systems which have been covered by this study, the potential instruments available for pursuing these objectives are as follows:
(i) To alter, by administrative act, the price at which milk processed by the parastatal DDE is sold through government-controlled retail outlets. At present, this milk amounts to about 15-18%1 of the total volume sold from all sources to consuming households in Addis Ababa, and it is bought by about 22-24% of both low and high income milk purchasers.(ii) To regulate, by legislation, the price at which other outlets sell to consumers.
(iii) To compel, by legislation, all producers, and in particular the intra-urban producers, to sell their milk to the DDE for processing and resale. Intra-urban producers account for 70% of the milk purchased by households in Addis Ababa and are the usual source of milk for about 40% of low income consumers and 50% of high income consumers.
(iv) To widen the range of dairy products sold through the government-controlled (i.e. official) outlets, e.g. Kebele shops. Cooking butter, although the sampling procedures used do not permit an accurate estimate, accounts for approximately 80% of the total volume of milk equivalent that finds its way in one form or another (butter, cheese, etc.) to consuming households in Addis Ababa2.
1 Sales through Kebele shops and other outlets directly controlled by the DDE amount to 15% of the volume of milk sold to consuming households. In addition, some groceries and supermarkets are government owned. All grocers and supermarkets together account for 3% of milk sales to consumers2 This is based on the premises that follow. The total sample of 482 households covered in the initial and subsequent surveys purchased 1241 litres of milk and 242 kilograms of cooking butter during the course of the survey weeks; Ayib, cheese, yoghurt, Ergo and table butter purchases are ignored. Approximately 23 litres of milk is needed to make 1 kg of Ethiopian cooking butter. It should be noted that the estimate is very rough because:
· The sample was biased towards including those who mostly purchase milk. It is not known whether the ratio between milk and butter consumption in this sample is the same as in the population at large.· Whereas milk is necessarily consumed at about the same time as it is purchased, butter can be stored over long periods; and it is not known whether the period of the survey was one, on balance, of stocking up or destocking in respect of households' purchases and consumption of butter.
There is a good case for rationalization of a more effective regulation of the dairy industry in Ethiopia. The Addis Ababa milk catchment area, and Ethiopian in general, is not self-sufficient in dairy products at current prices. The Ethiopia government's Dairy Development Enterprise (DDE) often operates its main Addis Ababa dairy plant below capacity, and the plant often utilises dairy recombination material (milk powder and butter-oil) to improve on capacity utilization. Such dairy recombination material, in the recent past, has been procured as food aid via the World Food Programme (WFP), and the proceeds from the sale of products made from such recombination material are expected to be channelled into dairy development projects in the country (see Mbogoh, 1984).
These factors and the results of this study suggest that there is some scope for a rise in the official price of milk at these government-controlled outlets without serious adverse consequences. Such a price rise might not greatly affect the actual prices paid by customers at these outlets since it might occur mainly at the expense of the unofficial "premium" of 10% -25% paid at the official outlets. If the actual retail price, however, did rise, only a small proportion of low-income households in Addis Ababa would probably suffer. There are probably just over 200,000 households in Addis Ababa, of whom about 160,000 are "low-income". Of these low-income households, only 22% (i.e. about 35,000) acquire milk at all, but only 23% of these (i.e. about 8,000 families) do so from the official outlets, i.e. the Kebele shops and DDE sales outlets. A 10 cents (14%) rise in the official retail price of milk per litre would, on average, cost these families the equivalent of about Eth. birr 1 per month.
A rise in the official retail price to consuming households might have two consequences. It would enable a bigger profit to be achieved from the sale of reconstituted milk made from the raw materials acquired free through food aid, and such a profit might be reinvested in dairy development. Secondly, it would enable higher prices to be paid to the milk producers around Addis Ababa, giving them greater financial incentive to produce more milk. It is beyond the scope of this study to predict whether the additional profits would be usefully invested or how big an effect on milk output a 10-cent per litre rise in the producer price would have. Neither the present total amount of food aid, nor the total present amount of milk collected by the DDE from the producers, would justify an expectation of a very dramatic effect in absolute terms.
Both low and high income groups of consuming households were found to be using other retail sources of milk considerably more frequently than they use the Kebele shops and other official DDE sales outlets, even though the latter offer milk for sale at lower prices. This factor and the reasons given for choice of source indicate that milk consumers are interested in things other than just obtaining the lowest price. A government policy which took these other considerations into account might best serve the interests of both low and high income classes. Particular attention might best be directed at:
· customer's convenience when collecting the product;
· regularity and reliability of supply;
· hygiene and clean premises from which milk and other dairy products are to be sold.
These appear to be the services most sought after by customers.
The results of this study have clearly demonstrated that no single organisational structure for dairy marketing systems can be said to be unique in terms of economic performance. There is need to evaluate local conditions and establish appropriate marketing systems to accompany dairy development projects in different regions or countries. A mix of private and government dairy marketing systems certainly appears to perform better than government marketing systems alone. In any case, the private dairy marketing systems were judged to be superior to government dairy marketing systems from the "consumer demand-responsiveness" criterion.
