General Notes

Geographical classification

Sub-Saharan Africa is usually defined as those states located south of the Saharan desert, but excluding the Republic of South Africa. In this handbook, for a variety of practical reasons, a number of other states, i.e, Djibouti, Equatorial Guinea, Namibia, Sao Tome, Reunion and Comoros, have also been excluded. The thirty-nine states which are included match the list contained in the World Bank's (1984) Towards Sustained Development in Africa. Annex 2 lists the countries by region and their "country codes" in the FAO Agricultural Production and Trade standard tapes from which many of the data are derived.

The classification and allocation of countries to regions (West, Central, East and Southern) differs between agencies; and the user of this handbook should carefully compare the allocation of Annex 2 with any other allocation with which he or she may wish to compare performance.

Income groups

The World Bank and other agencies often classify countries into income groups in order to compare performance between groups. Most of the tables in this handbook also do so, and the allocation of individual countries into groups follows that of the World Bank (1984) and is set out in Annex 1.

Regional and income group totals

In a number of tables data are not available for all the countries in a region or income group. Where that occurs, the totals in the tables of this handbook are totals for those countries for which data are available and no attempt is made to estimate what the data for the missing countries might have been.

Abbreviations

Abbreviations specific to particular tables are explained in the footnotes to those tables. The following abbreviations are standard throughout:

Livestock numbers and slaughter

The numbers given refer to all animals of the species indicated irrespective of their age or the place and purpose of breeding. An animal slaughtered refers to any animal killed within the country regardless of its origin. The data on meat production include both commercial and farm slaughter, given in dressed carcass weight. Carcass weight excludes the skin, offal and slaughter fats. The data on milk production excludes milk suckled by calves but includes (to the extent that data are available) milk fed to calves or other livestock.

Meat ouput

The meat output data (eg in Tables 4-9) refer to animals slaughtered in the country. They include, therefore, the meat from animals imported alive and slaughtered within the country, and exclude meat from animals exported from the country alive for slaughter elsewhere.

Meat and milk conversion factors

In a number of trade and consumption tables (eg 18-25,28-35 and 36-41) meat and milk products are aggregrated. In these aggregations the following conversion factors are used to aggregate dissimilar products.

For meats: in carcass equivalents

For milk:

In order to be able to compare in quantitative as well as in value terms, the relative importance of the different dairy products into which raw milk can be processed, the trade data for different milk products (in Tables 28-35) are converted to their liquid (fresh) milk equivalents (FAO, 1978).

The conversion factors are expressed as kilograms of whole liquid milk equivalent (LME) per kilogram of milk product:

Tropical Ruminant Livestock Units (TRLU)

A TRLU is the standard unit by which livestock of different species are compared. It is defined as an animal with the equivalent of 250 kg liveweight. Cattle, sheep, goats and camels are converted, each with a conversion factor, to the standard unit as follows:

Equines (horses, mules and asses), pigs and poultry are not ruminants and are not included in the TRLU totals. For the present purposes camels are defined as ruminants.

Animal health services expenditures

These are the total expenses incurred in trying to fulfill policy objectives such as the eradication of livestock disease, control through vaccination and research. Services may be paid for directly by the user, or indirectly through taxes and charges levied on livestock-related activities, for example, on production or export.

Relative changes and growth rates

A major purpose of this handbook is to present data in a summary form which enables the performance of different African countries to be compared. The main form of summarisation selected is the annual compound growth rate in two separate periods, l 961-1974 and 1975-1987, of just over a decade each. The selection of the periods was primarily determined by the availability of data. However, the choice of the break point between the two periods, at 1974/1975, is convenient in that it means that the second period starts at the point at which in many African countries, herds began their recovery from the great Sahelian drought of the late 1960s and early 1970s.

A comparision in terms of growth rates has both advantages and disadvantages. A major disadvantage occurs when changes are cyclical rather than linear and the peak or trough of the cycle occurs in the middle rather than at one end of the period over which the growth rate is being computed. The sign and size of the growth rate over the period as a whole are then poorly representative of reality. The growth rates of dairy imports in West Africa during the period 1975-87 is an example of this problem.

All annual growth rates are computed by least squares regression from log transformation of the data for each year of the period. The untransformed formula is:

Y_t = Y_o (1+r)^t

Where

Y_t = value of variable Y at time t
Y_o = value of variable Y at the beginning year
r = annual rate of change in variable Y

The consumption tables (Tables 36-41) were derived from the production and trade tables, i.e, consumption = production + net imports. Trade in pig meat in SSA is extremely small and has not been taken into consideration. In the most recent years, to include trade in pig meat would affect significantly the pork consumption figures in Cameroon and Côte d'Ivoire only. For this reason, the pig meat consumption table (Table 38) does not include net imports.

Prices

Two sets of price tables are presented. One set, Tables 43-45, present long-term (16-17 years) producer price series for three commodities, beef, mutton and cow milk, for 3-5 (depending on the commodity) selected countries (Côte d'Ivoire, Mali, Nigeria, Sudan, Zimbabwe). The prices for each commodity are quoted in local currencies at current prices and at constant prices, and also in US dollar equivalents, using official exchange rates to convert from local currencies to US dollars. These tables contain relatively high quality and reliable data. An ILCA researcher visited each country concerned in 1988 and had discussions with informed persons about the quality of the data. The data are discussed, especially in Appendix 1, in the source document which is: Williams T O. 1990. Impact of livestock pricing policies on meat and milk output in sub-Saharan African countries. LED Working Document 13. ILCA, Addis Ababa.

The countries involved in the first set of data were deliberately chosen to include seven countries which account for more than 50% of Africa's livestock output. In the event it was not possible to obtain all the required data (and in particular a sufficiently long series of data) on two of these seven, ie Ethiopia and Kenya. Tables 43-45 are not discussed further in these notes.

The other set of price data are presented in Tables 46 51. Some of the data were collected from published documents which are cited in the tables. Some were collected by postal questionnaire from officials already known to ILCA in the individual countries. An effort is made in these tables to distinguish between prices paid for different types (sex) of stock, at different seasons, and at different market levels between the producer ("farm gate") and the eventual consumer ("retail"). Some of these data are of high quality and reliability; some may not be, although a screening process was employed to eliminate apparently unreliable data. The detailed notes which follow apply only to this second set of data.

The countries involved in the second set of data were more "self-selective", in that those countries are included for which data were relatively easily obtainable for at least two years after 1980. The second set of data were sought and analysed by ILCA in 1987/88.

This second set of data on prices are compiled in six separate tables.

1 - cattle prices (Table 46)
2 - sheep prices (Table 47)
3 - goat prices (Table 48)
4 - milk prices (Table 49)
5 - feed prices (Table 50)
6 - feed price ratios (Table 51)

Differentiation of prices

Whenever possible, prices are presented by sex of the animals. All categories of the animals designated in these tables refer (unless specified) to the "average marketed representative" of their sexes. Thus cattle (male) refers to the average ox, cattle (female) to the average cow and so on. In cases where distinction is not made between sexes, the data are assumed to refer to the "average" of the animal of that species offered in the markets. In cases where the animals seem to deviate from the assumed "average", specific explanations are given (eg steer, lamb, etc.)

An attempt was made to distinguish between prices of high and low price seasons and annual average prices. Those prices marked with "H" are prices which prevailed during high price seasons and those with "L" during low price seasons. Prices which it was not possible to put into any of the price seasons are marked with a query mark "?". Annual average prices are not marked at all.

Prices paid at any one time differ from place to place within a country depending on the kind of market or "level" in the marketing system at which transactions take place. Five such "levels" were distinguished for livestock sales:

a) At the livestock owner's house or camp: commonly referred to as "farm gate".

b) At "primary" markets - in rural areas where most of the livestock are being sold by producers, either to other producers or to traders.

c) At "intermediate" markets, which may be in rural areas or in towns - markets where most of the livestock are being sold by traders who have previously purchased them at "primary markets", often to other traders who will take them on to "terminal" markets.

d) At "terminal" markets - where most of the livestock are being sold by traders to butchers or abattoirs.

e) At "abattoirs/wholesale butchers" - where abattoirs or other slaughterers sell whole carcasses.

For sales of cow milk three levels of sales were distinguished:

a) In "remote areas" - prices paid for milk sold in remote rural areas for making into butter or butteroil (ghee) or the value of milk that is equivalent to the price of local butter sold in such areas (ie if it takes 20 litres of milk to make 1 kg of butter, and butter sells locally for $20 per kg, then the "butter-equivalent value" of milk is $1 per litre).

b) In "urban catchments" - prices paid to farmers for raw fresh milk at "farm gate" or rural collection centres for milk which will be retailed as fresh milk in large cities or towns.

c) "At urban retailers" - prices paid by consumers in cities or major towns for fresh milk of not less than 2% butter-fat content.

Livestock prices

Prices are given in terms of liveweight on a per kg basis. When the original data on prices were obtained in terms of carcass weight such prices were divided by 2 (assuming 50% carcass weight) for cattle and by 1.8182 (assuming a 55% carcass weight) for small ruminants to arrive at prices per kg liveweight. In a few cases the original data were expressed "per head" for live animals. To convert such prices into prices per kg liveweight, liveweights of 300 and 250, 24 and 22, and 22 and 20 kg were assumed for the males and females of cattle, sheep and goats, respectively.

Milk prices

All milk reported here was assumed to have a butterfat content of not less than 2%. Data on milk from FAO "Prices received by farmers" are put in the urban catchment column. This is because "remote areas" refer only to areas which do not have access to regular milk markets (outlets). These FAO data are given on a per ton basis and are divided by 1000 to get milk price per litre. The implication is that equality is assumed between kg and litre.