CHAPTER 1 - URBANIZATION AND THE EXTROVERSION OF FOOD SUPPLY CIRCUITS

1.1 - Apparent parallels between huge urban growth and agrifood imports in sub-Saharan Africa (SSA)
1.2 - Links between urbanization and international supply and distribution circuits
1.3 - The rejection of false evidence regarding external urban supply circuits
1.4 - Typologies

1.1 - Apparent parallels between huge urban growth and agrifood imports in sub-Saharan Africa (SSA)

1.1.1 - The crisis in the agrifood systems
1.1.2 - Agrifood imports
1.1.3 - Rapid urbanization

It is obvious to many people that the urban explosion of the African countries is a major factor in the food crisis, creating a food deficit and making it necessary to resort to external food supply circuits.

1.1.1 - The crisis in the agrifood systems

According to official statistics, which are very inconsistent, Sub-Saharan Africa suffering from a crisis in the agrifood system as a whole, resulting in a sluggish growth in the availability of agricultural and food products. The nutritional status of the population appears to be declining. The number of people suffering from chronic undernutrition has apparently risen from 96 million (out of a total of 268 million in 1969/1971) to 204 million (out of 500 million in 1990/1992), which is over 49 per cent, and is likely to affect 300 million by the year 2010 (FAO forecasts). Average food availability was estimated to be around 2 150 calories a day in 1970 and only 2 200 in 1990 (FAO source cited by Azoulay & Dillon, 1993). Yet there are large differences between countries. Average food availability in Mozambique, Angola outside the French-speaking zone and Chad, Comoros, Burundi and Rwanda in French-speaking Africa is less than 2 000 calories.

Extensive agriculture, using fairly unsophisticated techniques, are vulnerable: according to FAO per capita food production, which remained constant throughout the ‘60s, has been falling by 1.2% per year since the early ‘70s. Taking 1970 as 100, the food production fell to 80 in 1991. Per capita food production for the whole of the SSA fell in terms of annual output (cereals, root crops) from 297.4 kilos (1974/1980) to 274.7 (1985/1987) and further still to 252.2 kilos (1990/1994). However, several other surveys have challenged these figures. WALTPS, for example, has estimated that the agricultural surplus marketed since 1960 has actually grown each year by 2.6%. Africa also has major supply problems within the various agrifood compartments (trade, storage, transport, processing, distribution, the lack of fluidity and rapidity in supplies, infrastructure, inappropriate legislation and the low productivity of small-scale units).

It is nevertheless necessary to define more accurately what is meant by food. Should it merely refer to a nutritive function, or should it also be seen as a system of social and symbolic relations? Should the agrifood chain merely analyse material operations of processing agricultural products, or should it be seen as an industrial process in which there are a variety of different branches and in which agriculture plays what is sometimes a secondary role? Is food insecurity due to a lack of food availability, to great instability in time and space, or is it not, as has been suggested by Sen (1981) due to a lack of rights (based upon trade, output from work, and gifts/transfers) giving rise to food access problems?

1.1.2 - Agrifood imports

The crisis in Africa’s agrifood systems can be gauged from the deteriorating agrifood balance[3]. Whereas food self-sufficiency stood at 98 per cent in 1960, by 1972 the figure had fallen to 90 per cent. In terms of value, agricultural imports increased 5.4-fold between 1970 and 1980, and food imports 7.2-fold. In terms of quantity, net cereals imports, which were negligible in the ‘50s, rose by 9 per cent per year throughout the ‘60s and ‘70s to reach 10.8 million tonnes of wheat and 4 million tonnes of other cereals by 1983/84, from 1974’s 5.1 million tonnes. Conversely, they fell back to 10.5 million tonnes in 1990 (World Bank). Cereals imports (maize, rice and wheat) appear to have risen from 507 million dollars at current prices in 1973 to 2 183 in 1980, 2 014 in 1985, 1 909 in 1991 and 2 631 in 1992. The actual figures, not counting Nigeria, were respectively: 446.9, 1 527, 1 526, 1 756 and 2 388. Food aid in the form of cereals totaling 2.4 million tonnes in 1980 rose to five million tonnes in 1984/85, before falling back to 2.8 million tonnes in 1990.

These figures, however, need to be seen in proportion. Food imports account for less than 10% of food consumption. The SSA food trade balance (in terms of nutrition value) is in equilibrium, and per capita food imports are low compared with other regions. Cereals imports rose as a percentage of goods and services exports by 3.8% (1973) to 3.8% (1980), 5.5% (1985), 4.2% (1991) and 5.2% (1992). Developments in French-speaking Africa vary enormously (see Table 1).

Table 1 - Cereals imports as a proportion of exports of goods and services

	1973	1985	1991
Benin	3.9	5.3	10.1
Burkina Faso	12	22.1	13.4
Cameroon	3.5	0.9	3.7
Congo	2.3	1.6	2.8
Côte d’Ivoire	5.8	3.0	4.9
Gabon	0.4	0.7	0.9
Mali	43.7	27.0	5.0
Niger	1.6	15.6	8.6
Central African Republic	7.9	3.2	6.7
Senegal	17.7	12.0	9.3
Chad	3.1	10.7	4.3

Source: Makaya, 1996.

Food dependency cannot, however, be likened to a mere food balance deficit. It is not only a matter of finished goods but all the inputs and imported equipment throughout the agrifood chain. Should external food dependency be defined as the impossibility of directly or indirectly meeting the demand for food consumption of own resources, or does it imply a lack of control over external relations and over the strategic links in the agrifood chains? Food imports not only constitute a balance that making it possible to bridge the gap between production and consumption. They are also competing commodities, which can stimulate the supply and are a source of income for those who manage to control them.

1.1.3 - Rapid urbanization

The process of urbanization in Africa is taking place very rapidly. While the SSA is still the least urbanized sub-continent in the world, the growth rate of the urban populations there is not only the most rapid in the world but it has tended to increase (5.3% on average throughout the ‘60s, and 6% during the ‘70s) before falling back during the ‘80s to around 4.5% (namely, the average natural growth rate plus 1.5%), mainly due to the decline in migration flows. Whereas the SSA was less than 10% urbanized thirty years ago, it is now undergoing an average urbanization rate in excess of 35%, and over 40% of the population will be living in the towns by the year 2000. Even though West Africa is twice as urbanized as Central Africa (20% v. 40%) the same internal demographic growth rate and the rural emigration trends are pushing the continent along.

While the population of the SSA countries almost tripled between 1950 and 1990 (from 164 million to 478 million) the urban population increased over ten-fold during the same period from 15 million to over 150 million (see Table 2).

Tableau 2 - Urban population growth in SSA

	1930	1950	1960	1970	1980	1990	2000*
Population of 45 SSA countries (m):	130	164	203	265	360	478	661
Rural population	122	149	176	207	257	323	400
Urban population
(Population over 5000)	8	15	27	52	103	155	260
Urbanization rate (%)	6	12	16	22	30	33	40
Town size:
> 1 million	0	1			10
> 100 000	10	28			160

* CERED FORECASTS
Source: Calculs à partir des statistiques Cour/ILTA, WALTPS, p. 7.

The urban explosion is accompanied by an uneven spatial distribution, with megalopolises growing with a weak urban backbone. There is a prevalence of main towns as opposed to a relatively weak network of secondary towns, while the large towns are very weakly locked into their rural surroundings. In the majority of African countries, the main towns are growing more rapidly than the secondary towns: the rate for the main towns has risen over twenty years from 35 to 45 per cent (Hugon & Pourtier, 1993). In 1980, SSA had 12 towns with more than one million inhabitants, 160 with more than 100 000 and about 3 000 with over 5 000 inhabitants (see Table 2).

Table 3 - Francophone Africa: total population, urban population (in millions), urbanization rate in 1960-1990

	1960			1980			1990			Urban growth rate
	Total population	Urban population	Urbanization rate	Total population	Urban population	Urbanization rate	Total population	Urban population	Urbanization rate	1980/90	1990/95
Benin	2.03	0.21	10.3	3.45	0.95	27.5	4.5	1.69	37.5	5.2	4.6
Burkina Faso	5.32	0.53	10.-	7.19	1.34	18.6	8.68	2.05	23.6	10.0	11.2
Burundi	2.46	0.09	3.7	4.12	0.19	4.6	5.33	0.28	5.2	6.9	6.6
Cameroon	4.48	0.79	17.6	8.64	2.54	29.4	11.48	4.75	41.4	5.4	4.9
Congo	0.83	0.29	34.9	1.64	0.91	55.5	2.30	1.64	71.3	5.9	4.9
Côte d’Ivoire	3.18	0.53	13.7	5.28	1.49	28.2	11.44	5.57	48.7	5.4	5.0
Gabon	0.45	0.09	20.-	0.73	0.27	37.-	0.94	0.55	58.5	6.0	4.7
Guinea	3.19	0.46	14.4	4.41	1.09	24.7	5.27	1.41	26.7	5.7	5.8
Mali	4.86	0.27	5.6	5.70	0.62	10.8	8.18	1.86	22.7	5.1	5.7
Mauritania	1.00	0.10	10.-	1.55	0.50	32.2	1.96	0.82	41.8	7.6	5.4
Niger	3.77	0.20	5.3	5.82	0.71	12.2	7.68	1.22	15.9	7.5	5.6
C.A.R.	1.25	0.26	20.8	2.00	0.76	38.-	2.44	1.02	41.8	3.0	3.4
Rwanda	2.41	0.05	2.1	5.11	0.23	4.5	6.92	0.36	5.2	4.9	4.2
Senegal	2.85	0.85	29.8	5.60	2.15	38.4	7.29	3.11	42.7	4.0	3.7
Chad	3.04	0.17	5.6	3.69	0.40	19.-	5.45	1.39	25.5	3.4	3.6
Togo	1.47	0.26	17.7	2.60	0.63	26.-	3.45	1.00	29.-	5.3	4.5
Zaire	16.65	3.46	21.-	28.15	7.46	26.-	36.60	12.55	34.-	-	3.9

Source: Arnaud, 1996.

1.2 - Links between urbanization and international supply and distribution circuits

It is evident that in terms of pace, dimension and the resulting structural changes, francophone African towns are making it increasingly necessary to resort to external food distribution and supply systems. Imported bread, rice and meat are rapidly becoming commonplace.

Resorting in this way to external supply and distribution circuits is increasing as multinationals play an increasingly important role in the agrifood chain and since international prices are not equilibrium prices. This is largely the result of over-supply due to protectionist measures and subsidies practiced in the industrial countries. There has also been a long-term slump in world prices. Between 1980 and 1993 real prices fell by 51% for cereals, 68% for beverages and 53% for other food products. This, coupled with what has often been an upward exchange rate has encouraged recourse to food imports.

There is a certain parallel between food deficits and the urban explosion in sub-Saharan Africa. In 1982/1984 cereals imports made it possible to feed 50% of the urban African population. However this percentage fell to 32-33% in 1985/86 and 1990.

Table 4 - Comparison of food deficits and urban expansion of the developing countries and in SSA

	Cereals Balance (Millions of tonnes)		Population fed (millions)		Urban population (millions)		3/5	4/6
	(1) DCS	(2) SSA	(3) DCS	(4) SSA	(5) DCS	(6) SSA
1934-38	-108.7	-	-	-	101	9	-	-
1948-52		-0.3	30	1.5	189	15	16%	10%
1958-62		-1.25	71	5.9	294	38	37%	20%
1968-72		-3.15	115	15.0	457	53	37%	30%
1973-77		-	190	-	-	-	-	-
1978-82		-8.80	709	42.0	706	102	44%	40%
1982-84		-15.0	-	71.5	-	115	-	50%
1985-86		-8.5	-	40.4	-	127	-	32%
1990		-10.5	1540	50	1823	150	29%	33%

(1) Excluding Argentina and China. (3)(4) According to the standard 210 kg/person.
Sources: Bairoch, for the developing countries (low and intermediate income economies). US Department of Agriculture for the SSA and the World Bank.

At the same time, the facilities and constraints created by opening up the dependent economies have often hastened urbanization: the blockages created by resorting to external fresh food supplies, the possibility of establishing “food security” in the towns with (free or purchased) contributions from abroad, the temptation to solve urban social conflicts by importing food and the difficulties found with marketing local products in the towns are all factors through which opening-up to external sources of supply reacts to the pace of urbanization and the forms it takes.

Figure 1 - The urban population and the cereals imports in Sub-Saharan Africa (1950-1990)

Many authors, such as Lipton (1977), have stressed an “urban bias” which has led to over-evaluating the exchange rate, before the adjustment measures, and to under-valuing agricultural products in order to boost the urban purchasing power. Other authors have emphasized the income which the African authorities make from cereals imports.

1.3 - The rejection of false evidence regarding external urban supply circuits

Statistical links and causal relations between urbanization and agrifood import circuits are by no means as obvious as the previous analyses seem to suggest. The complexity of the channels, the reversibility of the trends, the huge socio-economic disparities between African countries, and differences in the urbanization processes all make the magnitude of these external circuits relative.

This makes it necessary to set up typology groups. They lead to differentiating between the medium and long-term trends and factors of instability. They make it essential to reincorporate essential factors that have been ignored such as the instabilities in the agrifood systems and the mobility of the population, leading the agents involved to:

• adopt strategies to counter uncertainties;
• adopt different methods for regulation and to reverse the processed;
• cut off and segment the area, leading to a heterogeneity of agrifood supply circuits;
• informal economies operating outside the range of statistical data capture;
• effecting transfers linked to the rights and obligations of community membership which gets round the town/village rift (Mahieu, 1991);
• the role of economic policies.

In the long term there is no evidence of any significant increase in African imports of food compared with other types of imports. The ‘70s is an exception in this regard.

Trends differ from one major region to another. The cereal-growing savanna-covered Sahelian regions have deficits which are not found in the forested equatorial African countries which grow root crops and tubers. Oil-revenue countries saw a sharp increase in imports throughout the ‘70s, contrary to the experience of the agricultural product-exporting countries. There are considerable sub-regional and local differences, too.

Once the false evidence is removed, one finds that the least urbanized countries in Africa are the ones that are comparatively the most food-dependent in the form of imports or food aid, and those which have the lowest value-added by agriculture per rural inhabitant.

African urbanization, which takes a wide range of different forms, has not generally led to a break in links between the towns and the countryside for a variety of reasons. The agents belong to networks which overcome the town versus the countryside dichotomy.

The large urban areas are places in which trade is intensified. The town is not only a place where agricultural surpluses are to be tapped, but above all a place for wealth creation through the division of labour and the market, and for setting up a demand that can be met. In the towns there is the diversity and lack of uniformity of food regimes with overlapping confrontation and dynamics of their own. The examples that are always given, namely bread by rice, to characterize the “copycat models” must therefore always be taken in due proportion. There is a comparative ruralization of food consumption patterns (Requiers-Desjardins, 1989).

When one compares the demographic indicators of urbanization and food imports one can see the diversity between the results and the low level of correlation between these indicators.

The urban and rural African cereals circuit can be represented on the basis of Figure 2.[4]

Figure 2 - The cereal macroeconomic circuit in Sub-Saharan Africa (magnitudes in terms of millions of tonnes for 1990)

Table 5 - Urbanization rates and food imports into Africa (1983/1992)

	Urbanization rate		% food import/total imports
	1983 (%)	1992 (%)	1983	1992 (%)
Benin	33.4	39.60	13.5	12.32
Botswana	18.3	27.30	14.3	9.76
Burundi	4.7	5.60	12.07	9.05
Cameroon	34.5	42.10	48.1	15.20
Cape Verde	24.9	30.00	28.7	25.83
Comoros	24.5	29.00	31.6	27.34
Congo	37.1	41.70	9.6	19.09
Côte d’Ivoire	37.7	41.70	18.8	15.33
Ethiopia	11.2	12.70	13.25	36.94
Gabon	38.9	47.40	12.7	12.37
Gambia	19.6	23.80	23.2	34.01
Ghana	31.2	34.90	16.7	14.70
Guinea Bissau	17.6	20.80	11.55	35.01
Equatorial Guinea	27.6	29.40	14.9	17.30
Kenya	18.3	25.20	8.3	12.52
Lesotho	15.5	20.90	16.9	12.12
Liberia	38.1	47.50	21.5	59.93
Madagascar	19.9	25.10	21.9	15.17
Malawi	9.9	12.50	4.93	17.04
Mali	17.7	25.10	15.3	12.25
Mauritius	42.0	40.60	20.9	11.22
Mauritania	34.5	49.60	51.2	32.44
Mozambique	16.9	29.80	18.15	27.08
Niger	15.0	20.90	11.5	19.89
Nigeria	29.4	36.80	19.5	9.74
Uganda	9.1	11.70	5.52	4.75
Central African Rep.	40.8	48.30	27.7	18.19
Sao Tome & Principe	33.9	44.10	46.9	18.53
Sierra Leone	26.8	33.80	29.9	56.69
Sudan	20.2	23.30	14.7	18.19
Senegal	35.8	40.80	20.1	28.08
Seychelles	48.2	61.60	14.0	16.52

Figure 3 - Comparative urbanization rates and food import rates (1983/1992)

Source: FAO, World Tables

Figure 4 - Linkage between urbanization rates and value-added by agriculture per person (in dollars) (1980) and value-added per person by the primary and secondary sectors (1980)

As a first approximation, the ratio between the urban and rural population is an indicator of agricultural production for marketing by the rural population to supply the towns. Obviously in a more finely-tuned analysis it is necessary to draw a distinction between the agricultural population and the rural population, to incorporate the agricultural component of the urban populations and take account of their different rates of activity. The urban/rural population ratio rose from 0.08 in 1930 to 0.18 in 1950, 0.19 in 1960, 0.28 in 1970 and to 0.41 in 1980; the growth rate of this ratio may be put at 3.2% for 1950 and 3.7% in 1980 (Cour/ILTA).

This ratio can be related to the value-added by agriculture (Ya) and the non-agricultural sectors (Yna). If one takes the strong assumption according to which these levels of productivity reflect the differences between the towns and the countryside, the “rural” and “urban” levels of productivity can be compared both in time and in space (see Figure 2).

For the whole of Sub-Saharan Africa, “productivity” in 1980 was about six times higher for non-agricultural activities than for agricultural activities (771 dollars per person compared with 131 dollars). Conversely, over time this gap narrowed: it fell by 20% between 1970 and 1980 (6.9 to 5.9). One may reckon that in Francophone Africa the value-added by rural inhabitant is around 340 to 400 dollars, compared with a value-added by urban inhabitant is around 1,100 to 1,990 dollars. The gaps narrowed sharply between 1980 and 19990 but they are still around one to three. It may be estimated that the agricultural surplus marketed in the towns by African peasants has more or less increased at the same rate as the ratio between the non-agricultural to the agricultural population (WALTPS, 1994). The gaps vary widely between different Francophone African countries; the productivity ratios vary from between 1.98 in Benin to 26.5 in Gabon. Apart from the oil producing countries (Congo and Gabon) the gap has narrowed, which suggests that the per capita urban income has been falling in comparative terms.

The ‘80s were therefore characterized by a sharp narrowing of the gaps between the value-added in towns and in the countryside due to the implementation of adjustment policies (Cour, 1985; WALTPS, 1994). The wages of unskilled workers in the modern sector have fallen more steeply than the average wages. The so-called informal sector nevertheless plays a part of which little is known in cushioning and maintaining urban incomes comparatively steady.

1.4 - Typologies

1.4.1 - The effects of membership of the Franc Zone and the devaluation of the CFAF
1.4.2 - The four effects of urbanization on external supply circuits

The international supply circuits vary widely from one country to another. It is the coastal Sahelian countries which are most dependent in West Africa on external supply circuits (Cape Verde, Senegal, Mauritania, the Gambia). Using the “normal” values for agrifood imports (in terms of population, per capita GDP and urbanization level) West Africa draws comparatively less on external food supplies than the standard figures suggest.

A spatial analysis creates a distinction between the “parasite towns” that live off the rural world, with the “towns of independent means” playing a role or otherwise of redistribution vis à vis the rural world, the “cut-off towns” incorporated into the international area without any links with the rural world, and the “locomotive towns”, as poles of development vis à vis their hinterland. Depending upon their morphology, the magnitude of their networks and the different modes in which the space is structured, towns have different effects on supply channels.

The division of Africa into eight major regions[5] Nigeria, depending on a variety of geographic, economic, agronomic, demographic and food-related factors, shows a very uneven picture (Hugon, Coussy & Sudie, 1991).

In six regions out of eight, the increased urbanization rate was matched by a reduction in food import rates.

Table 6 - Food imports and urbanization rates in terms of the main regions of SSA

	1983	1983	1990	1990	1980	1990
	Imports	Import rates	Imports	Import Rates	Urbanization rates
	$	%	$	%	%	%
East Africa	2.30	8.29	1.70	6.51	9.50	13.50
SADCC	4.00	21.06	4.90	11.18	18.30	25.00
East Sahel	5.80	15.74	4.30	18.23	19.20	25.50
CILSS	8.30	11.34	7.70	9.91	22.20	32.00
Indian Ocean	2.20	11.41	2.20	15.35	23.30	31.70
Coastal West Africa	8.90	20.64	8.40	19.93	35.10	45.60
Central Africa	5.20	18.73	6.90	17.48	37.50	50.10
Nigeria	12.30	19.53	3.40	7.51	46.00	55.00

Value: billions of dollars; import rates % of food imports/total imports
Source: Makaya, 1996.

Source: FAO ILTA, Makaya

1.4.1 - The effects of membership of the Franc Zone and the devaluation of the CFAF

There is a substantial Franc Zone bias at the level of food imports. One may consider that the lack of any currency constraints, the convertibility of the CFAF and the fact that it was over-valued encourage food imports. The Franc Zone countries consumed on average 20 dollars per person of imported products during the ‘80s.

The countries with sustainable adjustment paradoxically imported more than the countries with non-sustainable adjustment or the non-reformers. Conversely, outside the Franc Zone, the adjustment effects do not seem to have cut imports.

Table 7 - Average imported food consumption per inhabitant (in US dollars, 1981) during the ‘80s

Country	1980’s average
Non-franc zone non-reforming (devaluation < 20%)	1,90
Non-franc zone non-sustainable adjustment (devaluation > 20%)	6,70
Non franc zone non-sustainable adjustment	7,05
Zone Franc non-reforming	7,50
Non franc zone sustainable adjustment (devaluation > 20%)	8,95
Non franc zone non-reforming	9,64
Non franc zone sustainable adjustment (devaluation < 20%)	11,09
Non franc zone non-reforming	12,97
Non franc zone non-sustainable adjustment	18,72
Non franc zone sustainable adjustment	25,58
Non franc zone non-sustainable adjustment (devaluation < 20%)	53,25

Source: Makaya, 1996.

1.4.2 - The four effects of urbanization on external supply circuits

Relations between urbanization and the dynamic of agrifood circuits are modelled in terms of macro-economic variables or economic policies. We can draw a distinction between four effects: wealth, income, security and poverty. The wealth-related effects cause the towns to drag the agrifood chain, while the income effect, otherwise known as Dutch Disease, has the reverse effect. The security-related effects lead to increased food imports, when production falls, while the poverty effect leads to the reverse result (Hugon, Coussy & Sudrie, 1991).

Table 8 - The different effects of urbanization on the agrifood circuits

Wealth effects (Côte d’Ivoire)

Reduced labour force drain by urbanization below increased matchable demand. Increased food imports lower than agricultural production increase.

Income effects (Congo, Gabon)

Income effects hasten rural exodus: labour force drain effects exceed rise in meetable demand. Combined effects of increased food imports and decline in agricultural production over the SSR.

Security effects (Sahelian countries)

Falling incomes and global imports and agricultural production. Pressure of urban demand creating increased food imports. Combined effects of reduced agricultural output and imports on the SSR.

Poverty effects (Madagascar)

Combined effects of lower production, lower agricultural output, lower imports and lower food imports. Economic adjustment by reducing the nutrition regime.

Notes:

Ya = agricultural output
Y = GDP
P_u = urban population
M = imports
M_a = food imports
SSR = food self-sufficiency rate or production below domestic use (final and intermediate demand)

[3] Food products are the ones listed in Section 0.1 and 4 and Division 22 of the CTCI (food products and live animals, beverages, tobacco and oil seeds).
[4] This figure is based upon strong assumptions. They provide magnitudes. The widely differing figures are cited by Azoulay and Dillon (1993) for 1980.
[5] We draw a distinction between the following:

4 non-Sahelian countries in West Africa (Burundi, Kenya, Uganda, Rwanda);

4 Sahelian countries in East Africa (Djibouti, Ethiopia, Somalia, Sudan);

4 Indian Ocean islands (Comoros, Madagascar, Mauritius, Seychelles);

9 Southern African countries formerly belonging to SADCC;

6 CEMAC countries

8 CILSS countries the coastal states of West Africa;