^{page 271 note 1} Cameroun's banking regulations governing lending to small and medium-scale enterprises, for example, distinguish them from large-scale firms on the basis of any one of the following: (i) payment of the flat-rate taux forfaitaire rather than the regular company taxes; (ii) short term rediscount limit at the Central Bank below C.F.A.F. 25 million; or (iii) equity below C.F.A.F. 50 million.

^{page 272 note 1} The investment constitutes a barrier (in an economy with limited access to capital) which is likely to prevent surplus labour from entering in sufficient numbers to drive the marginal product of labour to nil.

^{page 272 note 2} The term ‘small-scale’ is often used to refer to the entire range of firms between the ‘modern’ and ‘informal’ sectors, including artisans. In Cameroun, however, the term ‘petites et moyennes entreprises’ is used to refer to firms that are larger and more organised than artisanal establishments, but not sufficiently large to be considered in the modern sector.

The term ‘small-scale’ is therefore used in this article to refer to firms that are slightly larger and more organised than artisans, involving some delegation of management responsibilities. ‘Medium-scale’ firms resemble modem-sector firms in most ways except size and coverage by regulations and taxes. The term ‘intermediate sector’ is used here to refer to artisanal, small and medium-scale firms as a whole, and the upper limit in Africa could reasonably be set in the neighbourhood of 30–50 workers.

^{page 272 note 3} For further details, see Steel, William F., Small-Scale Employment and Prodution in Developing Countries: evidence from Ghana (New York, 1977), pp. 19–22,Google Scholar and Page, John, ‘Small Enterprises in African Development: a survey’, World Bank Studies in Employment and Rural Development, Washington, D.C., forthcoming.Google Scholar

^{page 273 note 1} The pattern of large-scale industrialisation is more likely to vary across countries in response to differences in demand, national priorities, local resources, and incentives. It is therefore difficult to generalise about the ways in which industrialisation is likely to replace (or stimulate) intermediate sector production in particular industries.

^{page 273 note 2} Food preparation involves some degree of food processing, and could therefore be considered a manufacturing as well as a service activity.

^{page 273 note 3} Ghana data are for 1973 and Cameroun for 1976, where available, because these are the years of the surveys presented. Deflation of the Ghana figure to adjust for its overvalued exchange rate would bring it close to $300.

^{page 274 note 1} Sources: Ghana, Central Bureau of Statistics, Population Census, 1970, Vols. II, III, and IV, worksheets; United Republic of Cameroun,Google ScholarCensus Bureau, preliminary estimates; World Bank, World Tables, 1976 and 1978 (preliminary),Google Scholar and World Atlas (annual).

^{page 275 note 1} Sources: Yaoundé: preliminary results of a 1976 survey of the Service de l'Artisanat, Department of Commerce, Ministry of Economics and Planning, Government of the United Republic of Cameroun. Accra: 1973 survey by the author (for comparison to the small-scale sector, including firms with up to 29 full-time wage employees, see Steel, op. cit.).

^{page 275 note 2} The second-largest artisanal industry in Accra, however, is food preparation using the adjusted figures, and vehicle repair if firms with fewer than five full-time wage workers are included regardless of the number of other employees.

^{page 276 note 1} Its ranking falls, however, as larger firms are included; it ranks eighth in Accra using the unadjusted figures (Table 1, column 5), and eleventh among all manufacturing firms with fewer than 30 wage workers. Steel, op. cit. p. 192.

^{page 276 note 2} This is true even when industries that may not have been enumerated in Yaoundé are taken into account. Services, however, contribute a larger share of G.D.P. in Cameroun. It may be that the modern-service sector is more highly developed there, perhaps in connection with its larger expatriate sector.

^{page 277 note 1} Source: see Table 2. The 1976 population of Yaoundé was 315,000; with 28·8 per cent economically active in the age group 15–54, and with 98·3 per cent of its work force outside primary activities, the non-agricultural labour force was estimated to be 89,178 (based on preliminary Census statistics). The 1973 population of Accra, projected from 1970 on the basis of 1960–70 growth, was 752,000, and its non-agricultural labour force was estimated to be 275,900.

^{page 277 note 2} This concentration diminishes the greater the number of industries that are included. The top 12 industries account for 89% of overall employment in Yaoundé and 86% in Accra, while within manufacturing the figures are 93% and 92%, respectively.

^{page 278 note 1} Apprentices account for 25·0% of the total artisanal employment in Accra (adjusted); family members, 7·4%; and part-time non-family workers, 2·0%. Family members are important principally in food preparation. Apprentices account for 28·9%; family, 5·6%; and part-time labour, 1·8% of the total artisanal manufacturing employment.

^{page 278 note 2} The Centre National d'assistance aux petites et moyennes entreprises was designated as the co-ordinating Agency for assistance to the intermediate sector in 1974.

^{page 279 note 1} Although a guarantee fund has been established to encourage banks to lend to smaller, riskier firms, this too has a natural tendency to favour larger and more established firms that are least likely to result in a poor default record. The same problem affects lending by the Cameroun Development Bank, although it does charge a higher interest rate for small-scale loans in order to cover the cost of special assistance that is provided.

^{page 279 note 2} See Steel, op. cit. p. 182.

^{page 281 note 1} Source: see Table 2.

^{page 282 note 1} Payoff of government contracts with small firms in Cameroun reportedly take from six months to as much as two years after completion of the work. A similar fate awaits firms that try to claim rebates for tariffs and taxes paid on machinery and materials.