Seasonal migration is usually necessary in order for people to be able to take advantage of complementarities in the timing of production and income-earning opportunities at the macro-environmental level. Two forms of this have already been discussed: hunting-gathering and nomadic pastoralism. The present chapter will concentrate on a third pattern, which, in terms of numbers of people involved, is of very much greater importance: the seasonal migration of agricultural labour.

Migration theory has occupied a fairly central position in the development debate since at least the early 1950s, and, in the course of their evolution, views on the subject have tended to polarize around two positions. Some writers view it in a positive light, regarding it as a rational response to changing economic opportunity, part of a continually evolving income maximization-cum-risk minimization strategy. Early writers on the subject (especially the ‘zero marginal product’ school of thought mentioned in Chapter I) presented what is now recognized as a simplistic version of this view: surplus labour from the rural areas would migrate townwards in search of greater economic opportunity, thus simultaneously providing a source of labour for industrialization and raising labour productivity in agriculture (see especially Lewis 1954, and Fei and Ranis 1964). These theoretical views were subsequently modified in the light of empirical evidence, later theories taking account of the risk of not finding employment in town (e.g. Harris and Todaro 1970).