^{page 312 note 1} Migration by families is likely to have been more common in the Federal period, 1953–63, when restrictions on the movement of citizens of Northern and Southern Rhodesia and Nyasaland were relaxed.

^{page 313 note 1} Nyasaland Government, The Report of the Committee on Emigrant Labour (Zomba, 1963).Google Scholar

^{page 313 note 2} According to Boeder, Robert B., ‘Malawians Abroad: the history of labour emigration from Malawi to its neighbours, 1899 to the present day’, Ph.D. dissertation, Michigan State University, East Lansing, 1974, 487,000 Malawians were out of the country in 1972. Between then and the 1977 census, approximately 212,000 returned to Malawi; this left some 275,000 outside, namely 5 per cent of the de facto population.Google Scholar

^{page 314 note 1} For a more detailed analysis, see Kydd, Jonathan G. and Christiansen, Robert E., ‘Structural Change in Malawi since Independence: consequences of a development strategy based on large scale agriculture’, in World Development (Oxford), 10, 5, 1982, pp. 355–75.Google Scholar

^{page 314 note 2} Since these estimates use only the officially marketed production by peasant farmers, it understates their economic importance. Since independence the geographical coverage of, and therefore the purchases by, the Agricultural Development and Marketing Corporation (ADMARC) has increased significantly. This makes the initial dominance of peasant production that much greater, and the growth of estates even more impressive.

^{page 314 note 3} Thus, for example, the reliability of the estimated trend during 1970–81 for the value of estate tobacco growth is very high, while that for the changes in peasant tobacco production is relatively weak. It should be made clear that the growth rate for crops with low r² values is very sensitive to change by the addition of new values. In such cases it is better to treat these growth rates as not significantly different from zero, i.e. as displaying no trend.

^{page 315 note 3} Since the late 1960s, two notable changes have taken place in the official marketing of maize. First, an important component of the rural development programme has been the expansion of both the geographical coverage and density of the official marketing system. The second factor stems from the dependence of Malawi's urban population on food supplies from peasant holdings. The withdrawal of labour from this sector, coupled with land scarcity and modest technical progress, has hampered the ability of peasants to supply the greatly increased marketed surpluses required. Thus, in fulfilling its obligation to supply food to the urban and estate sectors, ADMARC has acted to raise the price of maize sharply relative to other peasant-grown cash crops. This has not only induced a substitution of resources in production, but has also caused some switch between marketing channels. Thus, over the 1970s a growing share of the food surplus of the peasant sector has been captured by the official marketing network.

^{page 316 note 1} We recognise that this is not a device which originated with post-colonial authorities. The practice of maintaining a differential between the prices paid and received by the marketing board was/is common throughout anglophone Africa.

^{page 316 note 2} See Kydd, and Christiansen, , ‘Structural Change in Malawi Since Independence’.Google Scholar

^{page 316 note 3} The lack of a trend in the real value of payments per capita to labour employed on own-holdings occurs over a time period in which real world prices for many crops were increasing. Evidence of this can be found by looking at a summary of ADMARC's profits on the crop trading accounts. These profits come exclusively from the peasant sector, since estates are permitted to deal directly with international buyers and need not sell through ADMARC. Over the seven-year period 1971–2 to 1977–8, profits (in constant 1980 million Kwacha) generally rose, as follows: (1) 17·67, (2) 12·04, (3) 13·62, (4) 15·43, (5) 16·58, (6) 34·74, and (7) 40·18, making an average annual profit of K21.46 million. By way of contrast, with the recent slump in world prices, the ADMARC surpluses fell in 1978–9 to 5.58 and in 1979–80 to 0.83 million Kwacha.

^{page 316 note 4} This includes the impact of returning international migrants. According to the final report of the 1977 census, the Government estimates that natural increase occurred at a rate of 2.6 percent.

^{page 319 note 1} Kydd, and Christiansen, , ‘Structural Change in Malawi Since Independence’.Google Scholar

^{page 319 note 2} Over the course of the 1970s, commercial bank advances to the estates increased from about 15 to over 50 per cent of total advances. The Government also pushed them reluctantly into the role of agricultural development banks- i.e. as sources of medium- to long-term capital for estate development.

^{page 319 note 3} According to Boeder, op.cit., 51.2 per cent of these migrants were in Rhodesia. South Africa is estimated to have had 179,000 (36.9 per cent) of those abroad; Zambia, 43,000 (8.8 per cent); Tanzania, 8,000 (1.6 per cent); and the remainder (1.4 per cent) are listed as being elsewhere. When compared to the figures cited on p. 313, we see that while Rhodesia remained the most important host country for Malawians abroad, it was less dominant than in 1936.

^{page 321 note 1} With the cessation of hostilities in Mozambique, some Malawian residents are known to have returned there. The precise size of this outflow is not known, but was not likely to have been large enough to have had a significant impact on net migration figures.

^{page 321 note 2} It has been estimated that the percentage of the Malawian population who were enumerated in 1966 and again in 1977 is 83.2 per cent; Robert E. Christiansen, ‘The Pattern and Causes of Internal Migration in Malawi, 1966–1977’, Malawi Social Science Conference, Zomba, July 1982. The annual rate of increase during the inter-censal period was 2.6 per cent, although this applies to the entire population rather than to any specific age group. (However, since we are concerned with such a large part of the population, this should not cause a serious distortion.) This implies an inter-censal growth of 32.6 per cent for the economically active population. Since the actual figures were 51.2 per cent for males and 45.8 per cent for females, we have estimated that the net in-migration is 175,000 males and 158,000 females. A word of caution should be added at this point. These estimates are somewhat sensitive to our assumption about the rate of natural increase for the economically active population. For instance, if we assume that the rate for women is 2.7 per cent, then the number of female in-migrants turns out to be 141,000. However, since we believe that it is unlikely that calculations for the rate of natural increase are in error by a factor of more than 0.1 (4 per cent of our estimated rate of natural increase), we suggest that our figure of 333,000 is subject to an error of no more than 5 per cent, or 17,000 individuals.

^{page 322 note 1} According to a South African labour union official; personal communication to one of the authors.

^{page 322 note 2} The position of Malawian migrants in South Africa, especially of those working outside mining, is a subject which the authors have not researched. Thus we are unable to say whether any changes in their employment or residence status influenced the homeward movement which gathered pace in the early 1970s. Further, since the mid-1960s, President Kamuzu Banda had been making speeches in Malawi discouraging international migration and, as noted in the text, in 1974 he took strong steps to inhibit further migration. It may be that some measures to discourage migration and to encourage expatriate Malawians to return were used prior to 1974.

^{page 325 note 1} This is based on an estimating procedure developed by Christiansen, op. cit.

^{page 325 note 2} See Government of Malawi, Economic Report, 1983 (Zomba, 1983).Google Scholar

^{page 325 note 3} In March 1983 the prices paid directly to peasant farmers were increased by 35–100 per cent, in nominal terms, as part of the package of policy reforms introduced under the I.B.R.D. ‘structural adjustment loan’.