^{page 513 note 1} Koehn, Peter, ‘Ethiopia: famine, food production, and changes in the legal order’, in African Studies Review (Waltham), XXII, I, 1979, pp. 51–71.Google Scholar

^{page 513 note 2} Cohen, John M., ‘Analysing the Ethiopian Revolution: a cautionary tale’, in The Journal of Modern African Studies (Cambridge), XVIII, 4, 12 1980, pp. 690–91.Google Scholar

^{page 513 note 3} Cohen never specifies how and when he secured access to these studies, although he admits, ibid. p. 688, that in 1977 he jointly argued ‘that production in 1975–6 was up 10 per cent, without any knowledge of the crop-production studies cited here’; Cohen, John M. and Koehn, Peter H., ‘Rural and Urban Land Reform in Ethiopia’, in African Law Studies (Los Angeles), XIV, I, 1977, p. 24.Google Scholar

Some of the restrictions placed on the circulation and citation of official reports in Nigeria are discussed in Wallace, Tina, ‘“Agricultural Bonanza?” Some Critical Issues Raised by the World Bank Agricultural Development Projects in Nigeria, in Nigerian, Journal of Public Affairs (Zaria), IX, 05 1980, pp. 67–8.Google Scholar

^{page 514 note 1} Cohen, loc.cit. p. 689 fn. 2, 3. In an earlier draft of this note, Cohen more forthrightly stated that the data gathered in these studies, and subsequently presented in Table I on p. 689, ‘must be acknowledged to be risky’. He further admits on p. 699 that ‘we can only guess how much political pressure or censorship was exerted to shape the annual figures, for it was probably in the interest of the P.M.A.C. to claim low food-grain production in order to qualify for donor assistance.’

Holmberg, Johan, Grain Marketing and Land Reform in Ethiopia: an analysis of the marketing and pricing of food grains in 1976 after the land reform (Uppsala, 1977), Scandinavian Institute of African Studies, Research Report No. 41, p. 33, indicates that the Derg has, indeed, manipulated official statistics in order to inflate Ethiopia's food-grain deficit. He also notes on p. 30 that ‘Data on the production and consumption of cereals, oilcrops and pulses are difficult to obtain and subject to wide error margins.’ Moreover, in 1976,Google ScholarStahl, Michael found that ‘peasants deliberately underestimated both [crop] size and yield because of their deep-rooted suspicion of state authorities’; New Seeds in Old Soil: a study of the land reform process in Western Wollega, Ethiopia 1975–1976 (Uppsala, 1977), Scandinavian Institute of African Studies, Research Report No. 40, p. 12.Google Scholar

^{page 514 note 2} Cohen, loc.cit. p. 686.

^{page 514 note 3} Ibid. pp. 686–7. Cohen nevertheless proceeds on p. 690 to make 10 major points regarding post-reform agricultural production patterns based upon ‘tentative analysis’ of the data from these census surveys.

^{page 514 note 4} D'Silva, Brian and Rwang, Pam, ‘Planting Strategies of Semi-Subsistence Farmers in Pan Hauya Village, Zaria, 1979–1980’, Department of Agricultural Economics, Institute for Agricultural Research, Ahmadu Bello University, Zaria, September 1980.Google Scholar

^{page 515 note 1} See, for instance, Norman, D. W., ‘Economic Survey of Three Villages in Zaria Province: 2. Input-Output Studies, Volume 1, Text’, Institute for Agricultural Research, Ahmadu Bello University, Zaria, 1972, Samaru Miscellaneous Paper No. 37.Google Scholar

^{page 515 note 2} Moreover, while there are certain advantages in combining yield surveys with other on-going farm surveys, the volume of data generated may prove to be immense and the attendant strain on enumerators, coders, etcetera, is likely to have an adverse effect on the quality of the results. See the related points made by Bondestam, Lars, Some Notes on African Statistics — Collection, Reliablility and Interpretation (Uppsala, 1973), Scandinavian Institute of African Studies, Research Report No. 18, p. 5.Google Scholar

^{page 515 note 3} Slade, Roger H., ‘The Monitoring and Evaluation of Funtua, Gusau, and Gombe Agricultural Development Projects’, Agricultural Project Monitoring, Evaluation, and Planning Unit, Federal Department of Rural Development, Ministry of Agriculture, Kaduna, 1980.Google Scholar

^{page 515 note 4} The complexity of African farming systems suggests that individuals attempting to generalise about conditions such as aggregate levels of agricultural production must first attempt to understand the nature of those systems. For example, it is only through detailed study of rural household decisions that one is able to discover the reasons why farmer a plants one-third of his land in sorghum, while farmer B plants twice as much in the same crop. In short, the local decision-making process provides important insights into patterns of agricultural production.

^{page 516 note 1} This recommendation is consistent with Dahlberg's, Kenneth A. call for alternative, intermediate data systems and categories that are neither too sophisticated nor dependent on western interests and dominant local élites; Beyond the Green Revolution: the ecology and politics of global agricultural development (New York, 1979), pp. 172–4.CrossRefGoogle Scholar

^{page 516 note 2} To his credit, Holmberg – the analyst most familiar with the crop-production survey techniques of the Ministry of Agriculture – has characterised both Ministry and E.P.I.D. estimates of the first post-reform Ethiopian harvest as ‘little more than conjectures’; op.cit. p. 19. See also Bondestam, op.cit. pp. 36 and 39; Lars Bondestam, ‘Reliability of Population Data in Ethiopia’, Central Statistical Office, Ababa, Addis, 1972, pp. 2 and 17;Google Scholar and Marina, and Ottaway, David, Ethiopia: Empire in Revolution (New York, 1978), p. 201, fn. 34.Google Scholar

^{page 516 note 3} Koehn, loc.cit. p. 61.Google Scholar

^{page 516 note 4} Cohen, loc.cit. p. 687.

^{page 516 note 5} Gilkes, Patrick, ‘Ethiopia: more decentralization as land reform progresses’, in African Development (London), X, 7, 1976, pp. 664–5, cites the Permanent Secretary, Ministry of Land Reform and Settlement.Google ScholarHolmberg, John, ‘Pricing Strategies form Agricultural Produce in a Changing Society: rural/urban contradictions in Ethiopia’, 19th Annual Meeting of the African Studies Association, Boston, 1976, pp. 11 and 18, notes that ‘the Ministry of Agriculture…estimated the 1975[-6] production by smallholders to be 20% higher than the previous year, while EPID's estimates indicated in increase by 10%.’ Holmberg worked for E.P.I.D. at the time. Robert Liebenthal, ‘Certain Development issus in ethiopia and their Relationship to Rural/Urban Balance: a perspective based on World Bank experience’,Google Scholaribid. pp. 6 and 9, appears to rely on Holmberg's E.P.I.D. fugures. The Ottaways, op.cit. p. 201, later observed (August 1977) that ‘The only report available on the 1975/76 crop estimated an increase of roughly 10 percent over the previous year when the harvest had been good’. The document cited is John Dalton, ‘Report on prospects of the Ethiopian Harvest of 1968 E.C. (1975/76 G.C.)’, Prepared for the Relief and Rehabilitation Commission of the Government of Ethiopia and the United Nations Development Programme, Addis Ababa, 23 December 1975.

^{page 516 note 6} On the basis of one of the most systematic field studies undertaken in the post-reform period, for instance, Ståhl, op.cit. p. 48, discovered that ‘the 1975/76…harvest as a whole in Western Wollega showed a lower yield per hectare compared to previous years…mainly due to the bad quality of chemical fertilizers distributed through EPID in 1975’, although peasants ‘put a larger area than ever before under cultivation.’ By way of contrast, the official crop-sampling data promoted by Cohen, loc.cit. p. 687, fn. 6, ‘show the reverse, namely that yields increased while crop hectarage decreased’.

^{page 517 note 1} Cohen, loc.cit. pp. 688–9.

^{page 517 note 2} Cohen, John M. and Weintraub, Dov, Land and Peasants in Imperial Ethiopia: the social background to a revolution (Assen, 1975), pp. 86–7.Google Scholar In his 1980 note, loc.cit. p. 686, Cohen still maintained that ‘the dislocations of major land reform, even without a revolution, almost invariably lead in the short-run to substantial declines in production’. In 1975 and 1976, according to the Ottaways, op.cit. pp. 72–3 and 77, embittered former landlords did organise armed bands that harassed peasant cultivators, and military activity disrupted farming in much of Eritrea. However, these serious dislocations proved insufficient by themselves to cause a major decline in overall food production.

In sharp contrast to Cohen's, position, Eckholm, Erik contends in The Dispossessed of the Earth:land reform and sustainable development (Washington, D.C., 1979), Worldwatch Paper No. 30, p. 21, that ‘taken as a whole, the record supports the notion that land reforms can unleash higher agricultural output’.Google Scholar

^{page 517 note 3} Holmberg, Grain Marketing, p. 19. See also the Ottaways, op.cit. p. 76. Food production constitutes only one of several vital issues raised in relation to the rural policies of the Derg. The debate over harvest estimates should not be allowed to obscure the fact that the radical land and institutional measures introduced by the military Government have dramatically altered the rural power structure that prevailed under the imperial régime. In the immediate post-reform period, this resulted both in significant peasant empowerment and in improved nutrition for the many rural families who were able to increase their consumption of food. Holmberg, op.cit. p. 20; Gilkes, loc.cit. p. 664; Stahl, op.cit. p. 70; Cohen and Koehn, loc.cit. pp. 20–2; and Ottaway, Marina, ‘Land Reform and Peasant Associations in Ethiopia: a preliminary analysis’, in Rural Africana (East Lansing), XXVIII Fall, 1975, p. 54.Google Scholar

^{page 518 note 1} Cohen, loc.cit. pp. 686, 688, and 690. In 1976,Google Scholar he argued in Cohen, John M., Goldsmith, Arthur A., and Mellor, John W., ‘Rural Development Issues Following Ethiopian Land Reform’, in Africa Today (Denver), XXIII, 2, 04–06 1976, pp. 20–1, that ‘Balancing these complex, man-made crosscurrents, the net of their effect seems uncertain and quite possibly of less importance than the weather variations frequently encountered. Thus, it may be that weather and not the temporary disruptions caused by the government's land policy, will remain the most important determinant of crop size in the immediate post-land reform period. As of late 1975, just prior to the harvest, it appears that good weather had indeed facilitated production of a bumper crop.’Google Scholar

^{page 518 note 2} See Koehn, loc.cit. pp. 61–3; also Ståhl, op.cit. p. 48.Google Scholar

^{page 518 note 3} Specifically, the north and north-east of Ethiopia experienced inadequate rainfall; the Ogaden, southern Bale and Sidamo, and parts of Gemu Gofa continued to be afflicted by drought; and prime crop-producing areas in Shoa, Arussi, Hararghe, and northern Bale received ‘below and much below normal rainfall’ during the month of August. Holmberg, Grain Marketing, p. 19; Gilkes, loc.cit. p. 664;Google ScholarShepherd, Jack, The Politics of Starvation (New York, 1975), pp. 75 and 79; the Ottaways, op.cit. p. 77; Ethiopia, Relief and Rehabilitation Commission Report, ‘August 1–20 Rainfall Review’, Addis Ababa, 1975;Google ScholarEthiopia, Consolidated Food and Nutrition Information Service Report, ‘Current Situation in Drought-Affected Areas’, Addis Ababa, 1975; Ethiopia, Technical Information Service Report, ‘The Consequences of Rain Failure in the Ogaden and Lowland Bale’, Addis Ababa, December 1971, p. 1; and Ethiopia, Food and Nutrition Surveillance System Secretariat, ‘The Belg Rains: a survey of their effect on food production’, Addis Ababa, 1975.Google Scholar

^{page 518 note 4} Cohen, loc.cit. p. 688, our emphasis. The full citation, found only in his 1976 joint article,Google Scholar is Torgerson, Dial, ‘Rebels, Marketing Problems Threaten Ethiopia's Bumper Crop’, in Los Angeles Times, 28 10 1975, p. 22.Google Scholar

^{page 518 note 5} Cohen, loc.cit. p. 689.Google Scholar

^{page 519 note 1} Ibid. pp. 688 and 690.

^{page 519 note 2} See Koehn, Peter, ‘African Approaches to Environmental Stress: a focus on Ethiopia and Nigeria’, in Barrett, Richard N. (ed), International Dimensions of the Environmental Crisis (Boulder, 1982), pp. 258–60;Google Scholar and Markakis, John, ‘The Military State and Ethiopia's Path to “Socialism’, in Review of African Political Economy (Sheffield), 21, 05–09 1981, pp. 22–4.Google Scholar