Also, the majority of DAC donors have been increasing their ODA share allocated to social and administrative infrastructure (on the average from 20.1% in 1975/76 to 23.5% in 1989/90). However, the largest bilateral donors, such as Japan and the USA, are among the donors which have provided the lowest shares to the social sector (respectively 18.9 and 15% in 1989-90). Comparable figures for multilateral agencies show shares as high as 39% for the United Nations agencies (including UNDP, UNICEF, UNWRA, WFP, UNHCR, UNFPA, UNFPA) and 32.1% for IDA.
A more refined but less encouraging picture is given in UNDP's Human Development Reports which present four types of ratios, namely (a) aid/human expenditure ratios (the percentage of a donor's GNP going to human priority areas in recipient countries or the amount of official development assistance received for human priority areas expressed as a percentage of the recipient country's GNP); (b) aid/social allocation ratio (the percentage of official development assistance that goes to the social sector); (c) aid/social priority ratio (the percentage of social sector official development assistance that goes to human priority areas); and (d) social priority aid, as percentage of total aid.
Human priority areas relate to basic education, primary health care, safe drinking water, family planning and nutrition programmes. These ratios are calculated for individual and major groups of recipient countries, as well as for individual donors. Australia, Germany and Japan are the donors with the lowest ratios in 1988/89.
UNDP proposes that "at least 20% of total aid should be allocated to human priority concerns" (of primary benefit to the poor) "three times the present 6.5%".
One reason for which bilateral donors may be reluctant to provide direct assistance for primary social concerns, such as primary health care and basic education, is that these are areas where recipient governments and local administrations are seen as bearing the primary responsibility. For donors to support social programmes and projects implies relatively less capital-intensive and administrative-intensive assistance (as compared to other types of programmes) but relatively more local and recurrent cost financing. Many donors traditionally prefer to finance the foreign exchange of investment costs and are reluctant to finance recurrent costs.
The World Bank, for its part, has stressed the difficulty of assessing the effect of aid projects on the poor, and has noted that sometimes assistance for specific anti-poverty projects indirectly finances undertakings that may not help the poor and may even hurt them. The DAC Creditor Reporting System (CRS) which is the most detailed in this respect provides only an incomplete record of flows for calculating sectoral aid allocations. Development Cooperation reports of UNDP and information from aid agencies are difficult to use (because of problems of comparability, classification, timeliness, [etc]). Information from NGOs is very scanty.