overty is, increasingly, perceived as an   urban phenomenon. With one or two exceptions, such as Rwanda, starving millions in the countryside have not been seen on the world's television screens on any scale since the 1984-85 famine in the Horn of Africa. Faced with the desperate challenges that arise from urbanization in the developing world, the AIDS pandemic, pollution of the seas, greenhouse gases and declining air quality, many in the donor community no longer regard agricultural development as the pressing concern it was 25 or even 15 years ago. Agriculture, it seems, is no longer an area of pressing concern for development. At ICARDA, we believe that this could be counterproductive. Recent research at ICARDA has established, first, that there is persistent, grinding rural poverty in the world's dry areas; and second, that the link between this rural poverty, and the fragility of the environment in these areas, is an explosive one.
        Where are the dry areas, and how poor are their people?
For the purposes of this study, 41 countries and three territories were selected on the basis of plant growing seasons limited to 180 days. These were, for the most part, whole countries, but not exclusively: India, for example, figured in the work because of three of its States that fall within the definition of dry areas. Of the countries included, 34 were in the West Asia and North Africa (WANA) region.
       It was found that, out of a total population of one billion, 696 million had a per-capita GDP of below US$2 a day. About 300 million people lived below nationally-determined poverty lines, which are normally linked to nutrition. The poverty lines are not a comprehensive definition, as they do not take into account material deprivation, isolation, lack of influence, poor food and water security, or freedom of choice. Agriculture has a key role here.
       The  link between agriculture and pov-

rich states is just 30 million out of 627 million for WANA as a whole, and in any case, their per-capita GDP is US$8100 (oil states plus Cyprus). This is very high by the standards of the region, but not really that high by northern standards. Far more typical are Lebanon or Tunisia with just under US$1500. Ethiopia and Eritrea have US$52; Somalia, just US$36. It is these poorer countries, with their larger populations, that have higher percentages of their people living in poverty; in Sudan, for instance, 71% of its people, and 93% of its rural population, were reported by UNDP to be living below the poverty line in 1992.
      As part of this research, it was thought necessary to properly quantify rural poverty in the countries concerned. This was done by devising a rural poverty indicator (RPI). This gave us a cross-country poverty line so as to eliminate inconsistencies caused by different costs of living. Briefly stated, a welfare indicator based on per-capita income (using per-capita GDP as an estimate) was used to produce a poverty indicator which was then related to the percentage of rural poor. It was more complicated than this; allowances had to be made for differences in distribution of income and purchasing power.
       The result showed a close correlation with the UNDP Human Development Index and with indicators for infant malnutrition. For example, Ethiopia topped the scale with an RPI of 0.832; examples in the dry areas included Turkmenistan at 0.216, and Lebanon at 0.040. UNDP's 1994 estimate of malnourished children under 5 in these countries in 1994 was 38.2%, 10.6%, and 8.4%, respectively. This seems to indicate that rural poverty has a direct correlation with the nutritional standard of  nations included in the study.
       Grinding poverty is an offence against humanity. Alleviation of rural poverty should need no further justification.

erty is masked by the relatively low share of GDP accounted for by the agricultural sector. It varies between 27% in the lower-income countries outside the WANA region, to just 3% in the higher-income countries within it. However, this can lead to a dangerous misunderstanding: in fact, of those 300 million poor, 72% are rural and are dependent on agriculture for their income.
       The relatively low GDP percentage arises

partly because much production is consumed on-farm--especially in the poorer countries, where the rural population may be heavily dependent on subsistence crops like food barley and pulses. This keeps some production out of the GDP figures. However, in the poorest countries studied--Ethiopia, Eritrea, Somalia and Afghanistan--agriculture does account for about 60- 80% of GDP.
       At this point, it is, perhaps, worth exploding another myth--that the presence of oil-rich states in WANA results in a relatively high per-capita income, and that the per capita income of the region cannot really be bad. This is not the case. The population of the oil-