In China and South Asia, which are the areas that have contributed to the weakening of the war against poverty in the Asia-Pacific, rising internal inequality does seemed to have played a role. Though recent data on inequality is difficult to come by, the available evidence does suggest that the years of reform are worsening inequality in China. Before the start of reform in 1978, the ratio of urban to rural income had declined to 2.36 from 3.48 in 1978. Price and production reform in agriculture, ensured that this trend continued till 1985. However, after 1985, the ratio began to rise again and stood at 2.61 in 1994. Further, within rural and urban areas the gini coefficient of income distribution rose from 0.31 and 0.19 respectively to 0.41 and 0.37 between 1986 and 1994. This worsening of income distribution which appears to have continued since then has neutralised some of the significant benefits in terms of poverty reduction ensured by the rapid rates of growth in aggregate income China has managed to ensure since the beginning of the reform.
 
The adverse effects of rising inequality resulting from the process of structural adjustment and reform are visible in a country like Pakistan as well. The move to encourage exports during the 1980s, with the removal of export restrictions, devaluation of the currency and provision of a range of export concessions to the textile sector, did result in a rapid expansion in textile exports. As a result cotton textile manufactures such as yarns, fabrics, apparel and clothing accessories, which accounted for 35 per cent of Pakistan's exports in 1980, registered an increase in share to 75 per cent in 1998. The profitability and rapid growth of these exports rendered cotton cultivation an extremely lucrative activity, resulting in resumption and leasing in of land for self-cultivation by large farmers. Cotton output tripled between 1980 and 1992. Simultaneously, farm holdings of size greater than 50 acres, which accounted for 0.3 per cent of farms and 8.4 per cent of area in 1981, came to account for 8.4 per cent of farms and 23.8 per cent of area. The rural gini coefficient is estimated to have risen from 0.32 in 1978-79 to 0.41 in 1990/91. This increase in inequality would have adversely affected progress on the anti-poverty front. Further, this inequality would have made the consequences of the slowdown of growth in the 1990s particularly adverse for the poor.
 
The proximate role of inequality in worsening poverty or limiting gains in terms of poverty reduction has conventionally been situated in arguments that point to a complex combination of influences on poverty in predominantly rural communities. The early literature attempted to focus on two kinds of variables: those that capture movements in agricultural production or productivity and those that capture movements in prices that are seen as impacting directly or indirectly on real income and poverty. Poverty analyses tended to focus on the relative significance of each of these sets of variables as well as the best way in which their effects can be captured. Thus, as noted above, even though movements in agricultural production would obviously impact on poverty, the significance of that impact would depend on whether agricultural growth is accompanied by an increase in inequality or whether there is a simultaneous increase in sources of non-agricultural income in the rural areas.
 
As for prices, what needs to be looked at is not the nominal price level of food, but the relative price of food and the rate of increase of nominal food prices. A faster rate of increase of agricultural prices can have two contradictory effects on poverty. First, to the extent that agricultural growth is stimulated by a shift in terms of trade in favour of agriculture, and assuming that inequality does not increase, the rise in agricultural prices would contribute to a reduction in rural poverty to some extent. Second, since a large part of the rural population consists of agricultural labourers, small farmers and non-farm workers, who are all net purchasers of food, a sharp increase in the price of food would squeeze real incomes and worsen poverty.
 
This dual effect on poverty of higher agricultural prices has implications for those who emphasise the importance of “getting prices right”. These economists have traditionally argued that protection in developing countries, which in all cases has been exclusively provided to or concentrated on industrial goods, was indefensible on grounds of both “efficiency” and “equity”. In their view, protection for industry and none for agriculture has in the past tended to skew the industry-agriculture terms of trade in favour of industry. This is seen to have adversely affected incentives for investment in agriculture, resulting in slow agricultural growth that limited poverty reduction or even worsened rural poverty. Liberalisation, by reversing the skewed structure of incentives, it is argued, would spur agricultural growth and reduce poverty.
 
What this argument, given its exclusive emphasis on growth, ignores, is the adverse effects on poverty that higher food prices can have. So long as the prices of food do not rise faster than rural money wages and the prices of non-food agricultural products, the growth effect of higher agricultural prices would manifest itself and impact positively on poverty reduction. However, in practice the liberalisation of agricultural trade by subjecting non-food agricultural crops to international competition and by displacing domestic jobs keeps the prices of the former down and dampens money wage increases. On the other hand, shifts of acreage out of food into non-food crops and the release of food prices after liberalisation, tend to increase the price of food. In such an event, policies of stabilisation and structural adjustment worsen poverty, by neutralising whatever positive effects any increase in agricultural growth may have.
 
Even from a growth point of view, liberalisation has not benefited developing-country agriculture substantially, because of the limits to trade expansion generated by the subtle protectionism in the two principal developed country markets: the US and the EU. Although world trade as a whole as grown much faster that world GDP, this is more true of manufacturing trade and production than of agriculture. For those developing countries which, expecting to gain from an expansion of agricultural exports, liberalised agricultural export trade, this was a major setback. And since, agricultural growth is one of the important means to poverty reduction, it indirectly contributed to a slowing of the pace of poverty alleviation.
 
Agricultural Growth, Non-agricultural Activity and Poverty
If inequality is increasing, we should expect that the effect of any increase in per capita agricultural output on poverty would be weaker. However, there have been a number of experiences to the contrary. In India, for example, the Green Revolution of the 1970s and 1980s, while leading to some increases in agricultural output per capita, was characterised by some increase in concentration of operated area and marketed surpluses, as well as a substantial increases in regional inequalities in agricultural production. Yet, the incidence of poverty during these years was declining significantly, forcing researchers to look to other factors that could explain the decline in rural income poverty.
 
This decline was all the more surprising because evidence indicated that the output increases during the Green Revolution years and later were accompanied by a decline in the output elasticity of the demand for labour in agriculture. There seemed to be one factor which was neutralising the effect of these trends, viz. a rise in agricultural wages, which was then seen as an important influence on poverty. But why were real agricultural wages rising , if employment in agriculture was inadequately responsive to agricultural growth? The empirical answer seemed to lie substantially in an increase in rural non-agricultural employment. Over the 15-year period from 1972-73 and 1987-88, the share of rural male non-agricultural employment in total rural male employment rose by 9 percentage points and that of female non-agricultural employment in total rural female employment by 5 percentage points. This seemed to make the growth of rural non-farm activities and rural non-farm employment an important cause for reduction in rural poverty – a conclusion supported by experience in other Asia-Pacific countries, especially China.
 
Here too, the role of the State is crucial, not only in the form of State support for rural enterprises. Rather, a significant part of the impetus for growth in non-agricultural activities could come from outside the rural sector, mediated in large part by government expenditure on the provision of extension services, on social service provision and on employment generation. That is rural incomes are no longer based on or derived only from agricultural production, but by the specific forms in which rural areas are being integrated into larger macro-economic decisions regarding expenditure. If such decisions favour larger expenditure in rural non-farm activities, the effect on poverty alleviation is likely to be positive. And inasmuch as liberalisation and structural adjustment result in international competition that displaces small enterprise production and in fiscal policies that reduce government expenditures, countries which were benefiting from the poverty-reducing effects of the increase in non-agricultural activities, may suffer a setback in their poverty alleviation efforts in the wake of globalisation.
 
These aspects of the poverty problem make clear that the war on poverty requires that the State plays a significant role in economic activity and implements a well-worked out anti-poverty plan. As the UNDP's Poverty Report 2000 put it: ‘Why have a plan at all?’ some might ask. Isn’t planning old-fashioned in market driven economies? Perhaps, but markets don’t promote social justice. That takes organised public action.