viously have had access to skills training, extension services or credit facilities.

     A major social equity issue in agriculture is the perpetuation of poverty from one generation to the next due to the high incidence of child labor. Approximately 70% of all child labor is found in agriculture. Unpaid work on the family farm may or may not have an incidence on the child's school attendance and performance, depending on the hours and conditions of work. However, time lost to education, particularly if low achievement levels lead to early drop-out, has lifelong consequences on earnings. Much child labor in commercial agriculture is invisible and unacknowledged, although it may account for a considerable portion of family earnings (WDR, 2007).

     Social equity issues, such as child labor, must be addressed if broad-based agricultural development is to contribute positively to both economic growth and poverty reduction. The principal challenges are twofold: raising the living standards of those working in agriculture, particularly the poorest among them, and lessening the demographic burden on agriculture by providing opportunities for more diversified and rewarding economic activity outside the sector. Educating rural children and preparing them for a productive future addresses both those concerns and AKST can be instrumental in achieving this in a number of ways. For example, well targeted AKST can enable poor farmers to increase their earnings sufficiently to keep their children in school, rather than at work. The adoption by parents of innovative farming practices can teach children the experience of lifelong learning, openness to technological change and the benefits of applying knowledge to production. Incorporating AKST into rural school programs could provide young people with practical skill sets to help them make the transition to more productive work in agriculture or in rural support services, or could inspire them to pursue other science based studies.

     The labor requirements of various crops or cultivation methods are an important variable that needs to be considered. AKST is not employment-neutral, nor can it be if it is to improve the livelihoods of the rural poor. In some poor communities and households, the greatest challenge is to generate productive employment for able-bodied workers. In such circumstances, the development of high-value, highly nutritious, labor-intensive crops may offer opportunities for improving livelihoods and well-being. In other cases, laborsaving crops and techniques may offer better outcomes, for example, for labor-poor female headed households, or rural communities suffering from a high incidence of HIV/AIDS or other debilitating illnesses.

     Many observers note a dichotomy between small-scale agriculture and industrialized agriculture. Indeed, the uneven competition that has emerged between small- and large-scale production systems raises serious social equity issues within the agricultural sector as a whole. The two systems differ greatly in terms of resource consumption, capital intensity, access to markets and employment opportunities. The economic and political power of agribusiness enterprises and their relative importance in national economies enable them to influence decisions regarding domestic support packages, infrastructure investment, the direction of agricultural research and development and the setting of international

trade rules in ways that small-scale farmers cannot. Another major difference lies in their capacity to provide employment. Large-scale production systems are often in a position to offer better terms of employment, but they tend to shed labor as productivity gains are realized through technology and more efficient work organization. Although the number of persons working in small-scale agriculture has decreased as a percentage of the global population in recent decades, it has steadily increased in absolute numbers and is estimated to include approximately 2.6 billion people or 40% of the world's population (Dixon et al., 2001).

     While the notion of dichotomy may be useful in drawing out such contrasts, it tends to mask the wide range of ownership patterns, relationships to the land, forms of labor force participation and employment relationships that generate profound social equity issues. It is instructive to consider how just one set of rights-property rights-affects the livelihoods of various stake-holders in the agriculture sector: plantation owners, medium to small-scale owner-cultivators, tenant farmers, share-croppers, squatters, landless laborers, bonded laborers, migrant workers, or members of an indigenous community sharing common lands. These categories are not discrete; indeed, there is frequent overlap among them, and cutting across all these categories are issues of gender, which further define or delimit rights of ownership, access, use and inheritance of the land.

Choices to be made: agricultural productivity and poverty reduction

Most discussions of broad-based agricultural development focus on the interaction of five main factors-innovation, inputs, infrastructure, institutions and incentives (Hazell, 1999). Equity issues are inherent, though they may not be explicitly evoked, in the policy decisions that guide the investment of resources in these areas. For example, agricultural research and development is needed to generate productivity- enhancing technologies, but choices must be made as to the orientation of research efforts. The improvement of local food crops to better satisfy nutritional needs, the development of drought-resistant breeds to provide a more reliable harvest to those living on marginal lands, or the development of horticultural produce suitable for export may all be worthy goals in themselves, but have very different potential beneficiaries. Whether or not these activities lead to improved livelihoods for the poor depends on many factors, not least among them being the social characteristics of particular rural communities and the convergence of innovation with other productivity factors. Ownership or control of land and other assets, knowledge and skill levels, roles and responsibilities with regard to production, access to affordable credit, and rights with regard to distribution of services vary considerably across and within social groups. Ethnicity, class, sex and age all affect the capacity of those who work the land to access and use new technologies effectively and profitably, but take-up can be modified with well-targeted interventions. Productivity enhancement is not so much a technical issue, as one of political, economic and social choices and constraints, hence an issue of equity (HDR, 2006).

     This is well illustrated by a number of "equity modifiers" that have been suggested as a means to reduce poverty