and produce energy (Balce, et al., 2003; Ghosh, et al., 2006; IEA, 2006b). Despite the fact that production costs can be competitive in various settings, in the past many attempts to promote wider distribution of modernized bioenergy applications have failed. Common problems included technical difficulties and the failure to take into account the needs and priorities of consumers, as well as their technical capabilities, when designing promotion programs (Ezzati and Kammen, 2002; Kartha, et al., 2005; Ghosh et al., 2006).

Bioelectricity and bioheat production can be competitive with other sources of energy under certain conditions, especially the combination of heat and power generation within industries producing waste biomass.

The competitiveness of bioelectricity and bioheat depends on (1) local availability and cost of feedstocks-many of which are traded on market with strong prices variations both regionally and seasonally; (2) capital costs and generation capacity; (3) cost of alternative energy sources; and (4) local capacity to operate and maintain generators. Generally, bioelectricity production is not competitive with grid electricity but generation costs can compete with off-grid option such as diesel generators in various settings. Key to competitiveness is a high capacity utilization to compensate for relatively high capital costs and exploit cheap feedstock costs. High capacity factors can best be reached when proven technologies (e.g., thermochemical combustion) are employed on site or near industries that produce biomass wastes and residues and have their own steady demand for electricity, e.g., sugar, rice and paper mills. Estimates for power generation costs in such facilities range from US$0.06-0.12/kWh (WADE, 2004; REN 21, 2005; World Bank, 2005a; IEA, 2006b). In combined heat and power mode, when capital investments can be shared between electricity and heat generation, electricity generation costs can decrease to US$0.05-0.07/kWh, depending on the value of the heat (REN 21, 2005; IEA, 2006b). Thermochemical gasification can have higher generation costs and low capacity utilization due to weak electricity demand, and technical failures caused by improper handling and maintenance can lead to even higher production costs (Larson, 1993; World Bank, 2005a; Banerjee, 2006; Ghosh et al., 2006; Nouni et al., 2007). Data on electricity production costs with anaerobic digesters are not widely available, because most digesters are not installed commercially but through government programs to provide (1) energy access for rural households and villages, often solely for the provision of cooking fuel or heating or (2) methane capture on environmental grounds (e.g., in several industrialized countries). Overall, the economics of biomass power and heat can be improved through carbon credits.

3.2.3. Impacts of AKST on livelihoods, capacity strengthening and empowerment

3.2.3.1 Methodologies and approaches for assessing impact

Assessing the evidence for the contribution of AKST to improving livelihoods and empowerment is complex. While

there is evidence of contribution to increasing productivity of agriculture and sustainability of natural resource use, the extent to which this is relevant to specific groups of people and translates into improved livelihoods, is more complicated, involving differential impacts between and within populations. The difficulty of attribution applies similarly to negative outcomes. The paths of causality are complex and highly contingent on specific conditions (Adato and Meinzen-Dick, 2007) involving interactions between AKST and the policies and institutional contexts in which AKST products are promoted and adopted. Hence the assessments of impacts are sometimes contradictory or controversial. The methodological challenges of impact assessment are considerable; especially when going beyond economic measures of impact or individual case studies. Thus it is difficult to make broader statements on the poverty and livelihood impacts of AKST investments and products across different geographical regions and client groups. Impact assessments rely on comparison-before and after a specific intervention or change, or a "with" and "without" situation (the counterfactual either being empirically measured, or theoretically constructed assuming the best available alternatives are pursued). This approach has been helpful in establishing the economic returns from agricultural research and the contribution of increased productivity, but is more difficult to construct for the livelihood dimensions.

3.2.3.1.1 Assessment of the economic impacts of AKST Past assessments of impacts of specific AKSTs have documented adoption, productivity increases and financial returns and consequences for national food security (Hazell and Ramasamy, 1991; Evenson and Gollin, 2003a). There is evidence that agricultural productivity growth has a substantial impact on poverty reduction, although this is conditional on contextual and socioeconomic conditions, e.g., equitable land distribution (Kerr and Kolavalli, 1999; Hazell and Haddad, 2001; Jayne et al., 2003; Mathur et al., 2003; Thirtle et al., 2003). Economists have developed techniques to quantify the total economic value of the multitude of products and services (social/environmental and local/ global) from agricultural programs, such as agroforestry (Pearce and Mourato, 2004).

Impact assessments of investment in agricultural research have shown that it has been highly cost effective.

Investment in research has resulted in substantial economic gains from increased productivity. For example, in the case of the CGIAR system benefit-cost ratios for research have been between 1.94 (significantly demonstrated and empirically attributed) and 17.26 (plausible, extrapolated to 2011) (Raitzer, 2003). Three innovations-MVs of rice (47% of benefits), MVs of wheat (31% of benefits) and cassava mealy bug biocontrol (15% of benefits) account for most of the impact using the most stringent criteria, and are worth an estimated US$30 billion [at 1990 values] (Heisey et al., 2002; Evenson and Gollin, 2003b; Hossain et al., 2003; Raitzer, 2003; Lantican et al., 2005). While focused on a very narrow range of species, as a measure of this