Ethiopia Study

CIMMYT carried out a wheat DNA fingerprinting study with CSA two years before our work, so CSA did not want to integrate a protocol for sampling wheat again. CIAT has also been working on DNA fingerprinting of common beans in a separate sample. The data from these two studies and those of the crops we did include (maize, barley, sorghum, sweet potatoe, chickpea) are therefore not directly comparable.

The report shows the reach of innovations to which the CGIAR research has contributed. In all cases, other actors also contributed to the research leading to the innovations or to the dissemination of those innovations. The intent is not to attribute the wide-spread adoption of innovations to CGIAR’s efforts relative to other actors – a task that is arguably impossible.

CGIAR research has informed major donor-funded programs on soil and water conservation practices in Ethiopia. The impact pathway is fundamentally different from, for example, crop germplasm improvement where there are "embodied" technologies that emerge from CGIAR research partnerships.

The report documents 'reach', which together with careful causal estimates of the impact of selected innovations, can be used for a new way of calculating rates of return or cost-benefit ratios to CGIAR investments as a whole. See “SPIA approach to impact assessment for CGIAR (2020)” and the work of Michael Kremer at USAID’s Development Innovation Ventures for an exposition. Such an exercise is outside the scope of this report.

The answer is innovation-specific and the report highlights some factors that deserve further investigation. For example, for crossbred poultry, the role of  the policy framework of the Livestock Master Plan, as well as the public-private partnerships boosting private sector investment, is worth understanding better. Improved barley varieties that were recently released yet found in farmers’ fields are another example that deserve more investigation for their relatively rapid diffusion.

We only covered the reach of innovations in this report and did not document the impact on household-level outcomes. This would require combining survey data with a causal research design. SPIA’s approach to impact assessment is exposed in this Technical Note.

We did consider extension services efforts in the stocktake, in order to identify innovations that should be collected in the ESS. We did not make the role of extension an explicit focus in the report but it is an excellent candidate for further analysis with the ESS data.

Capacity development was not part of the Stocktaking exercise, in part because it is often less well documented. Linking capacity development to adoption of specific innovations as measured in household surveys would also be challenging. 

This is an important question, but one that is hard to answer with the ESS data. Targeted studies considering the impact of specific targeting efforts could be designed to analyze this directly.

The report certainly does not intend to make a case for a unified ToC [1]. Instead, it proposes that the information on the types of households that are chosing to adopting a specific innovation (and equally important those that are chosing not to adopt) can be compared with the assumptions made in the ToC of that specific innovation. This allows documenting whether the specific innovation is reaching the type of households it intended to target.  

[1] Indeed, studies of SPIAs adapted strategies call specifically aim to test the impact of scaling strategies that are adjusted to the specific characteristics of an innovation (and the related ToC).