Our evaluation demonstrated that a strategy of mobilizing local vendors, using pricing incentives and promotional offers, and integration with household water treatment accelerated adoption of locally-produced, ceramic cookstoves (upesi jiko
) into a large number of rural Kenyan households. While there is an abundance of literature on improved cookstoves to mitigate household air pollution, this report offers a description of the specific processes and mechanisms for delivering cookstoves to households. Given that rates of cookstove adoption may be low and the factors that determine adoption are not yet well understood [26
], similar evaluations are needed for other improved cookstove technologies in other settings. Furthermore, lessons from programs that have successfully delivered other types of interventions (e.g., insecticide-treated bed nets, micronutrient supplementation packets) may be extrapolated to cookstoves.
The strategy can be understood in terms of the theories of Rogers’s Diffusion of Innovations
]. The networks of SWAP vendors accelerated the rate of adoption of the upesi jiko
, both by acting as persuasive community peers, with roles as opinion leaders, and by providing specific health and product knowledge that was transferred during project trainings and amplified through interpersonal communication channels. These same SWAP vendors were motivated by the potential for modest but meaningful income with the sale and installation of upesi jiko
. In fact, ~70% of SWAP vendors were actively involved in the upesi jiko
market, as evidenced by multiple transactions (i.e., a median of seven transactions). Also, pricing incentives and the promotional offers likely accelerated the adoption rate by increasing the relative advantage of upesi jiko
compared with traditional, three-stone stoves. As a result, a market that was attainable for many village members was established. Although these incentives and promotions were not necessarily sustainable because they require external funding, they were relatively inexpensive and probably indispensible; two-thirds of upesi jiko
transactions involved incentives/promotions during the project. Importantly, incentives and promotions were timed to follow crop harvests, which often provide extra, seasonal income. Other evaluations should determine whether the effects of pricing incentives are attributable to a lower absolute price or periodic price reductions acting as a stimulus.
Integration of environmental health interventions to improve air and water quality capitalized on existing NICHE and SWAP resources and infrastructure, and created additional opportunities to reinforce health messages and promote adoption of the upesi jiko
. For community members who had already adopted household-level water treatment, the upesi jiko
may have been compatible with an idea that was previously-introduced by SWAP; namely, products introduced at the household level can improve child survival. Furthermore, the upesi jiko
were compatible with needs reported by the community (e.g., desired reductions in smoke) and, in general, were likely perceived to have some beneficial attributes (especially savings in firewood expenditures). At the program level, integration of interventions has the potential to be more cost-effective and sustainable than separate, vertical efforts [28
]. However, the effects of adding new products such as cookstoves to the SWAP project need to be monitored for a potentially negative impact on the distribution of other products. The finding that upesi jiko
adoption was greater in households that had been treating water more frequently is consistent with the hypothesis that health messages and corresponding product interventions reinforced one another. It may also be true that households that accepted one intervention may have been more likely to accept other interventions because of other underlying factors, such as higher socioeconomic status or a greater degree of health education. Notably, our evaluation was not designed to determine the relative impact of each intervention while co-promoting household water treatment and upesi jiko
Our findings should be interpreted cautiously, as demographic and socioeconomic data were linked to a very small subset of all households that used upesi jiko
. The project resulted in adoption of upesi jiko
by numerous village members. Although the stoves were promoted without regard to demographics or socioeconomic status (i.e., village residents had equal access to the upesi jiko
), most of the upesi jiko
were acquired by households in the uppermost socioeconomic status quintiles and by households where older women reside. The implications of these findings are significant. For example, many of these households may have larger homes with more rooms or buildings, which would allow them to install upesi jiko
in separate buildings (i.e., away from where children sleep) more frequently than households with lower socioeconomic status. Also, we did not have adequate data to determine the proportion of households with young children that were able to install an upesi jiko
. Therefore, the overall success of the project so far has been diminished by the inequitable adoption of the cookstoves [30
] and possibly by incompletely reaching the intended target population, young children, for reduction of exposures to household air pollution [31
]. Although demographic and socioeconomic characteristics are important determinants, other research in Africa suggests that the attributes of the cookstoves themselves also determine whether cookstoves will be adopted at the household level [33
]. Indeed, adoption of improved cookstoves is a complex process; even when new cookstove technologies are adopted, they may be used in addition to (rather than instead of) traditional alternatives [34
]. A subsequent evaluation will determine whether the adoption of upesi jiko
has had long-term success though the ongoing, appropriate use of the upesi jiko
. Furthermore, it is understood that our evaluation of a multifaceted strategy, including local vendors, promotional incentives, and product integration, did not allow for separate assessments of the impacts of each of these components of the strategy. Specific limitations also may apply to each component of the evaluation. For example, water treatment was generally categorized with consistency when measured through self-reporting and chlorine residuals, but discrepancies were identified. The largest discrepancies were households with reporting that suggested sporadic treatment, but chlorine residuals indicated frequent treatment (37%) and households with reporting that suggested sporadic treatment, but chlorine residuals indicated no treatment (34%).
Although this project effectively incorporated these cookstoves into a large number of households of Luo ethnicity, additional applications will be needed to determine whether a strategy of local vendors, promotional incentives, and product integration can be applied equitably in other communities. Planning and design of these programs should include formal evaluations to determine the effectiveness of each component of the stove-adoption strategy. A considerable body of emerging research is determining whether reductions in fuel use, household air pollution, and the incidence of respiratory diseases result from the use of improved cookstoves. At the same time, integrated, scalable strategies for improving access to cookstoves in rural households are also needed to accomplish these reductions.