Abstract
Halfway through the final decade of actions towards the 2030 Sustainable Development Goals (SDGs), progress toward SDG7 is off track. It is estimated that by 2030, 660 million people, mainly rural populations within Sub-Saharan Africa, will be without electricity. One promising avenue to counteract this trend is the provision of decentralised, or off-grid, renewable energy. Our systematic review synthesised the rigorous evidence evaluating off-grid electrification interventions and provides policymakers, practitioners and researchers across the sustainable energy field with an updated and comprehensive analysis of the impact of off-grid electrification interventions. Our systematic review synthesised the available rigorous evidence on the effects of off-grid technologies in low- and middle-income countries. We assessed which off-grid interventions are effective at supporting access, climate and socio-economic development outcomes, how these effects vary by region, population and other intervention characteristics and the main challenges and facilitators for interventions to benefit participants. We conducted a systematic search in 18 academic databases and 29 grey literature sources. We supplemented our searches by conducting backward and forward citation tracking, publishing a call for additional studies and contacting subject experts. To identify additional qualitative studies, we performed additional searches for studies related to interventions from our included impact evaluations. We included experimental and quasi-experimental impact evaluations of interventions promoting the use and uptake of off-grid technologies in low- and middle-income countries. Interventions fell into one of four categories: the direct provision of technologies, the opportunity to purchase/market expansion of technologies, subsidies and credit to purchase technologies, information provision promoting the use of technologies. Studies in any language were included, though they must have been published since 2000. We also included qualitative studies to understand the main challenges and facilitators of intervention effectiveness. Quantitative data was extracted for all estimates deemed relevant and the risk of bias for each of these estimates was assessed independently by two reviewers. When data allowed us to do so, we calculated standardised mean differences for results from each study and used random effects meta-analysis to synthesise effectiveness findings for comparable outcomes. We provided forest plots and measures of heterogeneity for all outcomes and tested for publication bias in outcomes with more than 10 effect sizes. When feasible, we conducted moderator analysis to understand how effects varied by intervention characteristics and checked whether results were sensitive to the risk of bias score of estimates. For qualitative studies, we extracted and analysed data based on a previously developed framework for challenges and facilitators of sustainable energy interventions. Our review includes 47 impact evaluations, with the majority of studies conducted in Sub-Saharan Africa. Most studies evaluated an intervention either directly providing technology or providing participants financial support to purchase the technology. Solar home systems, solar lamps/lanterns and solar mini-grids were the most commonly implemented technologies. Our risk of bias assessment found that the majority of the evidence base, for both experimental and quasi-experimental studies, was of a high risk of bias. Spill-overs, cross-overs and contamination was the main bias dimension found in experimental studies, while confounding bias was most prevalent in quasi-experimental studies. For energy access outcomes, we found that interventions significantly reduced kerosene consumption and increased the use and uptake of off-grid technologies while also leading to a small increase in energy access. However, we found no statistically significant effect on other energy security measures, such as reliability and affordability, as well as lighting use and energy expenditure. We identified very few studies evaluating a climate-related outcome and found no significant effect on air pollution. For socio-economic outcomes, we found a small positive effect on income, time spent studying and women's empowerment. However, evidence of publication bias in evaluations of time spent studying suggests that this result should be interpreted with caution, as it may overestimate the true effect. We found no statistically significant effects on other measures of time allocation, school attendance and respiratory illnesses. Our qualitative results were derived from 19 qualitative and mixed-methods studies related to interventions evaluated across our included impact evaluations. We found that financial support is an important mechanism by which interventions may aim to increase the uptake and use of off-grid technologies. Local involvement in interventions was found to be a success factor, while information and marketing strategies were highlighted in multiple qualitative studies as a key factor in increasing participant engagement. Our systematic review synthesised the rigorous evidence on the effects of off-grid electrification interventions and can be used by policymakers, practitioners and researchers to inform decision-making on sustainable energy. Future research in this area should focus on filling gaps in the evidence base. Further evidence is needed in contexts outside of Sub-Saharan Africa, as well as on the impact of information provision. Our analysis, limited in many instances due to the small number and quality of studies, found positive effects across different energy access and socio-economic outcomes. Qualitative analysis highlighted the importance of ensuring that interventions are appropriate for local contexts and that local views and voices are built into intervention design.