The performance evaluation design and implementation will address the following overarching questions across the geography, activity scope and the Project Logic of the Kigoma Solar PV program:
i. How well was the program implemented? (Including analysis of Project scope, timing, costs, and public perceptions); were the output targets achieved? If not, why?
ii. What type of challenges were encountered during implementations?
iii. How well has the solar energy approach addressed the energy needs of the beneficiary population?
iv. What are the outcomes of the program on solar energy access, use and costs as well as productivity income etc.?
v. How sustainable are the outcomes?
vi. What lessons can be learned from the experience of the program?
vii. Was the project successful in catalyzing investments in the energy sector in Kigoma? If not what conditions will needs to be in place for the pilot to encourage additional investments?
viii. Are there any unplanned results due to the implementation of the Kigoma Solar project?
In addition to the above the evaluation design and subsequent data gathering activities will address the following key research questions on outcome, objectives and Compact Goal:
i. Has the Kigoma Solar Project contributed to an improvement in electricity service coverage across different customer types?
ii. Has the Kigoma Solar Project contributed to an improvement in the quality of electricity available, across different customer types?
iii. Has the Kigoma Solar Project contributed to an increase in consumption of electricity, across different customer types?
iv. Has the Kigoma Solar Project contributed to an increase in investment in economic activities across different customer types?
v. Has the Kigoma Solar Project contributed to an improvement in human capital accumulation across different customer types?
vi. Has the Kigoma Solar Project contributed to a reduction in poverty across different customer types, as measured by household income per capital?
The study is also expected to collect data that will feed into the formulation of the Economic Rate of Return (ERR) model for the Kigoma Solar program. This will include data on the sources of energy used prior to the installation of PV systems and their associated prices and quantities data, energy use and cost after the PV installation, ways and purposes for which the system are being used, and energy sources used in homes in the project area.
Kind of Data
Unit of Analysis
Community, School, Health Center, Village Market, Enterprise, Household
Anonymized dataset for public distribution
Producers and sponsors
Millennium Challenge Corporation
(a) Stages of sample selection
The evaluation included institution, business and household surveys. There were three stages of sample selection that were implemented at different levels. The first selection was to adopt the Kigoma Rural and Kasulu districts that are targeted by the program as study areas. The second selection involved sampling of units of observation for the institution, business and household surveys. The third sampling was selection of respondents for fishers, business and households surveys. Sampling of responding units was intervention based. It involved subjects covered by the intervention for the pre-post comparison evaluation method. This was complemented by including communities not covered by the Program as comparison group for assessing baseline status in the additional treatment and comparison groups' evaluation method. However, some facilities were eliminated from the comparison sampling frame because they were too far to get to. Units of observation were selected based on program beneficiaries and stakeholders provided in the ToR.
The detailed deployment plan for the program was used as sampling frame. The evaluation used Cluster sampling design with stratification because the communities are clusters while the sources of data are stratified into treated and non-treated. For efficiency purposes, villages with most of the types of the solar PV installations (village market, secondary school, dispensary, health centre, BMU and SACCOS) were purposefully be given priority. This was to ensure all types of installations, conduits and beneficiaries are covered by the evaluation. The selection of comparison of groups' villages and facilities was also done purposefully to ensure that they are located far from the target villages and facilities of the program to the extent that they cannot access services provided by the program.
The first selection was for villages to be covered. The second stage sampling was also done purposefully based on the distribution of installations in the detailed deployment plan for the program as sampling frame and the above sample selection guide. The selection included Beach Managemetn Units (BMUs) that are also given in the detailed deployment plan for the program. It also included selection of SACCOs based on their list obtained from MCA-T and updates by Cooperative Officers of the two districts. Once an institution was sampled as a unit of observation, the head of the institution such Headmaster/mistress was automatically or purposefully selected/identified because we had already planned to interview the institution. The third stage sampling involved selection of units of respondents for outcome indicators namely fishers/boat owners that purchased PV systems through BMUs, Businesses connected to the PV system of the village market, and businesses and households that purchased the PV systems through SACCOs. A total of 170 respondents were selected for the various surveys.