The research project will be attributed to and in close cooperation with the PBF project of the BTC in Uganda. The contacts with the responsible BTC project manager are already made and it has also been authorized by the coordinator of the health unit of the BTC. This cooperation will not jeopardize the neutrality of the research nor that of the researcher. However it will facilitate the gathering of information, the understanding of the project and the relations with the Ugandan partners of the PBF project. For the BTC and the Ugandan partners this will mean that they get an extra input of information which can help them in the implementation of the project. At the same time, the researcher finds himself at a privileged point between the academic and the policy world, being not only able to generate new valuable academic knowledge but also to influence future policies. The approach of Theory-Based Evaluation (TBE) (e.g. Astbury & Leeuw 2010) forms the basis of the general methodology of our research proposal. A TBE always starts with an overview of the literature and the findings on the program theory of the project. This was done by the researcher on the basis of an earlier developed analytical framework based on the Principal-Agent theory (P-A) (Renmans et al. 2014). From this overview we learned that several important elements of the program theory of PBF were heavily under-investigated. The next step was to create so called ‘Context-Mechanism- Outcome Configurations’ (CMOC) (Tilley 1999) which are “middle-theories, drawing on or linking to more abstract theory, but standing above the descriptive particulars of individual cases” (p. 59) and this resulted in the abovementioned research lines and hypotheses. These will be investigated through a mix of quantitative (questionnaires) and qualitative (interviews) methods within a comparative research design. The BTC project will be implemented in two regions through a fragmented roll-out of the program which will give us the possibility to compare the evolution of such variables as performance, motivation and perceptions between PBF and non-PBF health facilities. The several variables elucidated in the program theory and the five research lines, will be researched through the use of standardized questionnaires and semi-structured interviews with health workers, management and supervisors of the health centers. A standardized questionnaire on health worker motivation (extrinsic/intrinsic, general motivation) will first be created on the basis of known literature (e.g. Mbindyo et al. 2009) and discussions with the Ugandan partners to ensure the local embedding of the questionnaire. This will be complemented with semi-structured interviews on health workers’ motivation, perception of the PBF, the M&E system, the sanctioning mechanism, own performance and rent seeking behavior among colleagues. Semi-structured interviews with supervisors, implementers (BTC and Ugandan) and managers will inform the final research line on contract and implementation. A study of primary documents from the BTC, the controlling agency and the Ugandan government will complement the interviews, especially for the research questions on rent seeking behavior. These questionnaires and interviews will be executed at three different moments (before the project beginning 2015, after +/- 1 year and after +/- two years) to enable us to analyze the evolution of perceptions and motivation and use a difference-in-difference approach between the PBF and non- PBF health facilities. In the end, the (not) found differences will be linked to causal mechanisms retrieved from the many interviews. Eventually, the results obtained from the research will lead to a new and more complete program theory on performance-based financing which can help future projects to develop better and more effective projects or decide to go for another approach. It will also inform the BTC and the Ugandan government on the desirability and the options for expanding PBF nationwide.
|Effective start/end date||30/04/15 → 29/08/18|
IWETO expertise domain