With a major portfolio funded by a multi-donor trust fund initiated in 2005 and now fully committed with 35 country pilot programs and a rich evaluation portfolio which includes 28 rigorous impact evaluations and other evaluation studies like qualitative studies and enhanced program assessments, RBF in health is certainly no longer a “new kid on the block” for the World Bank. Whether the initial excitement remains at the same level or not, the truth is that there are still a number of unknown factors, and continued unpacking of the RBF box will not only help satisfy the intellectual curiosity of social scientists but also distill real lessons for in-country operations supported by the World Bank and others.
In this blog, I want to share what we learned from a recently completed set of mixed method studies (on Zimbabwe, Zambia, and Benin), conducted under the umbrella of an analytical product called “exploring human resource in health (HRH) issues to inform the design, implementation, and evaluation of RBF programs.” This is clearly not the first attempt made globally to bridge the link between incentive payment and health worker outcomes. An extensive literature - building on theories from different social science disciplines, lab in the field experiments, and limited empirical evidence in health and other sectors – is out there: it expresses concern (or not) on the impact of monetary and non-monetary rewards on the intrinsic motivation of health workers and public sector workers as well as on intrinsic and extrinsic motivation. However, this work is one of the first attempts to look in-depth into HRH issues in RBF programs in the HRITF portfolio. It seeks to complement the series of well-designed and funded impact evaluations, which focus primarily on RBF’s impact on service coverage and on the quality of care. The HRH study focused specifically on supply-side RBF – or PBF – which is the operation in the three studied countries.
The quantitative part of the three country studies is largely piggybacking on existing impact evaluations, and survey data was supplemented by a qualitative assessment. Because the quantitative part piggybacked on respective impact evaluations, its design is not standardized across the three studies and the richness of information also varies depending on what’s available. Specifically, for Zimbabwe, the outcomes of interest are health workers’ satisfaction and motivation; for Zambia, additional outcomes include some rough measures of attrition and productivity; while for Benin, the team was able to assess real performance aspects, which are clinical quality of care and patient satisfaction. Both Zimbabwe and Zambia studies are based on pre and post data while Benin’s analysis is based on post-RBF data only.
The Starting Point – The Assumed Causal Pathway
We started out with a diagram developed by our colleague Christophe Lemiere which I found very intuitive and appealing (reproduced below). The idea is: the ultimate HRH outcomes at the individual worker level – health worker’s performance – has multiple dimensions and that there are, in turn, many factors influencing these dimensions. Quality of care, responsiveness to patients, and productivity—the key performance dimensions—are driven by the “can-do” and “will-do” factors. “Can-do” factors include skills and knowledge as well as working conditions; “will-do” factors refer to the degree of effort by health workers, driven by motivation, self-advocacy, or pressure from supervisors and peers, etc. By design, RBF touches upon both “will do” and “can do” factors which, theoretically, should lead to better performance. Without being too nuanced on intrinsic versus extrinsic motivation or satisfaction, we also thought RBF should have a positive impact here.
A Closer Look at the Causal Pathway
A deeper dig into the causal pathway, and especially during the analysis of qualitative information already collected from the field, revealed that there are more nuances in the diagram than we thought. And this actually urged us to come back and consult the literature again, focusing on features that are common in the three country RBF programs as well as in many others. Turned out, the causal pathway is neither single-tracked nor linear: there are multiple paths from RBF to satisfaction, motivation and, ultimately, to performance; these paths can also work in opposite directions depending on how they are operationalized. The net effect is, in typical econometric jargon, “theoretically ambiguous.”
For example, we typically think that one of the important contributions of RBF is to enhance supportive supervision from a higher level health authority (district or county) to front line service providers. Without RBF, in many countries, many of these facilities stay forgotten and unvisited for a long time, and hence the mere fact that they now receive attention from the higher level already motivates the people there. Some IEs have “supervision only” as a study group in a multi-arm Randomized Control Trial. Digging deeper, there’s a dirt of literature that says, “enhanced supervision” could have a positive impact on motivation if perceived as “supportive.” However, if it’s too much/too rigid, it could be perceived as “controlling” and the impact on motivation can be totally reversed. Along the same line, but at a much higher level of complexity, is the question of effect of financial incentives which is the core in any of these RBF programs. To summarize in simple language, financial incentives could actually have a negative effect if they are perceived as “too small”; they can choke if perceived as “too big”; or they can backfire if perceived as “unfair”. They will however be motivating if they are perceived as “fair and right sized” (what is “fair and right sized” is a million dollar question of course). And the list goes on…. For sure we can all resonate to these from our own experience as workers in a multi-employee institution…
Quantitative evidence shows a mixed and rather puzzling picture:
- In Benin, the study showed no quantifiable impact of RBF on motivation (“will do factor”). RBF also didn’t improve volume and productivity of ANC services, but seemed to improve content of ANC care;
- In Zambia, RBF performed better than pure control group on attrition and some satisfaction/motivation dimensions. However, it has no quantifiable improvement compared to the (supposedly equivalent) additional financing group on any dimension rather than “satisfaction with compensation.” Additional financing appeared to compare better with pure control on HRH and some service delivery outcomes than RBF;
- In Zimbabwe, RBF’s impact on “satisfaction with compensation” is positive, however, it is consistently negative on a number of motivation factors. Yet, Zimbabwe program performed very well on population coverage and quality of care.
Qualitative findings are in general positive but also provide some hints which could help explaining quantitative findings:
- Incentives are perceived as unfairly distributed by some, especially lower level cadres. There seems to be no easy way to make everyone happy and to eliminate “free riders”;
- Supervision is appreciated, but sometimes perceived as too restrictive, lack of understanding of reality on the ground, and lack of coaching/mentoring content. Too frequent supervision may shock;
- Workers burn out due to the increased volume of patients may be a factor affecting motivation, especially in the case of Zimbabwe where service outputs increased drastically partly due to the removal of user fee in the RBF program;
- Some RBF features for “can do” may not be in place as they should be, such as autonomy. Some factors are beyond facility control; and
- RBF rules are not well understood.
Taking together, the three studies suggested some common patterns: (1) there was a positive effect on workers’ satisfaction with improved compensation, and, to some degree, working conditions; (2) the impact on motivation is mixed; (3) no significant effect was found for productivity; and (4) there was some positive impact on quality of care and attrition. All three studies showed a rather appreciable contrast between the (rather modest) quantitative results and (very positive) qualitative evidence, suggesting measurement limitations with both types of instruments. All studies point to the importance of getting the design and implementation features right to assure positive HRH outcomes, in particular fairness in the allocation of RBF bonus and quality of supportive supervision.
The Take Home Messages
Coming back to the point “theoretically ambiguous,” I guess it is not extremely helpful if we complete a rich set of studies with a conclusion that it all depends on how things are operationalized on the ground. Clearly, more work needs to be done to understand the human aspects in RBF, including the work to improve the measurement instruments on both quantitative and qualitative sides. However, putting together different pieces of evidence, some key lessons can be drawn which are hopefully useful for the RBF programs themselves:
- Attention could be paid on supportive supervision to make sure that they are truly “supportive.” This could be done with training the supervisors, giving them the skills and exploring the optimal frequency as well as timing of supervision;
- Inter-facility communication could be improved to make sure staff are clear about RBF rules, allocation mechanism and comfortable with it, and to foster team work;
- The issue of the size of incentives and fairness in allocation is of utmost importance and in itself would take at least another study if not a series of studies and trials. For each program, it is a trial and learn and the opportunity to share experience may be useful here;
- A closer investigation of the effect pathways in the conditioned financing (RBF) and unconditioned financing (additional funding) needs to be pursued. Maybe a mid-way configuration would make the most sense to optimize outputs while reducing transaction cost which is quite high in RBF programs.
At the end of the day, one may ask “why do we care about workers’ motivation? As long as they perform and produce, why do we care if they are motivated or not?” Without getting very philosophical, I would say – we do care – and among many reasons, we care because motivation is essential to the sustainability of good performance. And we know some about what affects motivation and performance in RBF but there is still a lot to explore to crystalize the lessons learned.
Bertone MP, Witter S. The complex remuneration of human resources for health in low-income settings: policy implications and a research agenda for designing effective financial incentives. Human Resources for Health. 2015;13(1):1.
Bertone MP, Lurton G, Mutombo PB. Investigating the remuneration of health workers in the DR Congo: implications for the health workforce and the health system in a fragile setting. Health Policy and Planning. 2016; czv131.
Bertone MP. Financial incentives for human resources for health: What do we know? What do we do? The case of Sierra Leone. BMC Health Services Research. 2014;14(2):1.
Deci EL, Koestner R, Ryan RM. A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin. 1999;125(6):627.
Farrell D, Rusbult CE. Exchange variables as predictors of job satisfaction, job commitment, and turnover: The impact of rewards, costs, alternatives, and investments. Organizational Behavior and Human Performance. 1981; 28(1): 78-95.
Heneman RL, Greenberger DB, Strasser S. The relationship between pay‐for‐performance perceptions and pay satisfaction. Personnel Psychology. 1988; 41(4): 745-759.
Judge TA, Piccolo RF, Podsakoff NF, Shaw JC, Rich BL. The relationship between pay and job satisfaction: A meta-analysis of the literature. Journal of Vocational Behavior. 2010; 77(2): 157-167.
Judge, Timothy A., Carl J. Thoresen, Joyce E. Bono, and Gregory K. Patton. "The job satisfaction–job performance relationship: A qualitative and quantitative review." Psychological bulletin 127, no. 3 (2001): 376.
Judge TA, Bono JE. Relationship of core self-evaluations traits—self-esteem, generalized self-efficacy, locus of control, and emotional stability—with job satisfaction and job performance: A meta-analysis. Journal of Applied Psychology. 2001; 86(1): 80-92.
Judson TJ, Volpp KG, Detsky AS. Harnessing the right combination of extrinsic and intrinsic motivation to change physician behavior. JAMA. 2015;314(21): 2233-2234.
Mobley WH. Intermediate linkages in the relationship between job satisfaction and employee turnover. Journal of Applied Psychology. 1977; 62(2): 237-240.
Tett RP, Meyer JP. Job satisfaction, organizational commitment, turnover intention, and turnover: path analyses based on meta‐analytic findings. Personnel Psychology. 1993; 46(2): 259-293.
Willis-Shattuck M, Bidwell P, Thomas S, Wyness L, Blaauw D, Ditlopo P. Motivation and retention of health workers in developing countries: a systematic review. BMC Health Services Research. 2008; 8(1): 247-254.
 The muti-donor trust fund is known as “Health Results Innovation Trust Fund” or HRITF