Topmost menu

Designing a Health Service Program for Rural Uganda: Application of Expected Utility Theory


A large proportion of the population in rural East Africa live beyond walking distance to a clinic. Their access to healthcare is further compromised by poverty, limited health infrastructure, lack of health information, severely limited qualified health personnel, and lack of familiarity with the practices, assumptions, and culture of modern health care. This has left the subcontinent with some of the worst health indicators in the world.

Use of community health volunteers (CHVs) is a common strategy for implementation of health interventions in low-income communities. CHVs are women selected from the rural communities, and trained in basic and primary health care services. These CHVs are provided with life-saving, non-prescribed medicines to sell to the households in their communities with a profit margin they get to keep for themselves. Typically going door to door within the communities in which they live, these volunteer CHVs can bring healthcare to the sick and vulnerable, and can provide them with essential access to family planning advice, basic diagnostics and referrals, and quality information on affordable medicines and health products.[1] A recent random control trial in Uganda has demonstrated that if suitably incentivized and supported, health volunteers can reduce under-5 child mortality by 25 percent.[2]

Although CHVs are essential for implementing interventions to improve maternal and child health in low-income countries with limited health infrastructure, motivating their performance is a widespread problem.[3] CHV’s services are considered as volunteer works and very little financial incentive is offered to them for their services. Many of these CHVs become inactive in their job and eventually drop out due to lack of financial incentives. In response to the demand for providing sufficient financial incentives to the CHVs, the donor communities and NGOs raise the question about cost effectiveness and long-term sustainability of the financial incentives. In other words, they believe that such financial incentives will make the program highly aid dependent and therefore will not sustain in the long run. However, the question of burdening the already poor health workers with social responsibilities and very little income opportunities remains unaddressed.

Incentives and Utility of being a Health Volunteer:

Social (non-financial) and financial motivations are equally important in a CHV’s job. The CHVs should have a strong motivation to contribute to the welfare of their communities. At the same time, there should be enough financial incentives for the CHVs from this work. A recent study by Erika Deserranno[4] shows that relatively higher financial incentives attract less socially motivated people in community health work jobs and eventually result in a high drop out of the CHVs. CHVs with higher financial incentives perceive the job as a private goal job that they do to earn money rather than as a social goal job that they do to improve health in the community (Eirka Deserranno, 2015). On the other hand, although the socially motivated health workers initially remain happy with the low financial incentives, eventually they become demotivated with the incentive and become inactive in their job.

Findings from Field Study:

Working with the BRAC (Building Resource Across Communities) and Cape Breton University research team, we conducted field observation and focus group discussions with community CHVs in Uganda with to the goal of understanding the motivation, challenges, and opportunities they see in their role as health workers. Social recognition and respect by the community people were mentioned as the most satisfactory factors in their job. However, a majority of the health volunteers showed their dissatisfaction over the financial incentives they receive compared to the time and efforts they expend for this job. Visiting door to door, checking pregnant women and other patients, often take a whole day for a typical community health volunteer. “This is a 24-hours job, people call you over phone and ask for help even in the midnight,” as one of the CHVs said. Often the CHVs have to provide free medicine or at lower prices to the poor households, which reduce the small income they generate from the work.

Discussion with the best performing CHVs revealed that many of them are involved with other income-generating activities and they think their work as CHVs in the communities also help them in their other businesses. “Many women only buy stuff from my shops, they do not go to others since I have developed close relation with them,” a health volunteer mentioned.

Designing Sustainable Financial Incentives for CHVs: Application of Expected Utility Theory

In the existing model of BRAC’s health volunteer program, the CHVs only earn profit from selling the health commodities. The income generated from selling health commodities is often very little. As mentioned earlier, many households in the resource-constrained communities cannot afford to pay for the health commodities and CHVs often sell the medicines with zero profit, or even a loss.

The findings from the field study show that the involvement of the CHVs with other income- generating activities helps them to perform well in their community health works. This offers an opportunity to support the CHVs with business entrepreneurial skills and help them to develop a business that is complementary to the community health works. However, involving them in a complementary business may also create an extra burden of responsibilities and expectations on them. A business involves risks, especially for women with little or no entrepreneurial skills. The risk of losing money and the extra burden of running a business may deter the CHVs to be involved in BRAC’s health program. On the other hand, it also involves a risk of making the health volunteers too profit oriented, which might negatively affect their efforts and time spent on community health activities. It again takes us back to the situation in which the lack of financial incentives makes the CHVs inactive (or less active) and high financial incentives make them less committed to the community health works.

The team at BRAC designed a research project to investigate whether or not creating other income opportunities for the health volunteers increases the utility of the CSVs. The research also wants to generate evidence on whether involving the health volunteers in businesses (complementary to their health works) can be a sustainable income-generating tool for them. They are now implementing a new project where three models of financial incentives for CHWs are being analyzed:

Figure 1: Experimentation of Financial Incentive Models

The first incentive model involves the existing traditional model where the CHVs generate income only by selling health commodities (such as oral rehydration solutions, anti-malaria drugs, contraceptives, etc.). In the second model, in addition to the health commodities, the CHVs will be selling non-health products, such as soaps, cooking oil, flour, solar lamp, and fortified porridge etc. In the third model, the product basket will be the same as the first group (health commodities), but the CHVs will receive an additional income support as a fixed honorarium. The size of this honorarium will be equivalent to the average income that the CHVs in the business enterprise group earn from the sales of non-health products.

The test of these three models will help answer the following questions:

  1. The impact of the three different incentive models on community health volunteers (CHV’s) performance – alternatively, which incentive model increases the perceived utility of the health work job?

The health volunteers’ performance will be measured in terms of improvement of health outcomes in their respective communities.

  1. Comparative effectiveness of a risk-free incentive model (fixed honorarium) and the incentive model involving risk (business enterprise model).

It is apparent that the latter two models (business enterprise and fixed honorarium) offer higher income opportunities to the health volunteers and, therefore, expected utility of these two models should outweigh the expected utility of the traditional incentive model. Having the traditional incentive model as one of the arms in the proposed experiment will help us to generate empirical evidence on its competitive advantage over other models.

Business Enterprise vs. Fixed Honorarium: Determinants of Preference

The perceived utility of the business enterprise and fixed honorarium model, to a large extent, will depend on the individual characteristics of the health volunteers. For instance, risk-averse health volunteers might prefer to take the fixed honorarium model and will avoid the burden of managing a business of selling commodities door to door in their communities. Because all of the BRAC health volunteers are women, in the context of rural communities in developing countries, they also have to work at home doing household chores, cooking, taking care of children, water fetching, etc. Therefore, they have limited time to spend on health volunteer works and managing the business. These imply that the preference for the business enterprise model over the fixed honorarium model will take place when:

  1. There is a high probability that the income from the business enterprise model will be higher than the income from fixed honorarium model
  2. The average minimum income from business enterprise will be more or at least equal to the fixed salary offered under fixed honorarium model
  3. The income from the business enterprise will remain constant or will have less variability across the year (in other words, the risk is relatively low).

Within a standard microeconomic expected utility framework, we can express the preference through following equations:

Let,  w = current wealth of health volunteers

W+$10 = current wealth + fixed honorarium (which is $10)

W+ ibe = current wealth + income from business enterprise

Business enterprise model will be preferred if:

 U (w+ ibe) > U (w+10)

Expected utility of income generated from business enterprise (U (w+ibe)) can be increased if the probability of earning more than $10 is much higher than the probability of earning less than $10 or no income from the business model. The best way to increase the probability of earning more than $10 for the volunteers is to spend more time on their health volunteer work.

 An earning of less than $10 from the business model will make the fixed honorarium model more attractive to the health volunteers. Similarly, if the fixed honorarium is further increased and the new honorarium is sufficiently large (which is higher than the average income earned from the business model), then health volunteers will shift from preferring the business model to the fixed honorarium model. Such a shift will increase the operational cost of the program and will risk the long-term sustainability of the program.


Experimenting with different financial incentives for health volunteers and generating evidence on effectiveness of the different models will have important policy implications not only in the public health sector but also in labor economics. Given the resource constraints in developing countries, the success of the business enterprise incentive model is desirable, because this will give a sustainable long-term solution to the problem of incentivizing the health volunteers in rural communities. However, success of this model also depends on individual characteristics of the health volunteers (entrepreneurial skills, risk aversion nature, interpersonal skill, etc.) and the economy of the communities. Though the study has been designed using a cluster randomization method and so addressing the endogeneity problem raised from differences in communities, observing the individual characteristics of the health volunteers will be crucial and should be addressed carefully.

[1] Lehmann, U. and Sanders, D. (2007). Community Health Workers: What do we know about them? World Health Organization2, 1-42.

[2] Nykvist, B., Guariso, M. A., Svensson J. and Yanagizawa-Drott, D. (2014). The Living Goods Model in Uganda: A Cluster-Randomized Controlled Trial. Working paper.

[3] Rowe, A. K., de Savigny, D., Lanata, C. F., & Victoria, C. G. (2005). How can we achieve and maintain high-quality performance of health workers in low-resource settings? The Lancet366(9490), 1026-1035.

[4]  Erika Deserranno, (2015). Financial Incentives as Signals: Experimental Evidence from the Recruitment of Health Workers.

Figure 2: Utility and Wealth from the Incentive Models


, , ,