Need to reform the remuneration system to initiate a system approach to the health sector in resource-poor countries


Corresponding Author Stefan Hanson, Formerly Division of Global Health, IHCAR, Karolinska Institutet, Stockholm, Sweden. E-mail:


Les ressources humaines sont le facteur le plus important pour le fonctionnement des systèmes de santé. Les agents de santé doivent être rémunérés de façon appropriée, mais les systèmes de rémunération actuels des pays à revenus faibles et intermédiaires en Afrique subsaharienne sont souvent injustes. Ils sont dans une large mesure basés sur les allocations et les per diems des programmes spécifiques à des maladies. Ils déforment le fonctionnement de l’ensemble du système et doivent être réformés. Cette réforme doit faire partie d’une évolution vers une approche plus large du système pour le secteur de la santé dans de nombreux pays. Toutefois, un tel changement de politique est combattu par ceux qui ont de puissants intérêts privés dans les systèmes actuels.


Los recursos humanos son el factor más importante para que los sistemas sanitarios funcionen. Los trabajadores sanitarios deben ser remunerados de forma justa, pero los sistemas de remuneración actuales en países con ingresos bajos y medios en África subsahariana a menudo no lo son. En gran medida están basados en un salario y per diems de programas enfermedad-específicos (PEEs). Distorsionan la función del sistema y deberían reformarse. Esta reforma ha sido parte de un cambio hacia un enfoque sistémico más amplio en el sector sanitario de muchos países. Sin embargo, a este cambio de política se resisten aquellos con fuertes intereses en los sistemas actuales.

Human resources are the most important factor for the function of health systems. Health workers need to be fairly remunerated, but current remuneration systems in low- and middle-income countries (LIC/MICs) in sub-Saharan Africa are often unfair. They are to a large extent based on allowances and per diems from disease-specific programmes (DSPs). They distort the function of the entire system and need to be reformed. This reform has to be part of a shift to a wider system approach to the health sector in many countries. However, such policy change is resisted by those with strong vested interests in current systems.

Reasons for policy change

Two important developments during the last decades conduce to policy change. First, financial resources for disease control have increased drastically, much of this from national sources. Second, there is an ongoing shift in the burden of disease from communicable to non-communicable diseases – now the main cause of disease burden in many areas even in sub-Saharan Africa (WHO 2008). Addressing non-communicable chronic diseases, maternal and newborn diseases to a larger extent than control of infections demands a functioning health system. Added to this is the requirement of long-term treatment of individuals with advanced HIV infection.

The epidemiological transition plus the possibilities created by more resources obliges even many poor LIC/MICs to move from focusing on control of selected diseases to establishing a functioning comprehensive health system with better care for common diseases and for individual patients. This is a shift from a public health approach for selected diseases to a more patient-centred approach for general disease problems with maintained but more effective public health functions.

It is probably still true that in many poor countries, a focus on control of selected diseases might reduce the burden of disease more than a comprehensive approach within a poorly functioning health care system, where resources are spread so thin that the system effect is negligible and the main effect is achieved through efficient work in ‘isolated islands’– the DSPs – previously the ‘vertical programmes’ (Hanson 2000). Such has been the thinking on control of mainly communicable diseases for many decades - probably for good reason considering the low resource levels and the poor governance in LIC/MICs (World Bank 2007a). Led by bilateral donors, international agencies and lately philanthropic organisations, this thinking resulted in a proliferation of global initiatives for selected diseases, here called DSPs. Among these initiatives, those directed at the control of HIV have been particularly well financed.

However, the costs of maintaining parallel systems for public services and numerous DSPs are now likely to largely surpass those of one joint system delivering comprehensive care and operating with a unified budget, management and monitoring. A policy change would mean changing from a national system forced to apply multiple donor sub-systems to one national system applied by all donors in the spirit of the Paris declaration. Perhaps, this would be more realistic for regional systems, for example, for the East-African region, considering the expense to donors of using different systems for each country. Current support for control of selected diseases has had major implications for the national systems:

  •  On the positive side, DSPs, particularly those related to HIV, have initiated the increase in funding for health and greatly expanded it.
  •  On the negative side, the proliferation of DSPs has increased the complexity of both aid architecture and national systems and led to fragmented and segmented systems and gradually increasing transaction costs.
  •  Few efforts have been made to establish functioning operational systems able to deal with any disease.
  •  The DSP approach with a focus on selected diseases has led to neglect of general curative care including treatment of both chronically and acutely ill patients.
  •   Utilisation of health services has remained low with few reported visits per capita and many health MDGs are not reached in LIC/MICs in sub-Saharan Africa and parts of Asia (United Nations 2010)

Above a certain resource level, a strategy based on a wider system perspective is likely to be more effective than one based on a more narrow disease control perspective. Many LIC/MICs might now have reached or are about to reach that level. Many sub-Saharan African countries have experienced rapid economic growth during the last decade and more than a third grew by more than 4% between 1995 and 2005 (World Bank 2007b). This has meant that more resources have been allocated to the health sectors.

Resistance to change and implications for DSPs

Donors, however, seems to have great difficulties in shifting from the ‘old’ approach (Wood et al. 2008), perhaps mainly due to vested interests of many stake holders as well as the loss of control it would imply. The current systems, including the remuneration systems, also favour staff at central level, including the policy makers, in the recipient countries. Initiatives at changing the current order, such as the International Health Partnership (IHP+), meet with considerable resistance.

Introducing a clearer system perspective may incur drastic changes to the current order. Resource allocations would have to consider the whole disease burden, including that caused by the chronic diseases not covered by DSPs. DSPs would have to shift from short-term goals to long-term goals and gradual dissolution of DSP structures at least below district level. For monitoring, it would mean a shift to a unified system; for pharmaceutical supplies, it would imply integration. A continued education system would have to be introduced, which would aim at increasing knowledge and improving the quality of services for general curative and preventive care not covered by DSPs. It would also have implications for human resource management including remuneration, which is built on allowances and per diems in much of sub-Saharan Africa and thus may be the main obstacle to nationally driven change.

The current remuneration system and its consequences

Allowances can be divided into so-called duty-enhancing allowances, such as training, and remunerative allowances, such as expenses for housing. The current remuneration system, particularly for central level government staff, to a large extent builds on workshop allowances and per diems – both particularly attractive as not taxed. ‘Very often these practices have dramatic impacts on the health care system’ as the ‘the players plan their action around the primary goal of acquiring per diems’ (Ridde 2010).

This form of remuneration is difficult to justify in any system, whether selective or comprehensive, as per diems for workshop attendance probably mostly only result in ‘negligible tangible outputs’ as revealed by an organisational tracking exercise in East Africa (Hepworth 2009). It would be regarded as particularly distortive if seen in a wider system perspective. The effects include:

  •  shift of focus from routine work to workshop attendance
  •  shift of focus to prevention away from curative care, which offers few workshops
  •  shifting resources from the periphery to the central level
  •  taking away the initiative from the poorly financed routine system to the well-financed DSPs through which allowances can be paid out

In a system approach, it is even more important that staff at central level are available in their offices to oversee the function of the whole system than that they attend workshops to improve operations of specific diseases. However, although it is difficult to find any proof, it is not only obvious from anecdotal accounts that staff in ministries are difficult to find in their offices as they spend so much time in workshops, but also economically understandable. They do what the system encourages them to do.

Staffs at central level are those that have the best access to workshops. According to a recent report from Tanzania, the ‘bulk of allowances (70%) accrues to the 30% of the civil servants working at Central Government’, and ‘the average civil servant in one of the Ministries of Central Government receives more than five times as much in allowances as the civil servant delivering services at community level’ (Policy Forum 2009).

Despite these obvious bad effects, the remuneration of government staff, at least in some countries, continues to tend to shift more and more towards allowances and per diems. Thus, in Tanzania’s budget 2009–2010, the ‘amount allocated to allowances was equal to 59% of the total wage bill’ and the actual amount spent on allowances more than tripled between 2001/2 and 2006/7 (Policy Forum 2009). In many settings, attending 3–4 days of workshops equals a month’s income. In Burkina Faso, health workers’ income from per diems was higher than their salaries (Ensor et al. 2006). As allowances were only vaguely related to costs incurred, it meant they had ‘become an incentive instead of a reimbursement’ and not linked to actions that improve service. Similar remuneration systems are used in India (B. Forsberg, personal communication).

Maintaining the system has also been in the interest of donors as workshop attendance and is one of the key measures of capacity building, a main activity of donor assistance (de Graaf 2009).

The ‘allowances culture’ not only permeates the government system. It has also been adopted by NGOs as a main source of income. It reaches out to the rest of society, making it impossible to gather villagers or village committees for meetings without paying allowances. The allowances’ culture approaches the corruption that permeates many societies (Chêne 2009).

However, contrary to all criticism, the system has also been defended for ‘mobilizing additional resources retaining qualified staff’ in the government services, who would otherwise have left for the private sector or international organisations (Conteh & Kingori 2010).


Countries with a decent level of governance that have experienced rapid growth have strong reasons to apply a system approach to the health sector, as performed in Rwanda, for example, (Logie et al. 2008). The initiative for such a change could either come from strong national leadership or from donors, although these so far have tended to be conservative. It ought to be the responsibility of the World Bank and has been taken up by the EC (Unger 2008). The subject should be discussed both in recipient and donor country parliaments. It has to include a change of the remuneration system as one of the initial steps.

In a wider system perspective, there is no room for remuneration based on allowances and per diems and not directly related to costs incurred. Regular, more fairly distributed salaries are needed, possibly combined with bonuses based on mid- and long-term performance (Soeters et al. 2006). Such an approach is likely to lead to a more effective system serving more people, reducing the burden of disease and reaching the MDGs – a shift from donor-dominated approach for specific diseases to a more patient-centred national system with more effective public health functions.