Bulletin of the World Health Organization

Health worker remuneration in WHO Member States

P Hernandez-Peña a, JP Poullier b, CJM Van Mosseveld c, N Van de Maele a, V Cherilova a, C Indikadahena a, G Lie a, T Tan-Torres a & David B Evans a

a. World Health Organization, avenue Appia 20, 1211 Geneva 27, Switzerland.
b. Geneva, Switzerland.
c. Statistics Netherlands, The Hague, Netherlands.

Correspondence to P Hernandez-Peña (e-mail: hernandezp@who.int).

(Submitted: 10 March 2013 – Revised version received: 05 July 2013 – Accepted: 08 July 2013.)

Bulletin of the World Health Organization 2013;91:808-815. doi: http://dx.doi.org/10.2471/BLT.13.120840

Introduction

A sufficient number of motivated health workers of different types is a precondition for achieving universal health coverage (UHC).1 The world health report 2006 refocused the world’s attention on the acute shortage of health workers in many countries – a persistent problem – but it also made clear that governments need information on the costs of training, hiring, deploying and remunerating health workers when developing plans to improve the availability, distribution, capacity and performance of their health workforces.17 Unfortunately, data of this sort are only available for selected countries.8,9 According to a report from 48 African countries developed in the mid-1990s, the proportion of total health expenditure spent on health workers ranged from 17% to 74% and was 46% on average.10 The World Health Organization (WHO) subsequently estimated the share of government expenditure devoted to paying health workers in 64 countries as ranging from 10 to 90% and being 42% on average.7 This variability reflects differences across countries in accounting procedures, wages and the price of other health inputs, such as medicines.11

Despite the shortage of data, the information that exists has been used to estimate resource gaps and implement policies affecting the health workforce. For example, in the 64 countries identified as having the greatest health worker shortages, the yearly cost of training and employing an additional 3.5 million doctors, nurses and midwives and 2.8 million other workers was estimated at 17 800 million United States dollars (US$) in 2006.3,4 If salaries were increased to try to retain the newly trained health workers, the costs would be substantially higher.

The data supporting the estimates just described pertain to a limited number of countries and only to medical and paramedical workers, since remuneration data for other types of health workers are not available.12 Even for these categories, comparisons across studies and countries are made difficult by variations in payment modalities, statistical practices and in the definition of the different types of workers and their respective roles.13 To obtain a complete picture, it makes sense to first identify and assess the data available from as many countries as possible using standardized definitions and accounting practices to facilitate comparisons.14 This paper, based on the data available for 2000 and 2010, reports on the first attempt to do so for all 194 WHO Member States.

Methods

We searched for data on payments made to different types of health workers by conducting a documentary review and directly contacting officials from all WHO Member States. These officials included WHO focal points in charge of health accounts – appointed by health ministries – and staff in statistics offices responsible for national accounts reports. The data reported here follow the international standards developed for the entire health workforce,14 which is defined as “all people engaged in actions whose primary intent is to enhance health”.7 This definition embraces anyone working towards promoting, restoring or maintaining health, including people in management and support jobs that are essential for health systems to function (e.g. health insurance personnel). We also report on the remuneration of government health workers.

General government health expenditure comprises all current and capital spending earmarked for the maintenance, restoration or enhancement of the health status of the population. Total health expenditure comprises all expenditure earmarked for the maintenance, restoration or enhancement of the health status of the population, regardless of the outcome of the goods and services consumed, and including public and private spending.15 We used general government expenditure on health and total health expenditure as the denominators for our calculations.

Health worker remuneration is frequently reported as the following two variables or indicators, though not by all countries:

  • Total remuneration of all salaried workers in the health system, which refers to wages and salaries (including benefits and allowances) and to social contributions paid on behalf of workers involved in providing health services. A subcomponent is the remuneration paid by government to its employees.
  • Remuneration of independent practitioners (e.g. self-employed health workers such as physicians and physiotherapists), which comprises their business income net of operating costs, taxes and capital consumption. In national accounts this is called “mixed income”.16

We break down these broad categories into:

  • total remuneration of salaried health workers, as a share of total health expenditure and gross domestic product (GDP);
  • remuneration of salaried health workers paid by the government, as a share of general government health expenditure, total health expenditure and GDP;
  • total remuneration of independent health practitioners, as a share of total health expenditure and GDP.

Data were collated in millions of national currency units per calendar year and are reported by World Bank country income groups for 2013.17

We also assessed changes in remuneration over time and compared data for health workers with data for workers in the whole economy. Only 62 countries had data available for 2000 and 2010 for both health workers and workers in the total economy, so we supplemented this with data from 43 countries having one data point close to 2000 (up to 2003) and another close to 2010 (2008–2011). To allow for differences in the number of years between data points, we report the average yearly change for the case of health workers (i.e. the compound annual growth rate) and we report the number of countries covered in the analysis in each table.

Data sources

General government health expenditure and total health expenditure are published annually by WHO in its Global Health Expenditure Database (GHED) for all its Member States after country data searches and consultations. This was our main source of data for denominators and remunerations.18 We obtained the remuneration given to salaried health workers in each sector (in total and by governments) and to all salaried workers in the economy, including all industries combined (in total and by governments), from reports on government finance, budget records and national accounts, labour and general government accounts and health accounts. These data are published by ministries of finance and health, statistical offices, central banks, health insurance entities and international organizations such as the International Monetary Fund (government finance statistics and country reports), the United Nations, the Organisation for Economic Co-operation and Development, the Statistical Office of the European Union and the World Bank (public expenditure reviews). Statistical yearbooks, institutional annual reports and national web sites were supplementary sources of information. The specific data on health worker remuneration and the source from which we obtained each data point are available on the GHED, as are the figures for each country separately.

Data caveats

Most of the data on the remuneration of salaried health workers pertains to direct health care provision. Payments to other types of health workers – e.g. to people who dispense medicines and lenses in pharmacies and other retail units or who work in the administration of health insurance – are not routinely reported. In addition, outsourced services are reported as service purchases without any breakdown of how much is paid to health workers.

Government health workers can also be paid through external funds or through special budgetary arrangements that are not systematically accessed, such as through the armed forces. Some public hospitals are reported as corporations and aggregated with private entities, which reduces the amount reported as having been paid by the government. The available metadata are not always clear about what they comprise but, if anything, our numbers probably underestimate the payments made to health workers.

Results

Remuneration of salaried health workers

The remuneration of all salaried health workers – in the public and private sectors combined – is available for only 136 countries and averages 33.6% of total health expenditure (Table 1). The average remuneration increases the higher the country income group: it is 38.1% and 28.7% of total health expenditure in the highest and lowest country income groups, respectively.

Data for independent practitioner remuneration are even more scarce (n = 89). The average remuneration is 9% of total health expenditure. It varies enormously across countries – from negligible to around 50% of total health expenditure – and this is one reason that the differences observed by country income group are not statistically significant.

More countries report the remuneration paid by government to salaried health workers (n = 179). Such remuneration accounts for almost 20% of total health expenditure and for 33.2% of total government health expenditure, on average (Table 2). The shares in low-income countries are significantly lower than in the higher-income countries combined.

Remuneration of all health workers

Total remuneration of all health workers is the sum of the remuneration of salaried and independent health workers. The 33.6% of total health expenditure comprised by remuneration of salaried health workers cannot strictly be added to the 8.9% pertaining to independent practitioners (Table 1) because the data are not available for the same countries. If we include only those countries for which data for the two components are available (n = 75), the total remuneration of all health workers accounts for 34.5% of total health expenditure, on average (Table 3).

Changes over time

We first focused on the 106 countries for which data on payments to salaried health workers were available for the period (Table 4). From 2000 to 2010, total health expenditure grew faster than the GDP; it increased as a share of GDP at an annual rate of 1.5%. Total payments to salaried health workers also grew but at a slower rate than total health expenditure, so the share of total health expenditure paid to these health workers decreased in country income groups (at an average rate of 0.5% annually). However, because total health expenditure was increasing as a share of GDP, expenditure on salaried health workers also increased as a share of GDP at an average annual rate of almost 1%, despite a more modest increase in upper-middle-income countries.

Salaried worker remuneration

The period analysed includes the first years of the recent financial crisis. The health wage bill for salaried workers continued growing more rapidly than the total economy (Table 4). On the other hand, the remuneration of all types of salaried workers did not change substantially over the period as a share of GDP (the average annual growth rate was 0.03%, i.e. not significantly different from zero).

Government expenditure on health workers

Government health expenditures grew more rapidly than both GDP and total health expenditure over the period, at an average annual rate of 2% (Table 5). Government expenditure on salaried health workers also increased as a share of GDP, but less rapidly than total expenditure on salaried workers, as shown in Table 4. In fact, payments to salaried health workers fell as a share of government health expenditure by over 1.3% annually and did so more rapidly in upper-middle-income countries.

Government payments to all types of salaried workers, however, increased at a faster rate than payments to health workers in all but the high-income countries.

Payments to independent practitioners versus salaried workers

Data on payments to both salaried and independent workers, both in the health sector and in the entire economy, were available for only 75 countries for 2000 and 2010. Payments to both types of health workers increased as a share of GDP: for salaried health workers at over 1.2% annually and for independent health workers at just below 1% (Table 3). The growth rate of payments to independent health workers showed more variability across country income groups, but the small number of countries in some of the groups resulted in differences that are not statistically significant. Total remuneration to independent health workers increased faster than remuneration to other types of non-salaried workers. In fact, the share of GDP comprised by the remuneration of independent workers in all sectors fell between 2000 and 2010.

Average level of remuneration

Total expenditure on health worker remuneration is a function of the number of workers and how much each worker gets paid. To get an idea of the average payment made to each worker, we divided total expenditure on remunerations by the reported number of salaried workers for the 92 countries with available data. The global average in 2010 is US$ 15 352 (18 102 in purchasing power parities) per employee. We recognize the limitations of this estimate. For example, the incomes of different types of health workers vary considerably, so the differences across country groups in Table 6 also reflect differences in the mix of health workers. However, the variations across country income groups are as expected – health worker payments increase with country income group – and this gives face validity to the data presented earlier.

Discussion

Appropriate health financing schemes are needed for progress towards UHC. Also essential are systems capable of delivering high-quality health services covering health promotion, disease prevention and treatment, and rehabilitation and palliation. These services need to be accessible and affordable,19 but this is not possible without sufficient health workers possessing the right skills and located close to the people who need them.

Health workers need to be paid and questions of how much remuneration will motivate them to provide good services, stay in rural areas and not emigrate, for example, have long exercised the attention of policy-makers.2 At a higher level, the critical questions are how much money has been raised to pay the health workforce and at what rate this amount has changed over time. These are the main questions addressed in this paper, which provides critical information not previously available but crucially important for planning the changes needed to attain UHC.

The data herein presented, which resulted from a laborious effort to obtain all the information available from WHO Member States on health worker remuneration, point to several important findings. First, payments to health workers comprise an important component of total health expenditure and GDP. Payments to salaried and self-employed health workers combined accounted, on average, for over 34% of total health expenditure and around 2.5% of GDP in the 75 countries for which both sources of data were available in 2010. In the 136 countries with information on salaried health workers, their remuneration alone accounted for 33.6% of total health expenditure. This suggests that for all health workers combined, including those who are self-employed, the share of total health expenditure surpasses 34%. Only pharmaceuticals account for a higher share of total health expenditure.20 On the other hand, our estimates are lower than those from earlier studies based on smaller samples of countries. For instance, Peters reported a share of 46% of total health expenditure, but those estimates were made more than 15 years ago, when health systems were less complex than now.10

Second, these expenditures can be expected to increase over time and as a share of GDP. This has occurred over the past 10 years, largely because the overall number of health workers has increased everywhere. Interestingly, payments to health workers have increased more slowly than other types of health expenditure, which suggests that the health sector is becoming more reliant on technology or more capital-intensive everywhere.

Our findings have major implications with regard to the type of strategies that would accelerate progress towards UHC. In most low-income countries, funding for health is simply not enough to allow all people to access even a minimum set of needed health services.21 Considerably more will need to be spent on the health workforce for this to change, but the path to UHC also requires investment in other components of the health system, including medicines, infrastructure and information systems. The dilemma for governments is how much of the funding, which is currently insufficient, should be devoted to its health workforce.

Third, payments to salaried health workers have increased more rapidly than payments to other salaried workers. Although the data do not allow us to determine if this is the result of changes in the number of workers or in their average remuneration, expenditure on the health workforce seems to have been protected despite the recent financial crisis.

These findings are important, but complete and accurate data are not yet available for all countries. Although expenditure on the total health workforce is frequently found as a line item in government budgets and national accounts reports, coverage is incomplete and more effort is required to build low-income countries’ capacity to track and report the remuneration of various types of health workers.

In addition, identifying the flows linked to financial incentives, delivered sometimes as cash payments and sometimes in kind, is fraught with difficulty. This means that the data reported here might be underestimating true expenditures. Other possible causes of underestimation include the inability to identify the component of outsourced services paid to health workers from existing accounts, and the exclusion of retail sellers of medical supplies and health insurance administrators from health expenditure records.4,22

The remuneration of independent health practitioners, especially in countries with large private sectors, warrants particular attention. How much is paid to independently employed workers relative to salaried workers reflects the way labour markets are organized in different settings. In many countries independent practitioners are self-employed but receive the bulk of their income from social health insurance or government payments, in much the same way that salaried workers do. Fortunately, the new system of health accounts, published in 2011, includes a component for monitoring factors of provision, and this could be expanded to ensure uniform reporting of expenditure on various types of health workers over time and across countries.15

In summary, this study confirms, on the basis of data from many more countries than previously available, that payments to health workers account for a substantial share of total health expenditure. However, this share has been decreasing over time. It also shows that payments to salaried health workers have increased as a share of GDP, while those to other types of workers have remained stable or fallen.


Competing interests:

None declared.

References

Share