3. Use of skills intelligence in health workforce policymaking

This chapter explores how governments and social partners use the information gathered from skills anticipation exercises relating to the health workforce, including those identified in Chapter 2. Common policy uses are to inform education, employment and migration policy. Exercises are often designed with particular policy uses in mind. Quantitative outputs at occupation or qualification level are often used to determine student intake in health education programmes or migrant inflows. Qualitative findings that describe the types of skills that a given occupation will require can be used to define education and training course content and to inform changes to the way tasks are allocated across occupations.

For skills anticipation exercises to be effective in addressing skills gaps in the health workforce, they must successfully translate into policy. However, in many of the countries studied, challenges were identified in achieving the policy purpose of the findings. An important barrier preventing the information obtained from skills needs anticipation exercises to be used more extensively in policymaking in the health sector is that scope of practice regulation limits the tasks a person in a given occupation is legally allowed to perform. Other important barriers include lack of funding, coordination, stakeholder involvement, and poor alignment between the skills intelligence and the desired policy purpose. These barriers may be especially challenging to overcome in LMICs.

As with the previous chapter, this chapter draws primarily from semi-structured interviews with representatives from institutions that carry out or use anticipation exercises in Australia, Argentina, Bangladesh, Canada, Colombia, Ethiopia, Finland, Germany, Ghana, Ireland, Republic of Korea, the Netherlands, Norway, South Africa, Sweden and the United States. It also draws from participant contributions during a virtual peer-learning workshop attended by stakeholders from 20 countries who are involved in either producing skills information in the health workforce or using it for policymaking.

Skills intelligence related to the health workforce is used to inform education and training policy, employment policy and migration policy. The social partners also make use of this information in the context of collective bargaining and informing training content.

One of the primary objectives for gathering information about future skill needs in the health workforce is to provide guidance for education and training policy decisions. Specifically, this information is used to inform how many students should be admitted to health education programmes. The information is also used to design and update the health education curricula and to inform career guidance.

Several countries in this study (including Finland, the Netherlands, Norway, South Africa and Sweden) use quantitative forecasts to determine the maximum number of students that are allowed to enrol in health education programmes (“numerus clausus”). This type of workforce planning is particularly important in the health workforce, given the time and costs involved in training new health workers. Previous research shows that nearly all countries impose a numerus clausus on education programmes for doctors, nurses and other health professions (Ono, Lafortune and Schoenstein, 2013[1]).

The Advisory Committee on Medical Manpower Planning (ACMMP) in the Netherlands uses the different scenarios of their quantitative forecasts to determine the minimum and maximum recommended intake into health education programmes. In addition to providing a bandwidth recommendation, ACMMP also provides education institutions with a recommended intake size, such as at the bottom, middle or upper bound of the recommended bandwidth. Some of the factors that feed into the decision for the recommended intake size are the size and duration of the shortage, and the speed at which the ACMMP recommends eliminating those shortages. For example, since specialists for intellectual disabilities and geriatric medicine experienced large and increasing shortages over several years, the 2020 ACMMP report included a recommendation for a relatively large intake into education programmes for these specialisations, so that shortages would be eliminated in 12 years, instead of the average 18 years for other specialisations (ACMMP, 2020[2]). The Allocation Decree of the Dutch Ministry of Health, Welfare and Sport decides on a numerus clausus for health education programmes each year based on the ACMMP's recommendations.

In South Africa, current to medium-term skills needs are determined by the Sector Education and Training Authority (HWSETA) through a sector skills plan. The Department of Higher Education and Training uses the sector skills plan to set student intake in order to improve the responsiveness of the post-school education and training (PSET) system. They do this in coordination with Department of Employment and Labour (DEL).

Although numeri clausi are a common use of skills intelligence in the health workforce, their effectiveness in mitigating shortages (or surpluses) depends on the quality of the skills intelligence, and on the uncertainty of the context in which they are implemented. For instance, in South Africa, the data for sector skills plans comes from employers, monitored through the Annual Training Report. However, many employers do not comply, compromising the data quality. There is also a risk of “yo-yo” effects when policymakers over-correct the numerus clausus in response to fluctuations in perceived or real shortages or surpluses (Ono, Lafortune and Schoenstein, 2013[1]). Large rises or falls in student numbers in health education programmes can create adjustment problems for education and training providers, by changing the required number of teachers, classrooms and training places required from one year to another. In LMICs, even when the need for more places in particular health education programmes has been identified, there may be issues in absorbing trained workers into the health care system. For instance, in Ghana, respondents reported that the government could not always afford to hire trained health professionals, even when shortages had been identified and workers trained.

While determining student numbers in health education and training programmes relies principally on quantitative data such as occupational forecasts, defining the content of programmes typically involves more qualitative data and approaches. The development of education and training programmes and relevant curricula can be informed by tools such as the WHO global competency framework for universal health coverage (Box 3.1), which provides guidance for creating competency-based education outcomes for health workers. Exercises that focus on skills needs, rather than workforce needs, are useful for helping to update curricula, since they allow for a focus on developing the composite skills required for a particular profession. For example, Germany’s Federal Institute for Vocational Education and Training (BIBB) is using qualitative methods to investigate how the continuing education and training (CET) curricula for nurses should be adapted following recent reforms and putting a particular emphasis on the skills needed to establish quality outcomes in the nursing profession. They set up an expert panel composed of field experts, education providers, employers, and researchers to develop the new curriculum. Similarly, the explicit purpose of Australia’s industry reference committees is to update competency frameworks (called “training packages”) for health occupations that require vocational education and training. Training packages are used to update course content for qualifications associated with those occupations.

Among the skills anticipation exercises identified in this study, there is a noticeable gap in terms of analysis that sheds light on the impact of digitalisation on the skills needed by health workers in order to inform curriculum content. The exception was a project carried out by the Michener Institute in Canada, a post-secondary institution embedded in a hospital. Funded by the Future Skills Centre and relying heavily on stakeholder consultation, the project sought to understand the readiness of health professionals to use new AI tools, and the skills that would be needed to use these tools. The objective of the project was to develop a new national certification programme in basic AI literacy for health workers, after testing the curriculum with 5 000 health workers. The project focused on various professions in the health care sector, recognising that different roles will be transformed by AI in different ways, depending on their reliance on personal interactions, data, repetitive tasks, and creativity.

Owing to factors such as limited resources and slower implementation of new technologies, forecasting of the impact of digitalisation on the health sector was limited in the LMICs studied. Nevertheless, several LMICs noted the importance of digital skills for the health sector. In Ghana, for instance, qualitative data gathered from hospitals indicated a need for improved information technology skills among workers and these findings fed into training programmes delivered at public institutions. However, the skills assessment approach was reactive, and did not attempt to forecast future skills needs. In Ethiopia, while some stakeholders recognised that digital skills were essential to accelerate digital transformation, they also noted that Ethiopia had a major challenge in finding highly skilled workers. They called for a greater focus on training a larger percentage of the general population in higher-level digital skills.

Several countries reported using skills intelligence to inform career guidance websites or other career guidance services. The forecasts conducted by the Netherlands’ Research Centre for Education and the Labour Market (ROA) are used to inform various career guidance websites, including those aimed at secondary school students. A field experiment showed that informing pre-vocational secondary education students about the job opportunities in their occupations of interest – based on ROA’s quantitative forecast results as well as expected hourly wages in those occupations – increases their probability of changing their preferences towards an occupation with better labour market perspectives (De Koning, Dur and Fouarge, 2022[4]). South Africa’s Health and Welfare Sector Education and Training Authority (HWSETA), which conducts skills anticipation and assessment exercises in the country, launched a digital career guidance portal in 2021, as face-to-face interaction was constrained by the COVID-19 pandemic. The objective of the portal is to address skills needs in the health care sector, combat youth unemployment and poverty, and inform people about the opportunities in the sector, including funding support for trainees, In Colombia, the Ministries of Health, Labour and Education all reported using the results of skills anticipation studies to inform prospective students about employment prospects in the health sector through career counselling and/or vocational guidance.  

In order to inform career guidance, it is useful to conduct both national and sub-national demand-side labour forecasts for the health workforce. National level forecasts provide skills intelligence on the demand for specific skills and/or health professionals in a country, and can help address shortages at a national level. However, in many countries, skills shortages occur at a sub-national level, and many trained workers look for jobs within their region. Sub-national forecasts are useful in this context and could also help to address regional shortages, or urban/rural divides in health care provision as identified in some countries studied.

Skills intelligence in the health workforce can feed into the development of employment policies. It can inform how work is organised in the health sector, the updating of occupational standards, or the development or updating of on-the-job training, retraining and upskilling courses.

Skills intelligence can be used to improve work organisation in the health workforce, by optimising the allocation of tasks between occupations, or the mix of healthcare staff in an establishment at a given time. A focus on these two areas, often broadly referred to collectively as the “skill mix”, can help improve health system performance, address shortages and build more resilient healthcare systems (Buchan and Calman, 2004[5]). For the purpose of this report, a distinction is made between interventions to change the allocation of tasks between occupation (referred to here as “task allocation”) and those to adjust the mix of staff in the health workforce (referred to here as “skill mix”). An example of task allocation is reviewing scope of practice legislation to allow other workers in other health occupations to do some of the tasks that specialised medical doctors do. Respondents in several countries highlighted that specialised medical doctors are spending a significant amount of time doing administrative tasks, and some countries are therefore exploring how to shift these administrative tasks to other less in-demand health occupations to help mitigate shortages and make better use of doctors’ skills. An example of skill mix is forecasting future demand for health services in order to project how many doctors, nurses and other types of health workers will be needed to meet demand.

One of the mandated objectives of the Norwegian Health Workforce Commission is to investigate if there are possibilities for different groups of health workers to start sharing tasks, or to change the way in which they are currently sharing tasks. This exercise is based on an analysis of task and skill similarity across health professions, but also on an assessment of the potential impact of innovation and digitalisation on how certain tasks will be performed. They will use this information to estimate the number of different groups of health workers that are needed in the next 20 years.

In some countries, skills intelligence in the health workforce is used to help determine the definitions of tasks and duties that health workers are expected to carry out, and the skills, knowledge and behaviours they must exhibit to work safely and effectively in their occupation. For instance, skills intelligence that highlights the increasing use of digital tools and AI in the workplace could be used to update the occupational standards for health workers to include digital skills. This can in turn feed into the development and updating of curricula, qualifications and scopes of practice to help address both current and future skills needs.

In Australia, the primary use of Industry Skills Forecasts in the health sector is to update the training packages for vocational qualifications in health. Vocational education and training is based on occupational skills standards which are set out in units of competency within training packages and accredited courses. Training providers have to use this training package in the training they provide. For instance, the skills anticipation exercises showed an increasing demand in the aged care sector for dementia training and palliative care, which was then incorporated into the training package for this sector.

Countries with qualifications-driven migration policies, like Australia, Canada and South Africa, can use skills intelligence to select migrants with skills, qualifications and work experience that are in high demand, and that are difficult to source among the domestic workforce.

In Australia, the Commonwealth and state governments prepare annual skilled occupation lists that are informed by skills intelligence. Prospective migrants with qualifications or work experience related to one of the occupations on these lists will be prioritized for work visas (OECD, 2018[6]). One of these lists, the Medium and Long-Term Skilled Shortage List (MLTSSL), is forward-looking1 and identifies occupations needed to meet medium-term projected skills shortages, as well as occupations required in the longer-term to build productive capacity in the economy. The MLTSSL is used to select migrants for all permanent skilled visas, as well as the medium-term Temporary Skill Shortage visa. Of the 212 occupations on the list, 60 are health-related occupations. As part of the annual review of the skilled occupation lists, the National Skills Commission considers various sources of data, including Industry Skills Forecasts, as well as occupation-level data on skilled migrant employment outcomes, reliance on temporary visa holders, rates of over- or under-qualification, low visa grants and projected employment growth. Using a points-system, they determine which occupations should be added to the skilled occupation lists and which should be removed or moved to a different list. Stakeholders are invited to provide feedback (OECD, 2018[6]).

In Canada, the provinces use the findings from ESDC’s occupational projections to negotiate with the federal government about how many migrants with a given type of skill or qualification to admit. While the primary purpose of the Netherlands’ ACMMP exercise (Capaciteitsplan) is to determine intake to health education programmes, the findings are also used to decide whether foreign health workers are needed if relying on domestic supply is unlikely to meet increasing demand.

In South Africa, skills intelligence that is developed through the Health and Welfare Sector Education and Training Authority is used by the Department of Higher Education and Training to develop critical skills lists on behalf of the Department of Home Affairs. The critical skills lists are developed every two years and are used to shape migration policies and facilitate the hiring of foreign nationals to fill short- to medium- term skills gaps. Nevertheless, respondents suggested that the effectiveness of critical skills lists were limited by unreliable data. Employers reported that skills that were in critical need at a hospital level were not reflected on these lists. This underscores the need for robust data and effective collaboration between stakeholders in translating skills intelligence into migration policy.

For other LMICs studied, concerns over migration in the health sector were primarily related to outward flows and issues of brain drain. Lower income countries face severe skills shortages in the healthcare sector and have limited resources and capacity to train new workers. When public healthcare systems invest in training new workers, only for them to move abroad, this represents a loss when finances are already severely limited. Some respondents reported that by lowering visa regulations in developed countries due to shortages in the healthcare sector, the COVID-19 crisis had contributed to an increase in outward migration of healthcare workers to these countries. In Ghana, respondents stated the need for international cooperation in order to stem the brain drain from the country. A bilateral agreement between the Government of Germany and the Government of the Philippines, to facilitate the placement of Philippine nurses and other healthcare workers in the German healthcare sector while minimising the effects of brain drain – was highlighted as a potential model for finding sustainable solutions for migration in both LMICs and higher income countries. Similar bilateral agreements have been implemented in other countries2.

This study revealed that trade unions and representative organisations of employers use the information generated by skills anticipation exercises for a variety of purposes.

Trade unions representing health workers use the information gathered by skills anticipation exercises in collective bargaining, providing scholarships and informing in-house training. In Argentina, the trade union that represents health workers (Federación de Asociaciones de Trabajadores de la Sanidad Argentina, FATSA) advocates for better working conditions and benefits for health workers in light of high current and projected job vacancy rates in health occupations. FATSA also shares information with the Ministry of Education, to request scholarships to prospective students in high-demand occupations, including nursing. They also interact with the Ministry of Health and the Ministry of Labour to request training for unemployed people to work as nurse assistants for elderly care. FATSA also uses this information to inform their in-house training that is provided through their own educational institutes, university hospitals and agreements with tertiary education institutions. The information determines the courses they offer, and the scholarships they provide. SEIU Healthcare Pennsylvania, a regional trade union for health workers in the United States, also uses skills intelligence to develop and update their in-house training.

The European Union of Private Hospitals (UEHP) is an example of an employers’ association which uses skills information in trying to address common challenges private hospitals face in recruiting and retaining healthcare professionals. Member hospitals face severe shortages of health workers, and UEHP makes regular requests to governments to increase intake in health education programmes, citing findings from skills anticipation exercises. They also use this information to motivate hospitals to improve the quality of work and to promote a good work-life balance in an attempt to retain staff.

Interviews with representative organisations of employers and workers in LMICs revealed a number of ways in which social partners were involved in both generating and using skills intelligence in the health sector. In South Africa, employers report to the Health and Welfare Sector Education and Training Authority, which develops sector skills plans to anticipate medium-term skills needs in the health sector. Furthermore, the Public-Private Growth Initiative and Hospital Association of South Africa worked in collaboration with the Department of Health to conduct gap analysis on the supply and demand of nurses up to 2030, to inform both national and private policy. Employers and workers also use the results of sectoral skills plans in collective bargaining. For instance, the National Council of Trade Unions in South Africa uses the information to negotiate on working conditions (including national health insurance and minimum wages) through the National Economic Development and Labour Council. Critical skills lists produced by the government based on industry surveys are essential for private employers to attract the skilled workers they need, though some employers report they do not accurately reflect needs at the sectoral level. In Ghana, the Health Services Workers’ Union of the Trade Union Congress worked in cooperation with trade union partners in Norway to identify a gap of skilled labour in geriatric care and were engaged in training of trainer programmes to improve healthcare provision to the elderly.

There are important barriers to making full and effective use of skills intelligence related to the health workforce for policy purposes. Scopes of practice may limit the use of skills intelligence for task reallocation and skill mix. Other barriers include lack of funding, a lack of political will to increase public spending on health and lack of coordination between the different ministries that are involved with developing a policy response.

One of the key barriers to achieving the full and effective use of skills intelligence for informing healthcare policy that was identified in this study are scopes of practice for health professions. Scopes of practice are the actions and processes that healthcare practitioners are permitted to undertake in keeping with the terms of their occupational license. Scope of practice laws apply to all healthcare practitioners that require a license to operate. By restricting which tasks people working in certain occupations are legally permitted to undertake, scopes of practice help to ensure quality care provision. At the same time, they can also limit the possibility for optimal task reallocation between occupations. Scopes of practice laws may be inflexible or not readily adaptable to the changing skills needs of the sector. As a result, they can pose a significant barrier to full and effective use of skills intelligence for task reallocation and for achieving the right skill mix in the healthcare sector.

In many LMICs, public funding is insufficient to train and/or hire the number of healthcare workers that are required in order to meet current and future skills needs. For instance, in Ghana, where the majority of healthcare provision is provided by the public sector, trained healthcare workers may have to wait years to be hired, due to insufficient funding. Some workers leave the sector through lack of opportunity. Meanwhile, the critical lack of funding means that the majority of resources go towards the hiring of workers, and there are fewer resources available for the development of training programmes.

Another key funding barrier that several stakeholders raised is low political will to increase public spending on health. Healthcare provision already represents a significant share of country’s economies – 8.8% of GDP on average in the OECD and 9.1% of GDP in South Africa, compared to typically lower but still significant shares in lower income countries (such as 3.2% in Ethiopia and 2.5% in Bangladesh). There can be reluctance to increase funding in the sector, which limits governments’ ability to address shortages by investing in attracting and training more skilled health workers. In some cases, countries or regions are reluctant to publicly report health workforce shortages because they lack the public funding to attract or train more health workers.

Low political will to increase public spending on health may also be related to the fact that returns on investments to train additional healthcare workers can take 10 to 15 years to materialise, which often extends beyond political mandates. Furthermore, once skills anticipation exercises have been conducted, it can take several years before recommendations based on their results (such as increased intake into health education programmes) are implemented in practice. Such delays create a barrier to swift and effective policy responses and also mean that findings may be outdated or less relevant once policy responses are implemented.

Some countries mentioned a lack of coordination between the different ministries that are involved in the policy response as a barrier to translating skills intelligence to policy. In many countries, the responsibilities for labour market and education policies in the health sector are shared between the ministries of health, education and employment. For instance, in the Netherlands, the Ministry of Health is in charge of university-level education for health workers, while the Ministry of Education is in charge of upper-secondary education for health workers as well as health education in universities of applied sciences. Skills anticipation exercises may be carried out by one of these ministries, but require coordination between all three for effective policy making. Problems can arise both in terms of coming to consensus about what the skill needs in the health workforce are and agreeing on an optimal policy response. In Ethiopia, for instance, some respondents highlighted a lack of collaboration between the Ministry of Health and Ministry of Education on updating the curriculum and applying consistent approaches to qualification and assessment. This led to poor outcomes in the delivery of health education.

Previous research has shown that a lack of consultation with stakeholders and experts in identifying skill needs is one of the key barriers to translating skills intelligence into policy (OECD, 2016[7]). While developing stakeholder relations may take time, excluding social partners from exercises to anticipate skill needs or discussing and validating the results before publication may slow policy implementation. In this study, countries that took time to build strong stakeholder involvement into the development or validation of their skills anticipation exercises (particularly Finland and the Netherlands) emphasised how this approach resulted in more effective and swifter policy responses.

Full and effective use of skills intelligence is sometimes limited because the findings of skills anticipation studies are not easy to understand, are not fit for the policy purpose or are not considered credible by policymakers. Previous research has shown that stakeholders often perceive the output of skills anticipation exercises as too technical or not sufficiently disaggregated and that the way skills are measured and defined do not map to useful variables in policymaking (ILO, 2017[8]) (OECD, 2016[7]). Strong stakeholder involvement in the development of skills intelligence may help prevent or overcome these potential barriers.

Poor reliability of data can also limit acceptance of the findings of skills anticipation studies, and willingness to translate them into policy. In all LMICs studied, the reliability and availability of accurate labour market data in the health workforce was cited as a major barrier to carrying out accurate skills anticipation and assessment exercises. For instance, a respondent from South Africa’s Health and Welfare Sector Education and Training Authority cited data as the single biggest barrier to translating skill assessment and anticipation exercises into policy and practice. Reliability of data also goes hand in hand with funding. Many LMICs lack the resources to improve data collection systems and are therefore forced to rely on less reliable data. For this reason, many opt for qualitative methods rather than quantitative or mixed methods to assess skills needs in the sector.

The policy response may also be hindered by factors unrelated to the skills anticipation exercises themselves, such as student preferences. Effective education policy responses to skills intelligence in the health workforce rely on having a sufficient number of teachers, trainers and professional training placements for medical students, as well as a sufficient number of students who want to enrol in medical specialisations that are forecasted to be in shortage. The ACMMP in the Netherlands found that actual intake in health education programmes is often below their minimum recommendation, which is in part due to student preferences (ACMMP, 2020[2]). Although research shows that students can be nudged into different study programmes, for instance through career guidance informed by skills intelligence (De Koning, Dur and Fouarge, 2022[4]) or by financial incentives, the extent to which educational choices can be influenced remains limited. In Bangladesh, a respondent noted that the perceived attractiveness of different medical professions is a key reason why there is a surplus of specialists in certain areas, and shortages in others, including professionals involved in critical healthcare.


[2] ACMMP (2020), Recommendations 2021-2024, http://capaciteitsorgaan.nl/app/uploads/2020/04/2020_02_12-Capaciteitsplan-2021-2024-Hoofdrapport-DEFINITIEF-EN.pdf (accessed on 21 April 2022).

[5] Buchan, J. and L. Calman (2004), “Skill-Mix and Policy Change in the Health Workforce: Nurses in Advanced Roles”, OECD Health Working Papers, No. 17, OECD Publishing, Paris, https://doi.org/10.1787/743610272486.

[4] De Koning, B., R. Dur and D. Fouarge (2022), “Correcting Beliefs about Job Opportunities and Wages: A Field Experiment on Education Choices”, Paper presented at ESPE 2021, Barcelona, Spain.

[8] ILO (2017), Skills Needs Anticipation: Systems and Approaches, Analysis of stakeholder survey on skill needs assessment and anticipation, International Labour Organization, https://www.ilo.org/skills/areas/skills-training-for-poverty-reduction/WCMS_616207/lang--en/index.htm.

[6] OECD (2018), Getting Skills Right: Australia, Getting Skills Right, OECD Publishing, Paris, https://doi.org/10.1787/9789264303539-en.

[7] OECD (2016), Getting Skills Right: Assessing and Anticipating Changing Skill Needs, Getting Skills Right, OECD Publishing, Paris, https://doi.org/10.1787/9789264252073-en.

[1] Ono, T., G. Lafortune and M. Schoenstein (2013), “Health Workforce Planning in OECD Countries: A Review of 26 Projection Models from 18 Countries”, OECD Health Working Papers, No. 62, OECD Publishing, Paris, https://doi.org/10.1787/5k44t787zcwb-en.

[3] WHO (2022), Global competency framework for universal health coverage, World Health Organization, https://apps.who.int/iris/handle/10665/352710 (accessed on 6 May 2022).


← 1. Other skilled occupation lists have a shorter time horizon. The Priority Migration Skilled Occupation List identifies 44 occupations that fill critical skill needs to support Australia’s economic recovery from the COVID-19 pandemic. Of the 44 occupations on the list, 15 are health-related occupations.

← 2. Bosnia and Herzegovina, India, Indonesia and Tunisia.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD/ILO 2022

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.