4. Framework conditions and policy recommendations

Based on the insights presented in the previous chapters and a review of country practices, this chapter discusses the framework conditions required to conduct skills anticipation exercises for the health workforce, i.e. the minimum requirements that a country or region would need to carry out skills anticipation exercises in the health workforce. These include data requirements, social partner involvement, effective governance, funding and human resources. The chapter highlights good practices for developing and using skills anticipation exercises for the health workforce in order to support countries that want to develop these exercises or improve their use for policymaking. It concludes by providing a flowchart for policymakers to follow when implementing a skills anticipation exercise for the health workforce.

The stakeholders interviewed in the context of this study were asked to identify the minimum requirements that a country or region would need to carry out skills anticipation exercises in the health workforce, based on their experiences. In countries where exercises were more limited, stakeholders were asked about the barriers they faced in developing more advanced skills anticipation exercises. Together, these insights suggest that a number of framework conditions must be in place to ensure that skills anticipation exercises for the health workforce are useful for policymaking.

A variety of data is needed to conduct quantitative forecasts or qualitative scenario analyses based on quantitative evidence. These may include data on demographic trends, job vacancies for health workers, or macro-economic outlooks. They may also include information on patient needs, changing disease patterns, epidemiological transitions and technological advances. Having a national statistical office to collect these data will be very helpful in this respect, so that the developers of the exercise can focus on developing the models and producing results, rather than on data collection per se.

Most of the stakeholders in LMICs who were interviewed are facing challenges with regards to the reliability and timeliness of labour market data, making it more difficult for them to conduct quantitative forecast exercises. Instead, these countries focused primarily on qualitative methods, such as focus groups, foresight methods, or employer surveys. Qualitative approaches may be easier than quantitative approaches to set up in contexts where data, budgets and statistical expertise are more limited. In countries such as Bangladesh and Ghana, qualitative surveys of skills needs at a hospital level were the primary anticipation exercise for the sector, in part reflecting the limitations of labour market data to conduct quantitative analysis. Even in upper-middle income countries such as Colombia, the absence of quantitative data was cited as a major methodological barrier to the development or updating of skills anticipation studies. A lack of data on rates of informality, distribution of employment by region, salary offers and training data at the post graduate level were identified as a major challenge to the implementation of sectoral studies in Colombia.

Effective governance of skills anticipation exercises implies involvement and collaboration of relevant actors to ensure that the exercise is designed to meet the needs of users, achieve consensus about skill needs, and agree on coherent and complementary policy responses. This is particularly important for health workforce planning, because it is a shared concern of ministries of health, employment, education, migration, as well as employers and workers and their representative organisations.

Previous literature on skills assessment and anticipation exercises suggests a number of practices to promote dialogue and consensus about skill needs (OECD, 2016[1]). Having an independent institution carry out the exercise can facilitate open discussions about the findings and optimal policy responses. Examples from this study include the Netherland’s Research Centre for Education and the Labour Market (ROA) and Statistics Norway. Another approach is to mandate stakeholder dialogue via legal norms governing consultations around skills issues, to require inclusion of stakeholders in advisory boards to different ministries, or to put a third party in charge of leading and co-ordinating the discussion around current or future skill needs. For instance, Finland’s Skills Anticipation Forum is governed by legislation that mandates stakeholder involvement, and Korea’s new “Act of Providing Assistance with Health Professionals” requires the involvement in skills anticipation of an Advisory Committee comprised of representatives from government, research institutes, labour unions, and the Association of Medical Institution Workers. South Africa’s HWSETA, which conducts sector skills plans and anticipation studies, has a tripartite board which is tasked with the leadership and direction of the organization and with promoting employer, worker, and public interest in skills development within the health and social sectors.

More informal governance mechanisms include the establishment of working groups or roundtables with diverse stakeholders around a topic related to skills needs. For instance, Canada’s Michener Institute set up a symposium to engage stakeholders across Canada about what skills were needed to accelerate the adoption of AI in healthcare. Respondents noted that engaging representatives from rural indigenous communities in Canada had been a challenge, in part because the pandemic had prevented in-person visits to these regions. Employment and Social Development Canada also sets up targeted consultations with stakeholders when they do not agree with their occupational projections, as has been the case previously with projections in the demand for pharmacists.

Representatives of employers and workers should be included in the governance of skills anticipation exercises in the health workforce, in order to promote social dialogue in the sector and to generate results that are reliable and effective. Including social partners in the design and validation of anticipation exercises draws upon their knowledge and expertise of sectoral issues and helps to ensure an accurate identification of the skills, qualifications and occupations that are likely to be in demand in the labour market in the future. Engaging social partners in the anticipation and assessment of skill needs also promotes a smoother adoption of new policies to address health workforce shortages, and facilitates the use of the results by employers and workers, including in training programmes, hiring policies and collective bargaining. In many of the countries reviewed in this study, social partners are invited to validate the findings generated from skills anticipation exercises. For example, Employment and Social Development Canada sends the results from its occupational projections both to the provinces and to sector councils for validation. In some countries, social partners are responsible for or directly involved in conducting the exercises themselves (such as in Australia, South Africa, Finland, and Ghana). Social partner involvement depends on a country having successful and constructive industrial relations and strong networks in the health sector. Where these are absent, a country should seek to strengthen industrial relations and related institutional development in order to facilitate social dialogue and the representation of employers and workers in the sector.

The amount of funding required to conduct skills anticipation exercises depends on the type of exercises and how frequently they are carried out. In this study, the majority of exercises were publicly funded, either by a single ministry or by multiple ministries, but there were instances of privately funded or public-private funded exercises. Given that most funding for skills anticipation in the sector is public, ongoing political will is necessary to ensure that exercises have continuous funding. The absence of continuous funding and political support was identified as a barrier to regular and systematic skill needs anticipation in several countries interviewed. Countries with more limited funding tended to opt for qualitative over quantitative approaches, with LMICs in particular citing limited resources as a major barrier.

Even the best skills anticipation exercise will not lead to effective policies to address health workforce shortages if there is a lack of political will or sufficient funding. Most policy recommendations to address health workforce shortages require additional funding, for instance to hire additional personnel, to train new workers, to upskill and/or reskill the existing workforce, or to invest in technologies that can relieve health workers of some tasks (for instance, the administrative burden) to focus their efforts on areas where critical skills shortages exist. The results of skills anticipation exercises may in themselves be used to help justify political action and additional funding for skills development in the health workforce. In several countries included in this study a lack of funding was reported to be one of the main barriers to translating skills intelligence into policies and programmes for the health workforce.

Among the exercises covered by this study, quantitative forecasts are usually conducted by a relatively small team (five people or less) with specific technical skills in labour economics, econometrics and statistical programming, while foresights and Delphi studies are conducted by larger teams (10 or more) with less technical skill requirements, such as communication and interpersonal skills. The latter type of exercises also depend on the participation of groups of experts (often 10 or more). Whether the people who conduct the exercise work on it full-time depends on the scope of the exercise and the frequency with which it is repeated. For instance, Employment and Social Development Canada (ESDC) has five people working part-time on the bi-annual quantitative whole-of-labour market forecasts, while the Netherlands’ Advisory Committee on Medical Manpower Planning (ACMMP) employs 12 full-time equivalent workers, as well as people from other research institutes such as NIVEL, to run its mixed-methods forecasts for detailed health occupations every 3 years. Many developers of skills anticipation exercises contract out part or all of the research work, because they struggle to find enough people with the required knowledge or skills to conduct the exercise in-house, or because it is more economical to outsource specific aspects of the exercise rather than hire full-time employees, particularly in case of one-off exercises. In some countries (particularly in LMICs), additional funding for training and upskilling may be needed in order to develop and interpret skills anticipation exercises for the health workforce. Effective implementation also depends on having skilled government workers who can interpret skills anticipation exercises.

Drawing on the best practices identified in stakeholder interviews and from the literature, a number of recommendations can be made to guide countries in anticipating the future skill needs of the health workforce.

The optimal design of a skills anticipation exercise for the health workforce will depend on the intended policy use. Choices about method, scope, skills definition, frequency, and time horizon need to be made with the final policy objective in mind. For instance, exercises that are national in scope are best suited for informing policies relating to maximum student places in health education programmes, or policies to target migrant workers with relevant skills and qualifications. Meanwhile, exercises conducted at the sub-national level can generate information to inform career guidance, which should ideally be at the local or regional level to meet the needs of adults who are place-bound. Similarly, a whole-of-labour market approach is appropriate for evaluating the relative size of shortages of workers in the health sector relative to other sectors, and for optimising the skills mix across sectors; while a health workforce specific approach is most appropriate to understand how skill needs are changing within particular health occupations in order to inform new training curricula and can normally be conducted more rapidly and with fewer resources. Table 4.1 summarizes some of the advantages and disadvantages of the different methodologies identified in this study for different policy uses.

There is consensus in the general literature that combining a variety of approaches, including quantitative and qualitative, is the best approach for achieving robust and reliable results (CEDEFOP, 2008[2]). A mixed methods approach allows a country or region to leverage the advantages and mitigate the disadvantages of different types of exercises, as summarised in Table 4.1. This review highlighted several examples of mixed methods approaches for skills anticipation exercises for the health workforce. For instance, both the Finnish National Agency for Education (OPH) and the Dutch Advisory Committee on Manpower Planning (ACCMP) conduct quantitative forecasts that serve as inputs for qualitative Delphi discussions about skill needs in the health workforce. In South Africa, qualitative stakeholder consultations are used to verify the validity of the quantitative forecasts. While conducting exercises with mixed methods is likely to be the most costly approach and may be more challenging for countries with limited resources and particularly LMICs, it is worth further investment as it generates the most reliable and robust data on future skill needs.

Most of the exercises reviewed in this study project future demand for health workers by occupation or qualification. It is rare to find examples of exercises that focus explicitly on how the skills required for the health workforce are likely to change over time in response to trends like technological change, partly because individual skills are poorly understood and hard to quantify. However, those exercises that do focus on skills facilitate a more proactive and dynamic approach to health workforce planning that goes beyond anticipating manpower needs to consider which skills will be needed to allow health workers to adapt to new technologies.

Though relatively uncommon, several of the exercises identified in this review (Australia, Finland, Canada and Norway) did focus on skill needs explicitly. A focus on skills is particularly useful in informing curriculum development for health education and training, task reallocation between health occupations, and skills mix. The slower adoption of new technologies in low and medium-income countries may reduce incentives to carry out such skills-focused exercises.

Tripartite social dialogue throughout the process is essential to skills needs anticipation. As already noted in the framework conditions above, social dialogue and cooperation among key stakeholders helps to ensure that skills intelligence is fit for policy use and promotes buy-in to the policy response among stakeholders. Social dialogue is key to making informed decisions in the processes of designing and implementing the exercises, to making practical sense of analytical results and to the effective application of the findings. Skills anticipation in the health sector should include appropriate institutional mechanisms for social dialogue, and capacity development for social partners in order to generate constructive input into and feedback on the information generated. By following this approach, more robust skills anticipation exercises that translate into effective skills development policies can be developed in the health sector.

The data sources analysed for this report corroborate existing evidence (such as (WHO, 2016[3]) suggesting that health workforce shortages are widespread, and therefore represent a global issue. Health workforce shortages exist across low, medium and high-income countries, but are felt more severely in LMICs. Migration flows of health service workers tend to consist of movements from lower-income to higher-income countries, and while these flows can improve flexibility and the transfer of skills, they can also have negative impacts for origin countries that lose skilled and trained health workers to destination countries. This can have particularly damaging effects where countries have dedicated scarce resources to training workers who are then absorbed into healthcare systems abroad.

Bilateral agreements between origin and destination countries may help to mitigate the negative consequences of migration. For instance, the Model Agreement developed by the ILO and used by the Philippines and various destination countries involves measures to compensate the origin country for its loss, including planning for the return and reintegration of health workers, the exchange of students and expert visits, scholarship programmes, joint venture and investments in origin country health facilities, twinning of health facilities, and support to improve education and training facilities and technology transfers (Makulec, 2014[4]). The ILO Guidance on Bilateral Labour Migration Agreements ( (ILO, 2022[5]) can provide guidance to ensure safe, orderly and regular labour migration based on international labour standards.

Given that health shortages are a global issue, international cooperation is needed not only in the policy response but also in planning and conducting skills anticipation exercises. The exercises identified by this study solely focused on national and sub-national efforts to anticipate skill needs in the health workforce. However, internationally-coordinated efforts may be appropriate for addressing global skills shortage issues, and particularly to inform migration flows. For instance, the OECD and ILO Skills for Jobs database is an attempt to develop a harmonised cross-country measure of shortage and surplus intensity. Knowledge sharing between countries is also needed to support low and medium-income countries to face barriers in conducting skill anticipation exercises. Cooperation and knowledge sharing between higher income countries and LMICs on skills anticipation could also help to promote development in the health workforce in these countries, and create mutually beneficial solutions to skills shortages in the sector.

The review of exercises to anticipate skill needs in the health workforce conducted in this study suggests that there is no “one size fits all” approach. Rather, different approaches may work better depending on a country’s policy objectives, resource constraints, data availability and governance arrangements. The following flowchart provides an overview of some of the key decisions that need to be considered when establishing a skills anticipation exercise for the health workforce (Infographic 4.1).

First, public authorities need to agree on the policy objective of the exercise (Decision #1). The most common policy objectives related to the health workforce among the exercises included in this study are education policy (such as student intake, course content, or career guidance), employment policy (such as skills mix, task reallocation and occupational standards), migration policy (such as skilled migration lists and international cooperation) and collective bargaining processes.

Having a clear idea of the policy objective will facilitate setting the key design parameters, including how skills are defined, frequency, time horizon, and scope (Decision #2). For instance, if the intended policy objective is to set student intake in medical programmes, a longer time horizon may be needed, given long lead times in health education programmes. If the policy objective is to review the way tasks are allocated across health professions or to inform upskilling of health professionals in the face of technological change, an exercise at the skill or task level may be best suited. Exercises at the skill level may make use of a pre-existing skills framework (such as O*NET or DigComp 2.0), or a national qualification framework. Since longer time horizons imply more uncertainty, the frequency at which such exercises are repeated may need to be higher. If the objective is to identify future skill needs of specific health occupations, then the scope of the exercise will need to be more targeted than a whole-of-labour market approach.

With the policy objective and design parameters in mind, the next consideration is to take stock of available input that could feed into the skills anticipation exercise (Decision #3a). What kind of data is available, and how regularly is it produced? Who produces it and is it of good quality? In addition to quantitative input, expertise that could help to inform a qualitative exercise should also be assessed, including among hospital managers, health professionals, and patients. In parallel, public authorities should consider the human resources (number of people available in-house and their skills, as well as outside consultants or think tanks) and financial resources that are available to develop and support the exercise, interpret the results and translate it into policy for the health workforce (Decision #3b).

Based on the above considerations, public authorities need to decide on the optimal skills anticipation method (Decision #4). The review conducted in this study suggested that a mix of quantitative and qualitative approaches was likely to produce the most robust and reliable results to anticipate skills needs in the health workforce. However, it also showed that countries with more limited resources and fewer high-quality data sources tend to opt for qualitative approaches such as focus groups or employer surveys rather than quantitative forecasts. Most countries that aimed to use the skills anticipation exercise for manpower planning in the health workforce or to determine student intake into health education programmes conducted quantitative forecasts or applied a mixed-methods approach.

After deciding on the method, public authorities will need to assess whether any additional input will need to be produced or sourced (Decision #5). With quantitative forecasts, data requirements are high and additional data may need to be sourced. With focus groups or foresight exercises, new partnerships may need to be established to engage social partners and other relevant stakeholders in the exercise.

The final consideration is to establish a governance structure for the exercise (Decision #6). This involves defining who will be responsible for conducting the exercise, which ministry or ministries or other stakeholders will be the end users, and mechanisms to involve social partners and other stakeholders in developing the exercises or and validating the results. In this review, ministries of health, education and/or labour were often involved in developing and/or validating the exercises. Formal or informal cooperation mechanisms (such as working groups or conferences) may need to be set up to engage buy-in from all the relevant stakeholders and social partners.


[2] CEDEFOP (2008), “Systems for anticipation of skill needs in the EU Member States”, CEDEFOP Working Papers, Vol. 1, https://doi.org/10.2801/24837.

[5] ILO (2022), Guidance on Bilateral Labour Migration Agreements, International Labour Organisation, https://www.ilo.org/global/topics/labour-migration/publications/WCMS_837529/lang--en/index.htm.

[4] Makulec, A. (2014), Philippines’ Bilateral Labour Arrangements on Healthcare Professional Migration: In Search of Meaning, International Labour Organization, Makati City, https://www.ilo.org/manila/publications/WCMS_320609/lang--en/index.htm.

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

[3] WHO (2016), Global strategy on human resources for health: Workforce 2030, World Health Organization, https://apps.who.int/iris/handle/10665/250368.

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.