1. The green transition and jobs: what do we know?

Climate change and environmental degradation are among the most formidable challenges the world faces that will change how we live, work, produce and consume. Leading scientists have pointed out the risk of exceeding environmental tipping points, i.e. critical thresholds at which a tiny disturbance can alter the Earth’s climate system (Lenton et al., 2008[1]). Surpassing such tipping points could expose the world to long-term irreversible changes, with potentially dramatic consequences for lives and livelihoods globally. Consequently, “the growing threat of abrupt and irreversible climate changes must compel political and economic action on emissions” (Lenton et al., 2019[2]) (OECD, 2022[3]). Environmental policies and regulation that support the transition to a net-zero economy will lead to changes in industrial production, consumption patterns and energy provision. They will also require a transformation across every industry in the economy, ranging from production processes and innovation to the adoption of new technologies and investments.

Climate action and policy will need to contribute to a “greening” of the labour market, which has four major labour market implications. First, it will lead to the creation of new types of jobs. Second, it will entail the loss of some existing, “old” jobs. Third, it will cause a shift in the skills required in many jobs in the economy. Fourth, the green transition has a strong local angle. Risks and challenges in terms of jobs are uneven and are often concentrated in specific regions. Economic opportunities and job creation, likewise, will not materialise everywhere, with some places likely to benefit more than others from the shift towards carbon-neutral and environmentally friendly jobs and sectors.

Despite the fundamental changes that the green transition will bring about for labour markets, there is a lack of systematic evidence on the green transition’s impact on local labour markets. Existing work is often either very descriptive or qualitative. Most analysis has focused on the risks of the green transition (e.g., job losses) while its opportunities remain an under researched topic and is often limited to only a handful of sectors (e.g., agriculture, renewable energy or carbon-intensive industries) or a few specific jobs. The discussion has yet to cover the full picture, because green jobs and green skills extend beyond those subsets of sectors or jobs. Finally, most existing work on green jobs focuses on national level analysis and therefore does not identify any of the (potentially significant) geographic discrepancies within countries.

This chapter fills that gap and provides an overview of analytical work on the green transition’s labour market effects and local implications. It explains different approaches of assessing the opportunities and challenges that the green transition poses for workers, firms, sectors, and local economies. It summarises the main existing evidence on the impact of green policies on jobs in different communities. Finally, it sheds light on what we know and what we do not (yet) know with respect to the uneven effects of the green transition.

In most OECD countries, policies on environmental change have been on the political agenda for decades but policy actions to address the labour market implications of environmental policies are more recent. Over the last 30 years, the stringency of environmental policies related to air pollution, energy and carbon emissions in OECD countries increased significantly (Figure 1.1) — particularly between 2000 and 2010. However, until recently, many of those programmes included only limited direct labour market support despite a general expectation that green policies and stricter environmental regulation will affect the labour market. A number of new initiatives put a stronger focus on labour market aspects.

Advanced economies around the world have recently passed major policy packages for supporting the green transition and generating green economic growth. The European Commission’s European Green Deal (EGD) calls for all EU countries to reduce greenhouse gas emissions by 2030 by 55% compared to their levels in 1990 and aims to support EU member countries by mobilising more than EUR 1 trillion to finance the transition (Box 1.1). In the United States (US), the Inflation Reduction Act provides USD 369 billion for climate and clean energy, making it the single largest climate investment in US history (Box 1.2).

Given the urgency of addressing the climate emergency, these green policy packages are unprecedented in size and ambition. They entail multiple objectives. They aim to facilitate the reduction of greenhouse gas emissions and ease the reliance on fossil fuels. At the same time, they aim to boost sustainable economic growth that benefits all, ensuring a just and inclusive transition. For example, the European Commission emphasises the importance of focusing on “the regions, industries, workers, households and consumers who will face the greatest challenges in terms of employment, social and distributional impacts when decarbonising their economy” (European Commission, 2021[5]).

In response to the COVID-19 pandemic, most OECD governments passed large-scale economic measures, often including green recovery measures. Green policy measures account for a significant proportion of the recovery packages so far passed in 44 countries, consisting of OECD countries, EU countries and selected large economies (see Box 1.3). Spending on green recovery measures in those countries and the European Union totalled USD 1 090 billion by April 2022. Thus, spending on green recovery measures rose by more than USD 400 billion compared to September 2021.1 At the same time, however, some governments also introduced COVID-19 recovery measures that will have a negative or mixed impact on the environment.

Despite their importance, only a very limited number of measures are directed at areas that will support the development of “green” skills. The green COVID recovery programmes mainly reflect broader grants, tax reductions/other subsidies, and regulatory changes. In contrast, a relatively small share of measures are directed at two areas that are essential for the creation of green jobs and the provision of workers with the right (“green”) skills for the green transition: 8% promote research and development (R&D) and 2% relate to skills development.

Skills development and retraining are vital to ensuring that workers have the right skills to prosper in a changing world of work and are a prerequisite for making the green transition a “just transition”. Many workers in polluting jobs or sectors will require targeted support so that they can move into new jobs, greener sectors or retrain if the requirement of their jobs change as polluting sectors green. Indeed, even without the changes the green transition will entail for jobs and skills, many OECD countries already experience skills gaps and mismatches as obstacles to economic growth and productivity gains (OECD, 2021[8]).

The limited number of green measures on skills development is even more striking in terms of committed funding and subnational focus. Around USD 15.6 billion of the total funding of USD 1 090 billion are allocated to skills training.2 While skills gaps and mismatches affect most OECD economies, they differ substantially within OECD countries. Thus, training needs and the challenges of developing skills for the green transition must take into account local challenges and conditions (OECD, 2018[9]). For example, designing skills training with a subnational lens can yield a more effective matching with the demand for skills and jobs by employers in a local economy (OECD, 2022[10]). However, as a percentage of all recovery funding, skills training at the subnational level is considerably less than 1% (0.0024%).

Green policies and environmental regulation can affect the economy in many ways. Empirical analysis has so far focussed on their impact on economic growth, labour productivity, investments, technology adoption, firm performance, international trade, and the demand for skills (Box 1.4).

An outstanding, and politically sensitive, issue is the impact of environmental policies on employment, particularly the differential impact across communities. So far, the evidence on whether green policies lead to job losses or gains is mixed and suffers from considerable evidence gaps and depends on the impacts on the economy more generally (Box 1.4). A number of studies find negative employment effects of environmental regulation, particularly for energy-intensive industries (Walker, 2011[11]), (Greenstone, 2002[12])), while other works suggest there is a relatively small effect on overall employment ( (Gray et al., 2014[13]), (Brucal and Dechezleprêtre, 2021[14])). Evidence on fiscal stimulus packages shows that targeted green investment stimulates significant employment creation, especially of green jobs (Popp et al., 202[15]).

Even if the employment effects of green policies have a small impact at the national level, they may have a strong localised impact. As a result of the reallocation of workers from energy-intensive (or polluting) to non-energy-intensive (non-polluting) sectors, regions that specialise in the former are likely to bear, at least in the shorter term, a greater share of the labour market costs of environmental policies. For example, while a 10% increase in energy prices leads to a reduction of manufacturing employment by 0.7%, this reduction reaches 1.9% for energy-intensive sectors such as iron and steel production (Dechezleprêtre, Nachtigall and Stadler, 2020[16]).3

The main body of work on green policies and employment has two major caveats. First, it lacks analysis of the geographic impact, i.e., the uneven effects. Second, it does not offer insights into how green policies change the composition of jobs in a labour market, which matters for a range of policies such as education, adult learning, vocational training, or active labour market policies. Even though the consensus is that job losses and job creation due to green policies and the green transition overall will not necessarily be evenly spread across different places, most existing work does not investigate the different impact across places within countries (OECD, 2017[24]). This matters even more because labour is relatively immobile in the short run, so that those risks of green policies can result in a rise in local unemployment and relevant adjustment costs in areas affected the most (Valero et al., 2021[25]). A just transition requires not only better understanding the spatial impact of green policies, but also their distributional consequences for different workers, which can then inform policy responses that help support workers negatively affected or in risk of displacement.

As highlighted for G7 environment ministers, more quantitative evidence on the employment effects of green policies is needed (OECD, 2017[24]). Better and consistent data play a vital role in filling evidence gaps across multiple dimensions, including numbers and characteristics of jobs that might be created or destroyed as well as estimates of the number of workers who will need to transition into different sectors or occupations. Additionally, effective policy design requires knowing whether those workers possess the skills and competencies required in newly created jobs or how they can re- or upskill to meet those requirements.

Subnational data are critical. Most of the existing evidence on the link between green policies and employment focuses on aggregate, national information. However, when it comes to employment effects, green policies inevitably have a spatial impact. They will affect particular jobs and industries, especially those with greater pollution intensity, the most. As those jobs are not evenly spread out but often concentrated in specific areas, the challenges in terms of employment loss or changes in job requirements will also differ markedly within countries. Likewise, the opportunities that green policies present for job creation and green economic clusters also require a local perspective.

These data are also essential to deliver the green transition in the first place, as a lack of “green” talent can hold back the green transition. Across the OECD, companies face unprecedented labour shortages (OECD, 2022[26]). For example, in the European Union, almost 30% of firms in both manufacturing and services encountered production constraints in the second quarter of 2022 because of a lack of labour. In the United States, the number of vacancies is almost double that of unemployed persons. Those labour shortages are exacerbated by widespread skills mismatches, which make it hard to fill vacant positions with unemployed individuals. To mitigate labour shortages and skills mismatches for green industries and firms, policy makers need to know what skills and tasks green jobs entail before they can effectively design policy.

The lack of consensus on measuring green jobs is a major reason why the evidence on job creation by green policies varies widely. Different studies and institutions use different approaches, which makes it hard to assess the impact of green policies consistently across countries. Comprehensive and solid evidence with comparable data is needed for designing and evaluating green policies that support the creation of green jobs. Additionally, data and evidence on polluting jobs and how they are affected by green policies is important for policy.

Despite the fundamental changes the green transition entails for labour markets, there is not yet a universally accepted definition of what a green job is. Thus, different studies have come to different conclusions about the proportion of green jobs in the economy or green job creation (Bowen and Kuralbayeva, 2015[27]).4 Consequently, it is difficult to compare studies on the magnitude of green job creation or the relative importance of green jobs in the labour market. Additionally, most work until now has focussed on single or a narrow set of countries.

Across international organisations and national governments, definitions of green jobs vary. The United Nations defines green jobs as jobs in sectors that contribute substantially to preserving or restoring environmental quality and minimise waste creation and pollution (UNEP, ILO, IOE, ITUC, 2008[28]). The ILO defines green jobs as “decent jobs in any economic sector (e.g., agriculture, industry, services, administration) which contribute to preserving, restoring and enhancing environmental quality.” Thus, the ILO’s definition combines two criteria. First, green jobs help ‘reduce the environmental impact of enterprises and economic sectors by improving the efficiency of energy, raw materials and water; de-carbonise the economy and bring down emissions of greenhouse gases; minimise or avoid all forms of waste and pollution; protect or restore ecosystems and biodiversity; and support adaptation to the effects of climate change’ (ILO, 2022[29]). Second, the ILO includes a social component, referring to ‘decent’ jobs that are productive, secure and deliver a fair income, among other things.

Other international organisations refer to skills as a means of defining green jobs. For example, the European Centre for the Development of Vocational Training (Cedefop) defines green skills as “the knowledge, abilities, values and attitudes needed to live in, develop and support a sustainable and resource-efficient society” (OECD/Cedefop, 2014[30]). Then, green jobs are those that require green skills. Another possible way of defining green jobs is to first determine what the green economy is and then include those jobs are part of it (see Box 1.5). These definitions are relatively broad and not operationalised, i.e., systematically applied to available data across the world.

The vast majority of the empirical literature draws on two dominant approaches to quantify green jobs (Figure 1.2). The first type consists of top-down approaches that identify sectors or industries that are green and consider that all employment in those sectors or industries is green. The second type, i.e. bottom-up approaches, exploits information on occupations. They define green jobs based on the skills or tasks different occupations entail and the extent to which those tasks or skills are green (for a more detailed discussion of the different approaches see (Valero et al., 2021[25])). A third strand of macro-modelling approaches does not explicitly measure the number or location of green jobs but can provide an important overall assessment of the impacts of environmental policies on the labour market.

The fact that there is no universally agreed definition creates a number of challenges. First, it causes some research to avoid the term completely (Bowen and Kuralbayev, 2015[31]). Second, it makes it extremely difficult to compare different studies, which differ with respect to the countries, parts of the economy and time frame they cover. Thus, studies that include estimates of the share of green jobs in the labour market come to widely different conclusions (Table 1.2). Those estimates range from 2% (Eurostat, 2022[32]) to 40% (Bowen and Hancké, 2019[33]).

However, data and statistics on the relative importance of green jobs and their characteristics are crucial for policy and supporting the green transition. They provide an understanding of the magnitude of the labour market implications of the green transition and environmental policies, and can give a sense of pending structural changes to local and national economies. Furthermore, such statistics are required for an effective design and evaluation of labour market and skills policies. This is particularly important in the current context, with widespread skills and labour shortages, which raise an important question for policy makers: Do we have enough green talent? Amid prevalent skills shortages, notably for skills required the most in green jobs, the job creation potential of green sectors and industries might not be fulfilled (ILO / CEDEFOP, 2011[34]).

To better understand why the existing work on green jobs produces wildly different estimates, the subsequent sections review top-down and bottom-up approaches of defining green jobs as well as more general macro-modelling forecasts. Those sections explain some of their advantages, limitations, and the assumptions they entail for understanding why policymakers sometimes choose particular approaches in line with specific policy objectives and how, in the end, those approaches are complementary (Martinez-Fernandez, Hinojosa and Miranda, 2010[42]).

Top-down approaches have in common that they classify green employment based on the production process and/or outputs. For example, a sector could be classified as green based on its output, and thus all employment (regardless of the occupation) in that sector would be considered green. Prominent cases of such sector-based definitions include, for example, estimates for the renewable energy sector. In practice, empirical studies adopting this approach tend to make use of existing sector and/or industry classifications. This, coupled with limited data availability, nudges studies towards measuring employment in certain sectors (as defined by a standard classification) of the economy first, and then refining their estimates, depending on the resources available. However, top-down approaches could also generally consider areas of the economy that produce environmental goods or services, and therefore include industries from different sectors.

Top-down approaches can vary significantly in their estimates and in the breadth of their analysis. For example, a recent report on renewable energy estimated that the sector provides around 1.24 million full-time equivalent jobs in the EU (IRENA and ILO, 2022[36]). Broader definitions often consider a larger set of activities, also referred to as the environmental goods and services sector (EGSS), which comprises industries centred on environmental activities, i.e. activities that either directly serve an environmental purpose or produce goods that are specifically designed to serve an environmental purpose.5 In the EU, estimates find that EGSS accounts for 4.4 million full-time equivalent jobs, or 2% of total employment (Eurostat, 2021[35]). The EGSS is a leading example of the top-down approach, and it is therefore relevant to describe it in more detail.

This top-down approach aims to quantify employment (and other measures such as GDP, GVA and exports) on the environmental goods and services sector (EGSS). It builds on the UN System of Environmental-Economic Accounting Central Framework (see Box 1.7). EGGS revolves around the concept of environmental activities, including examples such as the manufacturing of bio-fuels, installation of photovoltaic panels, or environmental consulting services. The EGGS includes the set of producers who engage in environmental activities, such as environmental protection activities, aimed at reducing and preventing greenhouse emissions or other harmful environmental impacts, and resource management activities related to energy.

A core challenge of this approach is that producers may engage in secondary activities, which may or may not be environmental. Ideally, those secondary activities should be separated and classified accordingly. However, in practice this is not always possible given the data limitations. In the end, it is up to countries to decide how and if to separate secondary activities, which might weaken comparability across countries.

Results for this approach show that the size of the green labour market is small, around 2.2% of the jobs in the EU in 2019 were in the EGSS. Within the EU, there is some dispersion across countries, but the share of green jobs remains low, ranging from 0.9% in Belgium to 5.1% in Finland. Although the EGSS labour market has grown faster than the total labour market since 2000 (43.3% for the former and 11.7% for the latter), the growth in the proportion of jobs in the economy has been limited to 0.5 pp (from 1.7% to 2.2%). In addition, 97% of growth in the proportion of EGSS jobs was between 2000 and 2011, with little growth over the last decade.

This framework also requires that each country develop its own environmental statistics. Ideally, each country would follow the same methods and definitions, but inevitably, discrepancies in the process exist. The work done by the UK’s Office for National Statistics (ONS) is a useful example of the EGGS in practice. ONS follows the definitions provided by the EGSS and applies them using a variety of sources such as the Low Carbon and Renewable Energy Economy (LCREE) Survey, among others.

Since 2014, the UK’s ONS has compiled data on the LCREE. The data are derived from survey-based estimates of turnover, exports, employment and acquisitions and disposals of capital assets in 17 low-carbon sectors6. This is part of a broader program by the ONS that compiles information on Environmental Accounts such as natural capital, atmospheric emissions, and energy use. Businesses are considered to be part of the LCREE if they report activity in one of 17 defined sectors.7 Results show that for 2020, employment in the LCREE was around 207 800 full-time equivalents (ONS, 2020[37]).8

In short, LCREE and EGSS have the same objective, but the former serves as an input for the latter as the EGSS has a broader focus. LCREE does not include some "green" activities, such as recycling and the protection of biodiversity.9 Furthermore, EGSS, uses a combination of other surveys, supply-use/input-output tables, and other complementary data to provide estimates in line with international guidelines.

  • Ease of interpretation: It is easier to interpret if the primary objective is industrial analysis, and, depending on the information systems available in the country, may provide details on specific activities that are not typically identifiable in Labour Force Survey data (the primary source for the bottom-up approach), as they are less detailed than the most detailed industry classifications.

  • Limited data collection needs: Given that quantifying employment using the top-down approach is a spin-off of quantifying GVA, import and exports (national accounts), a lot of the data is already collected. New data collection requirements are limited.

  • Ancillary green jobs: Because the approach captures all jobs engaged in a certain activity it also captures jobs in the ‘green’ firm, that may not in and of themselves be considered green but, rather, support green activities. However, at the same time, as shown below this can also generate challenges with comparability depending on the degree of outsourcing a firm may engage in.

  • Statistical unit: A key difficulty with the top-down approach is the choice of statistical unit. The preferred choice would be the Local Kind of Activity Unit used in the European Statistical System or the Establishment used in the international system but in practice many countries use other measures of statistical units, such as the enterprise, and commonly legal units, which can impair international comparability. In addition, the choice of enterprise for example can also create challenges for estimating sub-national breakdowns as it may assume similar production activities across all regions where the parent has operations.

  • Limited subnational data availability: A key caveat of most top-down approaches is that data are not reliable enough to construct sub-national estimates. National data might, however, obscure potential regional differences that may be relevant.

  • Jobs unrelated to green activities may still be considered green: As all jobs in the sector are included, estimates inevitably include jobs that are only indirectly related to the underlying green activity. A clear example of this is ancillary activities, accountants or security guards for example, who will be classified as green as long as their statistical unit is deemed to have an environmental purpose. Indeed, the argument that all jobs engaged in the ‘green’ firm should also be included creates an argument that all upstream jobs (including in non-green activities) should also be included.

  • The approach is affected by market structure: The degree of jobs that are considered as being in green activities are also dependent on the market structure of the economy given the varying degrees of outsourcing firms may engage in. For example, if a company in the recycling industry that has an accounting department decides to outsource that department, then those employees (previously considered to be in green jobs) would no longer be part of the estimates of green jobs if they were integrated by a third-party company fully dedicated to accounting. Such issues could lead to unstable results that may be impacted by arbitrary changes in the production structure.

Bottom-up approaches of defining green jobs are based on information from individual occupations. They classify occupations based on their characteristics and the work they entail (Table 1.3). Quite common among these approaches is identifying green jobs based on the tasks they require and the extent to which those tasks are green (i.e., they support environmental objectives). Alternatively, green jobs can also be determined by defining specialised green skills and examining the jobs that require them. Finally, as pointed out above, some organisations also include a ‘decent job’ requirement as a condition for a job to be green.

Most studies to date that take a bottom-up approach follow the methodology proposed by (Vona et al., 2018[45]). Their work rests on the fact that each occupation can be divided into different functions, called tasks, which may be carried out daily, monthly or even on a yearly basis. Therefore, each occupation can be divided into a unique set of tasks. Occupations can then be analysed and categorised based on the content of these tasks.

The task-based approach is the dominant approach in labour economics for studying the labour market impact of structural transformation. Pioneering contributions include work by David Autor and co-authors ( (Autor, Levy and Murnane, 2003[46]), (Autor, 2013[47])), who assessed the impact of computers and digital technologies on the labour market, as well as (Acemoglu and Autor, 2011[48]) and (Acemoglu and Restrepo, 2018[49]), who provided further theoretical consolidation/confirmation of the approach by examining the effects of automation and digitalisation.10

While different sources can be used to retrieve information on occupations, the most widely used source is the Occupational Information Network (O*NET) of the U.S. Department of Labor. Through a mixture of direct surveys, occupational expert consultation and case-by-case analysis, O*NET (see Annex for further details) provides information on a large number of occupations and the tasks they entail, the relative importance of those tasks and the skills generally required to fulfil those tasks (Dierdorff and Norton, 2011[50])11. While bottom-up approaches need not be task-based but could instead be, for example, skills-based, most research focuses on tasks. O*NET’s Green Economy Programme (Dierdorff et al., 2009[51]) provides a list of individual tasks for each occupation and classifies these tasks as green or non-green.12 Based on this information, an occupation can be classified according to the overall greenness of its tasks. The infographic below provides some examples of this.

A number of academic articles and reports have used the task-based approach to examine the share of green jobs in the labour market and to shed light on the effects of the green transition. They combine information from O*NET with granular employment data in different economies. Vona et al. (2019[40]) examine drivers of green job growth in US metropolitan and non-metropolitan areas between 2006 and 2014 and is thus one of the first studies to examine spatial differences in terms of green jobs. They provide estimates of the green intensity of employment, which is the employment-weighted average of the greenness of tasks in a given local labour market (see previous section).13 The green intensity of employment can be broadly interpreted as the share of tasks or time spent on green activities. They find that the green intensity of employment is around 3%, although this ranges from 2.4 to 3.9 within the US. Although they find a convergence of green employment over time across the US, it remains geographically concentrated, especially in high-tech areas.

Recent analysis extended the scope to 34 countries, mainly consisting of more advanced economies in the EU, South Africa, Mexico and the US (International Monetary Fund, 2022[52]). It finds comparable green intensity of employment across the economies covered, ranging from about 2 to 3%. Additionally, the study examines the average employment-weighted pollution intensity of economies (see Measuring green jobs at the regional level for further detail), which ranges from 2% to 6%. The study finds that high-skilled urban workers are in occupations with a higher green intensity, which also comes with a wage premium. A second study adopted the same approach with a global perspective (IMF 2022). This study focuses on the green intensity of workers (i.e., the share of their tasks that are considered green). The work estimates that high-skilled urban workers tend to have a higher green intensity, and that this measure ranges from 2% to 3% in the economies they study.

A related body of work applies the task-based approach but looks at a different outcome, namely green jobs. Instead of examining the green intensity of employment, those studies use the same occupation and task classification provided by O*NET to identify green jobs as either those that entail any green task or include a significant proportion of green tasks. Valero et al. (2021[25]) consider not only workers who directly contribute to preserving the environment and reducing emissions, but also indirect green jobs, which are jobs that may not directly contribute but are expected to be positively impacted by the green transition through increasing demand. Overall, they find that around 17% of the jobs in the UK are either directly or indirectly green, and that the share of green jobs differs widely across sectors. Furthermore, they document that a disproportionately large proportion of green jobs are held by men and highly educated individuals.14 For the US, (Bowen, Kuralbayeva and Tipoe, 2018[41]) estimate that around 19.4% of workers are directly or indirectly employed in green jobs. In accordance with the other studies which consider the sub-national dimension, this study finds relevant differences across regions as the share of green jobs ranges from 13% to 25%. Finally, (Elliott et al., 2021[53]) study eco innovation and job creation within Dutch firms. Defining green jobs as those occupations with a greenness index (the weighted share of green tasks among all tasks of an occupation) greater than the average greenness index, they find that around 30% of jobs are green.

Bottom-up approaches, especially those based on tasks of occupations, offer a number of appealing features that might be particularly helpful for policy. At the same time, however, they also have a few limitations. Below is a short description of the most important aspects.

  • Subnational information: When detailed LFS data is available, computing sub-national estimates is straightforward as it does not require collecting extra data or further estimation compared to national estimates. This is particularly relevant for policy because, like other megatrends, the green transition will have different effects across different places.

  • Socio-economic information: Unlike most top-down approaches, task-based definitions enable detailed analyses of workers in green jobs and how their characteristics such as education, skills levels, gender, age or wages differ from non-green jobs. Thus, it can provide useful information on how the green transition might affect socio-economic disparities.

  • Link to labour market policy: A key implication of the green transition is that it will lead to a reallocation of jobs and a change of skills that are in demand. Occupational task-based approaches lend themselves to examining how workers can transition into new jobs or sectors through retraining or upskilling and can inform, with greater precision, skills and active labour market policies.

  • Coverage of green ‘activities across sectors’: Workers outside of sectors traditionally considered green can also contribute to net-zero and other environmental objectives. Task-based approaches are “sector-blind” and include all workers that support such objectives directly, for example researchers employed in high-emitting firms that are engaged in activities aimed at reducing, mitigating, or eliminating emissions.

  • Crosswalks and US-based tasks: O*NET task data were built in and for the US. Applying the concept of green tasks to occupations in other countries assumes that the same set of tasks are green in other countries and indeed at the sub-national level. Additionally, information on occupations’ tasks needs to be translated from the US occupational system (SOC) into the occupational classification systems of other countries. In practice, official crosswalks, i.e., mappings, are available. However, such crosswalks are not perfect for all occupations. For example, SOC-2010 occupations “Technical Writers” and “Writers and Authors” (SOC-2010 27-3042 and 27-3042 respectively) are both mapped to ISCO-08 occupation “Authors and related writers” (ISCO-08 2641) meaning not all matches are one-to-one.

  • Green tasks over time: O*NET’s Green Task Development Project offers a static picture of what green tasks are at a given time. It does not include information on how the set of green tasks has changed over time, nor does it claim that future green tasks will be the same.

  • Occupational employment data: Although data on occupational tasks is available at a detailed level15, occupational employment data is not always collected or published at the same level of detail across countries. This might undermine international comparability of estimates when sufficient detail is not available in certain countries (see section “Measuring green jobs at the regional level” for ways to resolve this issue).

  • Exclusion of ancillary or indirect activities: the approach mainly includes direct green jobs and therefore excludes workers that are employed in ancillary activities or jobs. Examples of this include security guards or accountants in the renewable energy sector.

Macro modelling approaches do not directly estimate the share of green jobs but instead examine the possible impact of green policies on the labour market. They may also measure the impact of other environmentally inspired changes in the economy, such as changes in consumer demand, preferences, or productivity. Therefore, this approach is flexible with regard to the definition of green policy. Macro-modelling approaches usually consist of dynamic computable general equilibrium (CGE) models that provide projections on how the economy and the labour market may respond to a given policy change. Thus, they can for instance be employed to provide ex-ante projections of the effects of concrete policy objectives such as the Paris Climate Agreement or the European Green Deal. One such approach is the OECD ENV-Linkages model (Box 1.9).

Macro-economic models can be used to study different questions or policies. For example, the OECD ENV-Linkages model has been used to study the jobs potential of a transition towards a resource efficient and circular economy (Chateau and Mavroeidi, 2020[55])(Chateau and Mavroeidi, 2020[52]). Other work has explored the consequences on the labour markets of structural changes induced by decarbonisation policies (Chateau, Bibas and Lanzi, 2018[54]). The latter finds a relatively small impact on the labour market, as less than 2% of jobs across the world would be destroyed compared to the baseline scenario. The study argues that this impact is not expected to be the same across industries or worker-types, with workers in both the energy industry and energy-intensive industries to be the most affected (i.e., 80% of total job destruction). In producing these estimates, the model assumes that revenues from the carbon-tax issues to foster decarbonisation will be recycled, meaning that tax revenues will be redistributed.

Macro models face a number of limitations but also offer benefits that complement top-down or bottom-up approaches used to define green jobs.

  • Policy guidance: They contribute to policy options by providing projections of likely consequences.

  • Future outlook: Those approaches can help identify possible challenges ahead, which can then inform policy choices. They also provide a forward-looking perspective, which complements top-down and bottom-up approaches that mainly look at the current situation or recent changes.

  • No short run: As the main objective of these models tends to focus on the medium to long term, they do not provide insights into the short-term impact of green policies or the dynamic relationship between the short and medium term. However, the short-term effects of green policies can be very relevant for a number of reasons. First, the short-run effects might differ from the long-run impact. For example, some sectors of the economy, which see no or positive influence in the long run, might suffer strong negative shocks in the short-run due to, for example, adjustment costs. Second, short-run adjustment costs, especially if high or concentrated on specific groups of workers, companies or places, can leave long-lasting detrimental effects that make it hard to return to a positive growth path or lead to widespread labour market detachment.

  • No sub-national analysis: One limitation imposed by the prioritised global perspective is the fact that sub-national impacts are unclear. Even international analysis is limited as most models focus primarily on comparisons between world regions.

  • Limited job categories: Most macro-models offer limited information on job categories. For instance, the OECD ENV-linkages model includes five job categories, preventing the analysis of specific job gaps that serve as labour supply bottlenecks limiting the green transition.

Analysis of the green transition’s labour market implications at the regional level is scarce. In fact, no study has, to date examined the green economy or the share of green jobs at the regional level consistently across different OECD countries. Subnational data and analysis have been limited to individual countries such as the US (Vona, Marin and Consoli, 2019[40]), the UK (Valero et al., 2021[25]) (Broome et al., 2022[56]), or the Netherlands (Elliott et al., 2021[39]).

This report fills that void with novel estimates on jobs that consist, to a significant degree, of green tasks. Those estimates are at the regional level, covering annual data for a decade (2011-2020/21) in 30 OECD countries. Thus, this report defines green-task jobs as those jobs with tasks that contribute to the green transition (see sub-section below for details). This stands in contrast to jobs in green sectors and is in line with the bottom-up approach discussed above.

There are a number of reasons why the analysis in this report applies a task-based approach to defining green jobs. First and foremost, it enables regional analysis across a host of OECD countries because of the availability of regional information on employment by occupation (either via labour force surveys or direct data provision by national statistical offices). This contrasts with sector-specific approaches, for which detailed data on employment across regions is often unavailable or not available at a meaningful level of sectoral disaggregation.16 Furthermore, in supporting policy makers in designing effective policy for managing the green transition, especially with respect to active labour market or skills policies, understanding how green jobs differ from non-green or polluting ones is crucial. In order to help workers navigate the pending labour market changes, one needs to know what skills they would require and what tasks they would have to pursue if they were to transition into green jobs and help deliver the green transition. Therefore, a task-based approach might be better positioned to help design training programmes and re- or upskilling offers. Another possible advantage is that sectoral approaches might only capture a small subset of green jobs due to their narrower focus on selected sectors, which will exclude jobs in other “non-green” sectors that may also, however, contribute to the green transition. Additionally, the approach also accounts for the fact that many jobs can be partly green, i.e. consist of both green and non-green elements (tasks).

Following the work done by (Vona et al., 2018[45]), this report adopts a task-based approach to measuring and quantifying green jobs. O*NET’s Green Task Development Project (O*NET, 2010[57]) identifies occupations affected by the green economy activities and emergence of new green technology. It provides a list of individual tasks for each of these 8-digit SOC occupations and classifies these tasks as green or non-green.

O*NET identified two types of occupations with green tasks:

  1. i. Green New and Emerging Occupations – These are new occupations, which correspond to jobs that were created in the economy in response to green economy activities and the emergence of new green technologies. Examples include solar photovoltaic installers and wind energy engineers. These occupations did not previously exist in the O*NET taxonomy. All tasks performed in these occupations are considered green.

  2. ii. Green Enhanced Skills Occupations – These are occupations that underwent a significant transformation in terms of tasks performed in response to green economy activities and the emergence of new green technologies. Examples include construction and building inspectors and car mechanics. These occupations previously existed in the O*NET taxonomy. For these occupations a list of green tasks, associated with the impact of green activities and technologies, was created and added to the existing list of tasks in the occupation (non-green tasks).

As a result of the process, 1 386 green tasks were identified. All other tasks that were previously included in O*NET taxonomy are considered non-green.17

Using that information, Chapter 2 of this report computes the green intensity of an occupation, which can be broadly defined as the proportion of tasks within an occupation that are green. Therefore, this task-based approach provides a continuous measure of the greenness of occupations according to the share of tasks a worker completes on activities that contribute to the green transition. Infographic 1.2 shows examples of green and non-green tasks performed in selected occupation (non-exhaustive list of tasks), together with the green score (i.e., the share of green tasks) of the occupation. It is important to note that this measure focuses on jobs with green tasks rather than jobs in green sectors such as renewable energy. Although the two often coincide, this is not always the case.18

Based on the share of tasks in an occupation with some green tasks, a binary measure, which classifies an occupation as a green-task or non-green job, is constructed. This binary measure considers that a green-task occupation is one that consists of a considerable share of green tasks. In practice, an occupation is considered green if its green intensity is larger than 10% (0.1). This means that at least 10% of the tasks it entails are green.19 This 10% threshold also helps make estimates comparable across countries with different occupational employment information and occupation classification systems.

Indicators constructed at the occupation level are aggregated to regional (and then national) indicators. For this, occupational employment data at the regional level is used. The share of green-task jobs in a given region is simply the proportion of jobs within that region that are green. The green intensity of a region, on the other hand, is the simple weighted (by employment) average of the green intensity measures at the occupational level.

To identify polluting jobs, the approach by (Vona et al., 2018[45]) is used. This approach combines elements of the top-down and bottom-up approaches. It uses industry information regarding both employment and emissions. Furthermore, polluting jobs are a subset of non-green jobs, i.e. they entail no green tasks. Polluting jobs are determined by first identifying the occupations that are very prevalent in highly polluting industries.

Highly polluting industries are defined as those four-digit NAICS20 industries in the top 95th percentile of emissions of at least 3 (out of 8) contaminants. These eight contaminants are those controlled by the US-based Environmental Protection Agency plus CO221. This analysis yields 62 brown industries. Within these sectors, those occupations that are at least seven times more likely to be found in brown industries than in any other industry are identified as polluting occupations. For more detail and discussion on these occupations and their identification, see (Vona et al., 2018[45]).22

This analysis yields a binary indicator for polluting occupations. As with green jobs, this indicator is aggregated at the occupation level to obtain regional (and then national) indicators. The share of brown jobs in a given region is simply the proportion of jobs within that region that are brown. It complements the share of green jobs by examining the opposing side, which refers to the part of the labour market that is expected to be negatively affected by the green transition. Nevertheless, both these indicators originate from different definitions as green jobs are defined purely bottom-up while brown jobs are defined in a way that integrates elements from both the bottom-up and top-down approach.

Supporting OECD countries in their ambitions to move to a net-zero and green economy requires good data and evidence. The effective design and evaluation of policies aimed at supporting environmental objectives, green economic growth and the creation of green jobs benefit significantly from comprehensive evidence. Furthermore, setting up training and learning offers as well as targeted career guidance in supporting a just green transition requires information on green jobs and the skills they require. Finally, it requires a sound understanding of the green transition’s different impacts among workers, regions, and sectors of the economy.

A major caveat of existing evidence on the green transition is the lack of analysis of its uneven impact within countries, especially on jobs. Most analysis of jobs at risk or the contribution of the green economy focus on national data, individual countries or examine entire supranational entities such as the EU. Even less is known about the spatial dimension of green policies in the labour market. Within countries, regions differ significantly in their economic and industrial structure as well as the composition of their labour force. Therefore, they have both different exposure to risks of job loss and the potential to benefit from green job creation. Thus, the green transition might have implications for the spatial divide within countries and socio-economic disparities in and across local labour markets.

To help fill this gap, Chapter 2 of this report offers novel evidence from regions in around 30 OECD countries. It presents new estimates on the employment opportunities and risks at the subnational level while pointing out geographic differences across regions. More specifically, it shows that the shares of green jobs and polluting jobs, which face a greater risk of displacement due to the green transition, vary widely within countries. Furthermore, it documents the links between the greenness of local labour markets and a number of regional factors. Finally, it zooms in on the socio-economic impact of the green transition within local labour markets.


[48] Acemoglu, D. and D. Autor (2011), Skills, Tasks and Technologies: Implications for Employment and Earnings, Elsevier-North, https://economics.mit.edu/files/7006.

[49] Acemoglu, D. and P. Restrepo (2018), “The Race between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment”, American Economic Review, Vol. 108/6, pp. 1488-1542, https://doi.org/10.1257/aer.20160696.

[20] Albrizio, S., T. Koźluk and V. Zipperer (2017), “Environmental policies and productivity growth: Evidence across industries and firms”, Journal of Environmental Econonics and Management, Vol. 81, pp. 209-226, https://doi.org/10.1016/j.jeem.2016.06.002.

[47] Autor, D. (2013), “The “task approach” to labor markets: an overview”, Journal for Labour Market Research volume, Vol. 46, pp. 185-199, https://doi.org/10.1007/s12651-013-0128-z.

[46] Autor, D., F. Levy and R. Murnane (2003), “The Skill Content of Recent Technical Change: An Empirical Exploration”, Quarterly Journal of Economics, Vol. 118, pp. 1279-1334, https://doi.org/10.1162/003355303322552801.

[33] Bowen, A. and B. Hancké (2019), The Social Dimensions of ‘Greening the Economy’: Developing a taxonomy of labour market effects related to the shift toward environmentally sustainable economic activities.

[27] Bowen, A. and K. Kuralbayeva (2015), Looking for green jobs: the impact of green growth on employment.

[41] Bowen, A., K. Kuralbayeva and E. Tipoe (2018), “Characterising green employment: The impacts of `greening’on workforce composition”, Energy Economics, Vol. 72, pp. 263-275, https://doi.org/10.1016/j.eneco.2018.03.015.

[31] Bowen, A. and K. Kuralbayev (2015), Looking for green jobs: the impact of green growth on employment.

[56] Broome, M. et al. (2022), Net zero jobs: The impact of the transition to net zero on the UK labour market.

[14] Brucal, A. and A. Dechezleprêtre (2021), “Assessing the impact of energy prices on plant-level environmental and economic performance: Evidence from Indonesian manufacturers”, OECD Environment Working Papers, No. 170, OECD Publishing, Paris, https://doi.org/10.1787/9ec54222-en.

[54] Chateau, J., R. Bibas and E. Lanzi (2018), “Impacts of Green Growth Policies on Labour Markets and Wage Income Distribution: A General Equilibrium Application to Climate and Energy Policies”, OECD Environment Working Papers, No. 137, OECD Publishing, Paris, https://doi.org/10.1787/ea3696f4-en.

[55] Chateau, J. and E. Mavroeidi (2020), “The jobs potential of a transition towards a resource efficient and circular economy”, OECD Environment Working Papers, No. 167, OECD Publishing, Paris, https://doi.org/10.1787/28e768df-en.

[16] Dechezleprêtre, A., D. Nachtigall and B. Stadler (2020), “The effect of energy prices and environmental policy stringency on manufacturing employment in OECD countries: Sector- and firm-level evidence”, OECD Economics Department Working Papers, No. 1625, OECD Publishing, Paris, https://doi.org/10.1787/899eb13f-en.

[18] Dechezleprêtre, A., D. Nachtigall and F. Venmans (2018), “The joint impact of the European Union emissions trading system on carbon emissions and economic performance”, OECD Economics Department Working Papers, No. 1515, OECD Publishing, Paris, https://doi.org/10.1787/4819b016-en.

[50] Dierdorff, E. and J. Norton (2011), Summary of Procedures for O*NET Task Updating and New Task Generation, https://www.onetcenter.org/reports/TaskUpdating.html (accessed on 20 August 2022).

[51] Dierdorff, E. et al. (2009), Greening of the World of Work: Implications for O*NET-SOC and New and Emerging Occupations.

[21] Dlugosch, D. and T. Kozluk (2017), “Energy prices, environmental policies and investment: Evidence from listed firms”, OECD Economics Department Working Papers, No. 1378, OECD Publishing, Paris, https://doi.org/10.1787/ef6c01c6-en.

[19] Dussaux, D. (2020), “The joint effects of energy prices and carbon taxes on environmental and economic performance: Evidence from the French manufacturing sector”, OECD Environment Working Papers, No. 154, OECD Publishing, Paris, https://doi.org/10.1787/b84b1b7d-en.

[38] ECO Canada (2021), From Recession to Recovery: Environmental Workforce Needs, Trends and Challenges - Updated Labour Market Outlook to 2025.

[53] Elliott, R. et al. (2021), “Eco-Innovation and Employment: A Task-Based Analysis”, IZA Discussion Papers, Vol. 14028/Elliott, Robert J. R. & Kuai, Wenjing & Maddison, David & Ozgen, Ceren, 2021. "Eco-Innovation and Employment: A Task-Based Analysis," IZA Discussion Papers 14028, Institute of Labor Economics (IZA)., https://www.iza.org/publications/dp/14028/eco-innovation-and-employment-a-task-based-analysis.

[39] Elliott, R. et al. (2021), “Eco-innovation and employment: A task-based analysis.”, IZA Discussion Papers, Vol. 14028.

[60] European Commission (2021), Council recommendation on ensuring a fair transition towards climate neutrality.

[5] European Commission (2021), Staff working document (2021) 452.

[32] Eurostat (2022), Statistics Explained. Environmental Economy – Statistics on Employment and Growth.

[35] Eurostat (2021), Environmental economy – statistics on employment and growth.

[43] Eurostat (2016), Environmental goods and services sector accounts — Practical guide — 2016 edition, Publications Office of the European Union, https://doi.org/10.2785/688181.

[22] Garsous, G., T. Koźluk and D. Dlugosch (2020), “Do energy prices drive outwards FDI? Evidence from a sample of listed firms”, The Energy Journal, Vol. 41/3,, Vol. 41/3, https://doi.org/10.5547/01956574.41.3.ggar.

[13] Gray, W. et al. (2014), “Do EPA regulations affect labor demand? Evidence from the pulp and paper industry”, Journal of Environmental Economics and Management, Vol. 68/1, pp. 188–202, https://doi.org/10.1016/j.jeem.2014.06.002.

[12] Greenstone, M. (2002), “The impacts of environmental regulations on industrial activity: evidence from the 1970 and 1977 Clean Air Act amendments and the census of manufactures”, Journal of Political Economy, Vol. 110/6, pp. 1175-1219, https://doi.org/10.1086/342808.

[29] ILO (2022), What are green jobs according to the ILO?.

[34] ILO / CEDEFOP (2011), Skills for Green Jobs - A Global View.

[52] International Monetary Fund (2022), World Economic Outlook: War Sets Back the Global Recovery.

[36] IRENA and ILO (2022), Renewable energy and jobs: Annual review 2022.

[58] JRC (2021), Labour Markets and the Green Transition: a practitioner’s guide to the task-based approach, https://doi.org/10.2760/65924.

[23] Koźluk, T. and C. Timiliotis (2016), “Do environmental policies affect global value chains?: A new perspective on the pollution haven hypothesis”, OECD Economics Department Working Papers, No. 1282, OECD Publishing, Paris, https://doi.org/10.1787/5jm2hh7nf3wd-en.

[4] Kruse, T. et al. (2022), “Measuring environmental policy stringency in OECD countries: An update of the OECD composite EPS indicator”, OECD Economics Department Working Papers, No. 1703, OECD Publishing, Paris, https://doi.org/10.1787/90ab82e8-en.

[1] Lenton, T. et al. (2008), “Tipping elements in the Earth’s climate system”, Proceedings of the National Academy of Sciences, Vol. 105/6, pp. 1786-1793, https://doi.org/10.1073/pnas.0705414105.

[2] Lenton, T. et al. (2019), “Climate tipping points- too risky to bet against”, Nature, Vol. 575, https://doi.org/10.1038/d41586-019-03595-0.

[42] Martinez-Fernandez, C., C. Hinojosa and G. Miranda (2010), “Greening Jobs and Skills: Labour Market Implications of Addressing Climate Change”, OECD Local Economic and Employment Development (LEED) Papers, No. 2010/2, OECD Publishing, Paris, https://doi.org/10.1787/5kmbjgl8sd0r-en.

[57] O*NET (2010), Green Task Development Project, https://www.onetcenter.org/reports/GreenTask.html (accessed on 15 May 2022).

[44] OECD (2023), Regional Industrial Transitions to Climate Neutrality, OECD Regional Development Studies, OECD Publishing, Paris, https://doi.org/10.1787/35247cc7-en.

[6] OECD (2022), “Assessing environmental impact of measures in the OECD Green Recovery Database”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://doi.org/10.1787/3f7e2670-en.

[3] OECD (2022), Climate Tipping Points: Insights for Effective Policy Action, OECD Publishing, Paris, https://doi.org/10.1787/abc5a69e-en.

[10] OECD (2022), Future-Proofing Adult Learning in Berlin, Germany, OECD Reviews on Local Job Creation, OECD Publishing, Paris, https://doi.org/10.1787/fdf38f60-en.

[26] OECD (2022), OECD Employment Outlook 2022: Building Back More Inclusive Labour Markets, OECD Publishing, Paris, https://doi.org/10.1787/1bb305a6-en.

[17] OECD (2021), Assessing the Economic Impacts of Environmental Policies: Evidence from a Decade of OECD Research, OECD Publishing, Paris, https://doi.org/10.1787/bf2fb156-en.

[8] OECD (2021), OECD Skills Outlook 2021: Learning for Life, OECD Publishing, Paris, https://doi.org/10.1787/0ae365b4-en.

[7] OECD (2021), “The OECD Green Recovery Database: Examining the environmental implications of COVID-19 recovery policies”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://doi.org/10.1787/47ae0f0d-en.

[59] OECD (2020), “Making the green recovery work for jobs, income and growth”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://doi.org/10.1787/a505f3e7-en.

[9] OECD (2018), Job Creation and Local Economic Development 2018: Preparing for the Future of Work, OECD Publishing, Paris, https://doi.org/10.1787/9789264305342-en.

[24] OECD (2017), Employment Implications of Green Growth: Linking jobs, growth, and green policies.

[30] OECD/Cedefop (2014), Greener Skills and Jobs, OECD Green Growth Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264208704-en.

[37] ONS (2020), Low carbon, renewable energy economy, UK: 2018.

[15] Popp, D. et al. (202), The Employment Impact of Green Fiscal Push: Evidence from the American Recovery Act, https://doi.org/10.3386/w27321.

[28] UNEP, ILO, IOE, ITUC (2008), Green Jobs: Towards Decent Work in a Sustainable, Low Carbon World.

[25] Valero, A. et al. (2021), Are ‘green’ jobs good jobs?.

[40] Vona, F., G. Marin and D. Consoli (2019), “Measures, drivers and effects of green employment: evidence from US local labor markets, 2006–2014”, Journal of Economic Geography, Vol. 19, pp. 1021–1048, https://doi.org/10.1093/jeg/lby038.

[45] Vona, F. et al. (2018), “Environmental Regulation, Green Skills: An Empirical Exploration.”, Journal of the Association of Environmental and Resource Economists, Vol. 5/4, pp. 713-753, https://doi.org/10.1086/698859.

[11] Walker, W. (2011), “Environmental Regulation and Labor Reallocation: Evidence from the Clean Air Act.”, American Economic Review, Vol. 101/3, https://doi.org/10.1257/aer.101.3.442.

Estimating the share of green jobs depends on detailed employment data at the regional level. National statistical offices provide information on the total employment by occupation either directly or via micro data from labour force surveys. While many countries use International Standard Classification of Occupations (ISCO) developed by the ILO, several countries such as the US, Canada or Australia have their own classification systems. The level of detail of occupational classifications systems varies across countries as these are built according to the needs and objectives of each country. Nevertheless, one thing is constant with regard to their structure: occupation codes are constructed in such a way that each digit represents a category or subcategory. The first digits usually represent a major or broad category, the second digit represents a subcategory of the first, and so on. For most classification systems, one digit represents one category or sub-category, except for the US where, for example, the major category uses two digits as they have more than 10 major categories.

In addition, the structure of the datasets available on employment by occupation differs across OECD countries. For the empirical analysis in this report, microdata from the respective Labour Force Surveys (LFS) was used for European countries and Canada, while the analysis for the US and Australia was based on their aggregated datasets on occupational employment. The latter reports the estimated employment per occupation and TL-2 region.23 Information on occupational employment for European countries is less detailed than in the US, Canada and Australia. The European LFS (EU-LFS) provides information up to three ISCO digits.24 This means that estimates for the greenness of occupations need to be aggregated from more detailed 4-digit to less detailed 3-digit ISCO occupations, which can lead to an overestimation of the share of green jobs (see (JRC, 2021[58])).

To ensure comparability across countries, occupational data is aggregated for countries with more detailed information to a level that is comparable across countries, without sacrificing detail. Therefore, for the US and Canada, information is aggregated to a lower number of digits to make it more comparable with European data, which is only available at the three-digit level for ISCO-08. The digits of aggregation were chosen in such a way that the number of unique occupations is similar across countries without losing too much detail. Based on the examples of the US and Canada, choosing a lower digit of occupational detail does not lead to significant changes in the estimates (see Annex Chapter 2).

O*NET provides data for SOC-2010, therefore the green and brown indicators were computed for this occupation classification and then assigned to the rest of the occupation classifications using one or more crosswalks provided by national statistical offices. Crosswalks are matchings that provide a way to translate occupational classifications and therefore merge labour data from different countries. These matchings are not always one-to-one and can be either full or partial; this creates an issue with aggregation (see (JRC, 2021[58])), as the green score of two or more origin-occupations needs to be combined in order to assign a unique indicator to the destination-occupation. An approach that mitigates the impact of this is discussed in Chapter 2.

The green indicators are based on task information provided by the Occupational Information Network (O*NET). O*NET provides several relational datasets that describe occupations, tasks and task ratings, among other things. The latest available dataset (version 24.1 from Nov. 2019) with information on green tasks is used. This dataset provides information for 974 occupations (out of 1 110) using SOC-2010 occupation classification at the 8-digit level.

Information on emissions is extracted from the National Emission Inventory database and the Greenhouse Gas Emission for Large Facilities (2011), both of which are maintained by the Environmental Protection Agency (2011). This data is aggregated to the main 4 digit NAICS code.


← 1. Measures aimed at supporting the green transition in recovery programmes include: grants, loans and tax relief directed towards green transport, circular economy and clean energy research, development and deployment; financial support to households and businesses for energy efficiency improvements and renewable energy installations; new funding and programmes to create jobs and stimulate economic activity through ecosystem restoration; Control of invasive alien species and forest conservation (OECD, 2020[59]).

← 2. None of the skills and training programmes included in the recovery plans appear to have a negative environmental impact.

← 3. The authors contrast the effect of increases in energy prices with those of automation and globalisation, stating that they are 30% and 80% smaller, respectively.

← 4. A recurrent theme in work on the green transition is the lack of evidence and clear definitions of green jobs.

← 5. Examples include environmental protection activities, aimed at reducing and preventing greenhouse emissions or other harmful environmental impacts, and resource management activities related to energy.

← 6. These sectors were developed in consultation with LCREE experts from various government departments, who are also data users.

← 7. Note that for lower-level disaggregation of the survey data, for example, country or region, the accuracy of estimates can be low, limiting the level of detail of the results. In addition, the survey only collects data on direct LCREE activity. Furthermore, this activity does not have to be the main activity of a business for a business to be counted as active in the LCREE economy.

← 8. Notably, the survey has registered no significant change since 2015 in employment. Energy efficient products (excluding energy efficient lighting) and the low emissions vehicles sectors remained the largest sectors in the LCREE economy in 2020. The energy efficient products sector is particularly important in terms of employment as it accounts for 42% of employment.

← 9. The challenges of defining a "green job" - Office for National Statistics (ons.gov.uk).

← 10. Task-based analysis has been applied to many other labour economics issues such as remote working potential.

← 11. For more information on the data available and their structure, visit the O*NET data dictionary https://www.onetcenter.org/dictionary/24.1/excel/

← 12. O*NET includes such information for each 8-digit SOC (the United States’ Standard Occupational Classification) occupation.

← 13. Other studies instead examine the share of jobs with any green task or the share of jobs with a significant proportion of green tasks.

← 14. The authors also show that workers in green jobs enjoy a wage premium and have more secure work contracts.

← 15. Note that it is available at a detailed level for the USA, but by using crosswalks this detailed information can be translated and used in other countries.

← 16. Task-based approaches might also avoid an “arbitrary” designation of green sectors since they are based on the content of individual jobs.

← 17. Note that O*NET also provides and importance score for each task. The weighted proportion of green tasks yields qualitatively the same results as the unweighted shares and has the drawback of shrinking the dataset as the importance variable is not available for all tasks.

← 18. The measure of green jobs used in this chapter treats occupations with the same tasks equally, regardless of their sector. For example, accountants working in different sectors such as renewable energy or manufacturing would be consistently defined as non-green given the lack of greenness of their tasks.

← 19. The annex of Chapter 2 provides more information on this and alternative thresholds and further discusses this choice.

← 20. The North American Industry Classification System.

← 21. Specifically, these are: CO, VOC, NOx, SO2, Pm10, PM2.5, lead and CO2.

← 22. Web Appendix C, especially for data and other methodological details.

← 23. Information on employment is not reported for the Northern Territories in Canada, as access to this micro-data is limited.

← 24. In addition, note that Poland and Slovenia were excluded from the analysis given that for these countries the occupational codes are disseminated at 2-digit level only. This does not allow for sufficiently precise identification of green jobs.

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 2023

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.