In this section, we review findings and quantitative metrics from the literature as they relate to how many jobs can be created by deploying renewable energy and energy efficiency compared to supporting investment in fossil fuel energy production. We also consider net employment and economic effects of transitioning to a low carbon energy system at a national and regional scale.
In our review, most of the identified studies quantifying job creation by technology estimated gross employment. Therefore, following an approach taken by Blyth et al. (2014), we provide an approximate assessment of net jobs impacts by comparing gross employment factors for renewable energy and energy efficiency with those for fossil fuels. Sections 3.1. and 3.2. review gross employment factors according to different metrics identified in the reviewed literature, relating to the scale of technological activity and level of investment.
Evidence identified in our review on net job creation applies predominantly to international, national and regional scales, and tends to focus more on net effects as a result of low carbon transition across the power sector or wider energy system rather than exclusively as a result of deploying single technologies. In section 3.3 we consider studies which estimate how many jobs are created as a result of decarbonising the energy sector, net of jobs lost in more carbon intensive, fossil fuel activities. These studies attempt to characterise how net job creation, destruction, and linked economic impacts might be manifested at a national or regional level, e.g. as a result of climate change mitigation targets or Nationally Determined Contributions (NDCs). Section 3.4. discusses findings from the literature on the economic impact of low carbon energy job creation.
3.1. Comparing gross job creation of energy technologies and interventions by scale of activity
In Figures 1a and b, we have collated data from our review on job creation per installed capacity for different life stages of electricity generation technologies: manufacturing (in job-years/GW); construction and installation (in job-years/GW); and operation and maintenance (in jobs/GW). One job-year is one full-time job for one person lasting for a year (Dufo-López, Cristóbal-Monreal and Yusta, 2016). These units show the number of jobs created annually and are used to characterise manufacturing and construction/ installation jobs which are required in the first few years of projects and at a project-level tend to be shorter-term in nature compared to ongoing employment in operation/ maintenance. Operation and maintenance jobs are typically expressed in jobs/MW, as it is assumed that these jobs are more permanent in nature and should last over the lifetime of energy technologies (Ram et al., 2022). These more granular estimates broken down by technology life stage are only available in a limited range of studies in the literature. In Figures 1a and b, we have converted jobs and job-years from the original datasets so they are normalised by GW rather than MW. This helps to aid comparison of the potential volume of jobs created by technology.
Most of the documents from which this dataset has been derived estimate direct and/or indirect employment impacts using an analytical approach (see notes to Figures 1a and b). Two studies apply an input-output model, but only estimates from the global analysis of Jacobson et al. (2017) account for direct, indirect and induced jobs. Therefore, the data shown in these figures relates mainly to employment created directly in the activities shown, or in associated supply chains.
A number of key observations can be made with respect to this dataset. Firstly, there is a higher level of manufacturing job creation per GW for several types of renewables (including offshore wind, small hydro and solar PV) compared to gas or coal-fired electricity generation (Figure 1a). The average estimate of job-years created per unit of installed capacity is particularly high for offshore wind (16,800 job-years per GW). The potential to make use of high employment factors for manufacturing depends on the presence of a renewables manufacturing base in any given country. Moreover, many construction/installation and operation/maintenance jobs may effectively be exported overseas depending on the development and size of an export market for manufactured renewables. Simas and Pacca (2014) observe that the standard metric of manufacturing job-years/unit of installed capacity in a particular year may be misleading since it does not account for the proportion of imports or exports. It would therefore be possible for a country to have a very high index if installed capacity is low due to most manufactured technologies being exported.
The evidence on construction and installation (Figure 1a) indicates that this activity creates the most jobs for solar PV, biomass and small hydro, around between 15,000 and 18,000 job-years per GW. Construction of natural gas power plants is associated with the lowest level of employment (2,500 job-years per GW). Wind farm installation performs relatively modestly (averaging 4,200 and 7,200 job-years per GW for onshore wind and offshore wind respectively). Ram et al. (2022) observe that demand for construction, installation, and operation and maintenance jobs tends to be created locally, and these activities are therefore a fairly good indication of the potential to generate jobs within a country or region. However, a key uncertainty is the extent to which labour and supply chain services may be imported from other countries.
Figure 1b suggests that operation and maintenance is associated with the highest number of jobs over technology lifetimes for small hydro (1,600 jobs/GW) and biomass (1,100 jobs/GW). Natural gas and coal have the lowest employment factors for operation and maintenance (130 and 155 jobs/GW respectively).
The estimates of job creation presented so far pertain to electricity generation technologies. The review has found only limited examples of equivalent data for heating, low carbon or otherwise. Ram et al. (2022) include employment factors for a range of heating technologies in a recent global scenario analysis based on an energy transition to 100% renewables, with a high level of electrification, from 2015 to 2050. The authors present job intensities by manufacturing, construction and installation, and operation and maintenance activities, expressed in units of jobs (or job-years) created per megawatts of thermal (MWth) installed capacity. In general, these particular estimates suggest that CIM activities are associated with a higher number of job-years per megawatts thermal (MWth) for low carbon heating technologies and fuels sources (e.g. individual heat pumps, district heating sourced from heat pumps or biomass, or energy from waste CHP) compared to gas or oil heating. Estimates for employment creation in operation and maintenance (jobs per MWth) are very similar for individual oil, gas, electric and heat pump technologies, but highest for solar thermal and individual biomass heating systems.
3.2. Comparing gross job creation of energy technologies and interventions by level of investment
Figure 2 summarises the evidence identified across 14 studies which estimate the number of gross jobs created per USD million invested for different energy technologies. This represents total investment, across public and private sectors, and is largely based on studies which model estimates of job creation potential using input-output analysis (see notes to Figure 2). Half of the studies account for direct and indirect employment effects only, with the other half representing indirect and induced employment multipliers as well as direct jobs impacts. Employment factors extracted from five documents reviewed capture investment only or mainly in the CIM phase, with a further four studies quantifying jobs per investment in CIM and O&M phases. However, the remaining six studies are unclear on which phase/s are represented.
The comparison presented in Figure 2 suggests that renewables or energy efficiency can generate more jobs per US dollars invested than fossil fuel generation or nuclear power. Fossil fuel generation creates five jobs per USD million invested on average across the various studies, compared to eight jobs / USD million for nuclear power and 15 jobs / USD million for the renewable energy technologies shown in the chart. Building energy efficiency demonstrates the highest job creation potential per level of investment, creating an average of 22 jobs / USD million.
The relative employment outcomes of investing in higher carbon or low carbon technologies or energy efficiency may reflect a number of factors, including the ratio of spending on local content versus imports, the ratio of spending on labour as opposed to capital, and average wage levels for jobs in particular technology sectors and their supply chains (Pollin and Garrett-Peltier, 2009). The extent to which high employment factors are desirable as a policy objective is a point to which we return later. Lower employment factors for fossil fuel power generation may also be associated with greater market maturity of well established, conventional technologies compared to renewables or energy efficiency.
3.3. Does a shift to low carbon energy create jobs? Estimates of net job creation or destruction at a national or regional scale
Global analyses are in broad agreement that the net employment effects of climate change mitigation policy and a low carbon energy transition will be positive, estimating that by 2030 and 2050, more jobs will be created by a shift to low carbon, renewable energy than the number of jobs displaced from decommissioning fossil-fuel power plants (e.g. (ILO, 2019; Jacobson et al., 2019; IRENA, 2021; Pai et al., 2021). Nevertheless, key questions for policy makers are how job creation and job destruction will be geographically distributed, and how the impacts of this transition will fall in particular countries and regions at different points in time.
In Table 2, we summarise the scope, methods, and findings from 14 studies identified in our review which estimate how decarbonisation in the energy sector affects the overall number of jobs created and displaced at a national and/or regional scale. Taken together, these studies represent evidence from Argentina, Brazil, China, India, Japan and eight European countries (including the UK). Half of the studies in Table 2 focus on the impacts of decarbonisation in the power sector alone, but several studies consider other energy sectors such as heat, transport and industry (Sievers et al., 2019; Füllemann et al., 2020; Lee et al., 2022). An important variation between the studies relates to how they use counterfactuals to calculate net employment impacts. Across the 14 documents, overall job creation or destruction is typically estimated as being net of avoided investment in fossil fuels (either domestic industries or imports), or net of reference (e.g. continuation of current policy) scenarios. In terms of methods, input-output analysis is most commonly used (five studies) followed by CGE models (three studies), including Perrier and Quirion (2018) who combine both model types. Sievers et al. (2019) and Lee et al. (2022) apply a macro-economic and macro-econometric model respectively. Five of the identified documents account for direct, indirect and induced employment; five studies consider direct and indirect but not induced jobs; and two studies estimate direct jobs effects alone[2].
The documents reviewed in Table 2 address the net employment impact of climate change mitigation or decarbonisation scenarios over the next one to three decades, achieving higher shares or capacities of renewable energy and/ or substituting fossil fuels with renewables and energy efficiency. Nine of the 14 studies conclude that, at a national scale, there is likely to be positive net job creation overall from replacing fossil fuels with renewables / energy efficiency or as a result of energy sector decarbonisation. Whilst some of these additional jobs are relatively short term since they relate to the construction and installation phase, Arvanitopoulos and Agnolucci (2020) find that in the UK, there is still likely to be an increase in overall jobs in the long term. Perrier and Quirion (2018) observe that both an input-output and CGE model generate positive net employment outcomes in France for a €1 billion investment in solar PV installation or building weatherisation.
Three studies find that a mixture of job creation and displacement effects could result in China, India and Italy respectively (Cai et al., 2017; Mu et al., 2018; Sharma and Banerjee, 2021). Mu et al. (2018) model alternative ways of financing a feed-in tariff in China to attain a 1 Terawatt-hour (TWh) expansion target for solar PV and wind power. These lead to direct and indirect job creation but induced job losses, with net job creation arising overall in three scenarios and modest net job destruction in one scenario for wind power compared to a reference scenario.
Two studies report overall net job destruction or indicate a significant risk of such an outcome (Baran, Szpor and Witajewski-Baltvilks, 2020; Le Treut et al., 2021). Le Treut et al. (2021)’s CGE model analysis concludes that power sector decarbonisation in Argentina could cause a small proportion of job losses across the economy (-0.5 to -0.7%), and that net job creation in energy, construction and manufacturing would be offset by job destruction in other economic sectors. Additionally, significant structural change would be implied within the energy sector due to the fast decline of fossil fuels and rapid growth of low carbon power generation (Ibid.). Baran, Szpor and Witajewski-Baltvilks (2020) caution that a transition away from coal production and use in Poland could lead to net job destruction, without a well-managed plan to help displaced coal miners transition to new roles in renewable energy and energy efficiency sectors.
The aggregate balance of job creation and destruction across a national economy also depends on the extent to which new and displaced energy sectors utilise labour within or outside a given country. Cai et al. (2017) combine input-output and analytical models to evaluate employment impacts of solar PV, onshore wind, hydropower and bioenergy in Italy from 2006 to 2014. The authors contend that lower job creation for renewable energy CIM activities compared to three potential alternative investments in other economic sectors is due to many renewable energy components being imported. Conversely, O&M in Italy’s renewables sector used mostly local goods and services and generated higher employment factors than the counterfactual investments. Similarly, in an input-output analysis of a transition to low carbon electricity generation in Japan by 2050, Kuriyama and Abe (2021) find that a shift to renewable energy could lead to net job creation within the country by reducing the need for fossil fuel imports and reliance on overseas labour. Making the most of this domestic job generating opportunity depends on effective national and local government support providing training, recruiting workers and developing a labour force with requisite skills in renewable energy sectors (Ibid.).
Our review identified limited examples of studies which attempt to quantify the regional distribution of low carbon energy employment. Sharma and Banerjee (2021) find that retiring coal power plants in India and attaining a national 100GW solar PV capacity target would have mixed net jobs impacts across different regions. While the authors’ analytical model suggests an outcome of positive net job creation in six Indian states, net job destruction is projected in six other states, particular those which have considerable coal mining activity. Kuriyama and Abe (2021) conclude that transitioning to renewable electricity generation in Japan could be particularly favourable for creating stable, long-term jobs in O&M in rural areas. The study illustrates that early planning is required for conventional power plant phase out to avoid a surplus of workers, accounting for regional differences in the impact of this phase out and new opportunities available for renewables, energy efficiency or other low carbon energy sectors. Elsewhere, Sievers et al. (2019) investigate a low carbon transition scenario across the energy system using a macroeconomic model and an economic impact assessment model; their approach includes a regional breakdown of employment impacts. While the study finds that a low carbon energy transition could create around 1% more jobs in total nationally from 2010 to 2030, northern and eastern German states would likely gain the most economically given their high suitability of locations for siting renewable energy, and that they are less affected by the phase out of fossil fuel power plants.
3.4. Economic impacts of net job creation or destruction at a country or regional scale
The evidence reviewed in Sections 3.1 and 3.2 indicates that renewables and energy efficiency can generate more jobs than fossil fuels for the same level of investment. It does not automatically follow, however, that prioritising investment in these technologies will result in higher employment for any given national economy in the long-term. Investing in options with higher job multipliers such as certain renewable energy CIM activities or installing building energy efficiency interventions makes sense in a depressed economy in which aggregate demand is low compared to potential supply of goods and services (creating a so-called “Keynesian output gap”). In such a context stimulating additional employment in labour-intensive sectors is very likely to lead to higher overall employment (Blyth et al., 2014).
While the circumstances around the COVID-19 economic recession and the 2009 financial crisis are very different, evidence from the 2009 crisis indicates that the green measures (e.g. in renewable energy infrastructure) forming part of the recovery stimulus created more jobs than conventional stimulus measures (Allan et al., 2020). The CCC recommend that in the short term, “green stimulus policies can be economically advantageous compared to traditional fiscal stimuli. They tend to have higher short run multipliers and higher numbers of jobs created” (CCC, 2020, p. 141). Domestic construction projects such as insulation retrofits or building wind turbines may be particularly favourable and less prone to offshoring services overseas. In comparison to the renewable energy CIM phase, employment in the operation and maintenance of renewable power generation technologies is typically more permanent, with potential to last over technology lifetimes (Ram et al., 2022).
There is a debate in the literature identified in our review around the extent to which policies supporting renewable energy may contribute to longer term economic growth, notwithstanding short-term employment and growth benefits (e.g. Jaraite, Karimu and Kazukauskas, 2017; Safwat Kabel and Bassim, 2019). While we have not found extensive evidence to definitively answer this question, identified studies suggest that subject to geographic and contextual variations, low carbon energy shifts may promote modest economic growth effects and are unlikely to be detrimental to economies at a national level. Several of the studies presented in Table 2 suggest that supporting a low carbon energy transition is compatible with longer-term economic growth over the next 10-30 years. For example, Stamopoulos et al. (2021) note that increased investments in solar PV and wind power are key to their modelled outcome that the Greek National Energy and Climate Plan could contribute EUR 6.8 billion to Greek GDP by 2030. Similarly, Sievers et al. (2019) project a 1.6% increase in GDP versus a reference scenario in Germany from 2010 to 2030 as a result of energy sector decarbonisation. Elsewhere, Lee et al. (2022) analyse two Net Zero policy scenarios for Japan (with and without nuclear power phase out), finding that by 2050 decarbonisation across the energy system could add 4%-4.5% to GDP compared to a reference scenario.
On the other hand, and linked to the challenge of allocating displaced coal miners to new low carbon energy jobs, there is a significant risk that replacing coal power with renewables and energy efficiency in Poland could lead to reductions in GDP as well as net labour losses (Baran, Szpor and Witajewski-Baltvilks, 2020). The contribution of different renewable energy phases to net employment and value added in a given country may also depend upon the extent to which goods, services and labour are sourced locally. Cai et al. (2017) find that due to significant component imports, renewable energy CIM in Italy contributed less to value added historically than counterfactual scenarios. By contrast, the predominantly local content and higher labour intensities of renewables O&M led to higher value added than the counterfactuals. Pegels and Lütkenhorst (2014, p.529) found that renewables component manufacturing in Germany was “often located in the traditional industrial centres”, some with the “highest unemployment ratios nationwide”.
While well established, high carbon technologies and sectors may be close to their limits in terms of additional innovation and economic productivity gains, investment in less mature, faster growing low-carbon technologies such as renewables could contribute more to productivity through greater scope for innovation and learning by doing (CCC, 2019). Careful planning is required to optimise the design of subsidy schemes for renewable energy technologies to lower risks of increasing income inequality or regressive distributional impacts on low-income households, such as those which have been modelled in studies of Ireland and Germany respectively (Többen, 2017; Farrell, O’Donoghue and Morrissey, 2020). When designing stimulus programmes, it makes sense from a Keynesian perspective to support technologies and projects that have a positive, long-term impact on economic productivity beyond the timeframe of the direct stimulus effects. In a longer-term context, labour intensity is not in and of itself economically advantageous. If it implies lower levels of labour productivity (economic output per worker), then it could adversely affect prospects for long-term economic growth. On the other hand, critiques of Green New Deal approaches argue that they perpetuate a narrative around the need for full employment, high production, high consumption economies which are incompatible with conserving finite resources and minimising ecological footprint (Green, 2022). Developing on from Blyth et al. (2014), we therefore contend that policy should not be focused on maximising jobs per unit of investment in the long run. Rather, policy decisions should be based on whether investments can contribute to an economically efficient transition towards effective climate change mitigation, taking account of the need for a just transition, wider ecological impacts and energy security considerations.
[2] Two documents do not provide information on the representation of direct, indirect and/or induced jobs.