Estimating the Jobs Impacts of Different Carbon Tax Reforms This work is a product of the staff of The World Bank. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank or IFC, its Board of Executive Directors, or the gov- ernments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denomi- nations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Contents Introduction .................................................................................................................................................. 4 Carbon taxes, revenue recycling, and clean energy investment ................................................................... 5 The modeling approach: MINDSET ............................................................................................................... 7 Caveats .......................................................................................................................................................... 8 Scenario description ................................................................................................................................... 10 Results ......................................................................................................................................................... 12 Overview global results ........................................................................................................................... 12 Economic and employment effects by region and sector ....................................................................... 14 Summary of employment effects across countries................................................................................. 18 Conclusions and outlook ............................................................................................................................. 20 References................................................................................................................................................... 21 Table of Figures Figure 1: Employment effects by sector and income group, in relative differences of sectoral employment to baseline under ‘Gov’ scenario ................................................................................................................ 15 Figure 2: Employment effects by sector and income group, in relative differences of sectoral employment to baseline under scenario PR .................................................................................................................... 16 Figure 3: Employment effects by sector and income group, in relative differences of sectoral employment to baseline under scenario PR+RE .............................................................................................................. 17 Figure 4: Impacts of revenue-neutral carbon tax in payroll tax cut (PR) and government expenditure (GOV) scenarios on GDP and employment, relative difference to reference. The color represents the country region based on World Bank classification. ................................................................................... 19 Figure 5: Comparison of effects across countries. Comparing the PR scenario with the Gov scenario ..... 20 List of Tables Table 1: Granular policy & climate scenarios modeled simultaneously for 120 sectors & 164 countries: ... 7 Table 2: Scenario overview ......................................................................................................................... 10 Table 3: Result dimensions.......................................................................................................................... 12 Table 4: Impacts of different scenarios on employment, relative difference to reference......................... 13 Table 5: Impacts of different scenarios on GDP, relative difference to reference....................................... 13 Table 6: Impacts of different scenarios on output, relative difference to baseline. ................................... 13 Table 7: Median values and confidence intervals for the PR scenario........................................................ 14 2 Acknowledgements This note was prepared by Ulrike Lehr (Senior Energy and Environmental Economist, Jobs Group), Ira Dorband (Economist, EFI) and Gilang Hadardi (ETC, EFI). The note was reviewed by Elizabeth Ruppert Bulmer (Lead Economist, Jobs Group) and Hector Pollitt. The activity is funded by the WB Climate Support Facility. 3 Introduction There is broad consensus that sufficiently high carbon prices are important in moving countries to low- carbon pathways. Carbon pricing disincentivizes carbon-intensive activities, and the revenues generated could finance climate actions as well as support households and firms. When climate policies – such as renewable energy investments and energy efficiency programs – are designed in a way that stimulates job creation, a double dividend can be achieved: lower greenhouse gas (GHG) emissions and more jobs. The analysis reported here assesses different policy designs around a carbon tax reform, focusing on the direct and indirect implications for jobs. By filling this important knowledge gap, governments will be empowered to take better-informed and possibly more ambitious action. It is important for many governments to compare expected economic outcomes with outcomes in other countries, which they consider their peer, and to understand better why results might differ. Economic outcomes can be compared using several indicators. Often, governments want to compare with their regional neighbors, for instance, to find out if investment in clean energy and a carbon tax have the same effects in Egypt and Morocco. As countries in a region often differ in economic progress and income level, another interesting comparison might be within the same income group. More aspira- tional would be a comparison with a member of the next higher income group. Physical endowment, especially with fossil fuels, can influence the resulting employment effects under different scenarios. Hence, a comparison across producers of crude and processed fuels is suggested in this report. The possibility space is mapped with different scenarios, applying identical carbon taxes and revenue recycling schemes to all countries to allow for cross-country comparisons. After this introduction, a brief literature review provides the background for the policy scenario designs. It explains the benefits of carbon taxes, as well as the expected impacts of different revenue recycling schemes. The scenarios are described in more detail in section 4, preceded by a brief description of the modeling approach and a short section on caveats regarding the approach in sections 4 and 5. Section 6 explores the results along several dimensions, such as geographic regions, country income groups, and resource endowments, and shows the results disaggregated into 10 economic sectors. Section 6 concludes with an outlook on future research. 4 Carbon taxes, revenue recycling, and clean energy investment Price-based policy measures aimed at curbing greenhouse gas emissions encompass various strategies such as carbon taxes and cap-and-trade systems. Carbon pricing assigns a monetary cost to greenhouse gas emissions, incentivizing individuals and businesses to reduce fossil energy consumption. With carbon taxes, governments set explicit prices for carbon emissions, resulting in increased costs of fossil fuels. Revenue generated from carbon taxes can be used to fund various initiatives aimed at mitigating the im- pact of climate change. Under Emissions Trading Schemes (ETS), governments establish a maximum quantity of emissions and issue allowances accordingly. The companies covered by ETS trade these al- lowances, determining the price of carbon emissions through market mechanisms. However, ETS is high- ly administratively challenging and difficult to implement unilaterally due to carbon market liquidity problems, and their fluctuating prices hamper economic planning. Hence, this paper analyzes carbon pricing through taxation. Notably, carbon pricing may need to be part of a broader climate policy package to reduce GHG emis- sions effectively. Evidence, mostly from high-income countries, suggests that existing carbon pricing schemes have led to minimal GHG emission reductions: A meta-analysis of empirical studies shows that the carbon pricing schemes analyzed, on average, induced 0 percent to 2 percent reductions in the juris- dictions’ annual emissions (Green 2021). This remains below IPCC estimates of GHG emissions needed to fall by 45 percent below 2010 levels to limit warming to 1.5°C (IPCC 2018). Carbon pricing schemes seem not to have induced low-carbon investment or innovation (Lilliestam, Patt, and Bersalli 2021). Moreover, existing carbon prices are well below the price levels needed to achieve the Paris Agreement goals (cf. Stiglitz and Stern 2017). Hence, carbon pricing needs the support of other policy instruments to increase investment in renewable energy and efficiency. Carbon pricing has been shown to offer a range of economic benefits beyond reducing GHG emissions. By encouraging behaviors that reduce energy consumption and promote the adoption of clean energy, carbon pricing supports global efforts to combat climate change and efforts to improve local air quality (Sampedro et al. 2023). Furthermore, carbon pricing mechanisms provide political certainty about future increases in fossil fuel prices for investors. Carbon pricing can thus play a role in the innovation and adoption of low-carbon technologies, fostering investment in sustainable energy solutions (Calì et al. 2022)Finally, carbon taxes can generate significant revenue streams for governments. The receipts can be directed towards supporting further climate mitigation or development priorities (e.g., funding public expenditure initiatives on education or health) or supporting jobs-rich growth. Carbon taxes are hard to evade, so they increase fiscal space for overall budgetary uses or targeted measures. However, fiscal reforms entailing carbon taxes also come with certain challenges, including public ac- ceptance. One of the primary political concerns surrounding policies for low-carbon transitions is the level of public acceptance and perceived equity. Transitioning towards resilient and low-carbon struc- tures has the potential to yield favorable impacts on overall employment, extending beyond just the en- ergy sector, as highlighted by recent research (Marin and Vona 2019; Metcalf and Stock 2020; Schoder 2022). However, of particular concern is the heterogeneity of labor market responses and employment outcomes across different occupations, regions, income levels, and genders, which are still not fully comprehended. For consumers, higher fossil fuel prices resulting from carbon pricing may reduce household purchasing power and impact living standards. Similarly, energy-intensive industries may face 5 challenges in maintaining their competitiveness due to increased production costs, potentially leading to job losses and reduced industrial output. The revenue generated from carbon taxes can be used strategically to address existing economic chal- lenges and support sustainable development goals. This includes direct income support to low-income households, funding job retraining programs for affected workers, and investing in low-carbon technolo- gies. Revenues can also be spent on public infrastructure and welfare programs (Timilsina 2022). Fur- thermore, revenues can substitute existing distortionary taxes, mainly, income tax and indirect taxes. Reducing such distortionary taxes can lead to better economic outcomes in terms of GDP and welfare, compared to absorbing carbon revenues in government budgets or providing direct income support to households (Goulder 2013). Shifting energy investment towards renewable options leads to more jobs and growth at the country level. Evidence on such renewable energy benefits is broad, based on different modeling and empirical approaches, mostly for developed countries and under various scenarios. The aggregate results often show either close to no effects on jobs, meaning job losses net out job gains, slightly positive effects (for instance, for Europe, see (Barker et al. 2016)), or significantly positive effects. The literature agrees that net effects need to be studied, and job losses in high-emitting sectors are counterbalancing the positive effects of investing in clean energy. Such effects are even larger at subnational levels, but we did not go below the national level in the current analysis. Countries specializing in manufacturing renewable ener- gy technologies gain from trade and increased export opportunities, while countries depending on fossil fuel imports gain from trade effects by becoming less dependent when moving to alternative energy sources. At the EU level, transformation scenarios are also assessed in terms of their socioeconomic consequences by soft-linking energy system models with macroeconomic models (Fragkos, Fragkiadakis, and Paroussos 2021, Vrontisi et al. 2020). Nieto et al. (2020) analyze different pathways and find positive effects on employment. Other modeling results include Pai et al. (2021), who find an overall increase of energy jobs globally of 7.7 million by 2050, using an integrated assessment model augmented with country-specific labor productivity data. The above-listed revenue effects can enhance or alleviate the effects of the clean energy transition in general. More general effects include budget effects on fossil fuel importers and exporters. For the im- porting countries, GDP will be positively affected by a better trade balance, with renewable energy and energy efficiency options fueled by domestic resources and lowering the need for imported fuels. Oil exporters will see less export demand, reducing returns from royalties. Emission reduction reduces health risks, contributing to the overall target of mitigating climate change. These effects are independ- ent of the choice of mitigation policy instruments and tax design. The evidence for developing countries is still nascent but growing. Both the impacts of carbon taxes and additional investment in renewable energy have important implications for developing countries. With low tax incidence and tight public budgets, carbon taxes can increase the fiscal space and channel revenues toward achieving sustainable development goals. With roughly 675 million people globally without electricity access,1 2.3 billion people without access to clean cooking, and up to 3.5 billion households without access to reliable and affordable energy solutions (Ayaburi et al. 2020), investing in 1 https://www.irena.org/Publications/2023/Jun/Tracking-SDG7-2023 6 renewable energy provides a solution not only for the climate challenge, but also towards better eco- nomic opportunities, better jobs, and growth. Relatively few studies have considered the specificities of job implications in lower-income countries when analyzing the impacts of carbon prices, for example, accounting for informal labor or remittances. The World Bank’s Country Climate and Development Re- ports (World Bank Group 2022) give an overview of the economic impacts of price-based and invest- ment-focused climate policies in developing countries but do not focus on employment impacts. The modeling approach: MINDSET The Model of Innovation in Dynamic Low-Carbon Structural Economic and Employment Transfor- mations, MINDSET is designed to analyze short- and medium-term sectoral output, labor demand, and carbon-related cross-border competitiveness effects of climate and development policy mixes. The current version of MINDSET is a demand-driven global macroeconomic model with exogenous response functions. Based on the global GLORIA MRIO database2, the model supports multi- or single-country pol- icy scenario analyses, including impact analyses from third-party policies.3 Policy scenarios include the sectoral climate and development measures such as carbon pricing, border adjustments, income and payroll tax reforms, public spending and investment, electricity sector reforms (jointly with energy sector models), and climate damages (see Table 1 for detailed model inputs and outputs). Table 1: Granular policy & climate scenarios modeled simultaneously for 120 sectors & 164 countries: a) Carbon prices: Commodity or carbon tax, subsidy, and tariffs, differentiated by commodity, sec- tor, and country b) Carbon Border Adjustment Mechanism (CBAM): domestic carbon prices are complemented by border adjustments for user-defined energy-intensive and trade-exposed sectors, differentiated by sector and country for various global collaboration scenarios c) Labor tax reforms, differentiated by skill level d) Income tax reforms e) Sectoral public spending policies, cross-funded or additional f) Sectoral public investment, cross-funded or additional g) Climate damages h) Clean energy investment Model outputs inform various metrics of interest to support policy design: • Sectoral output multipliers of tax and spending policies, • Sectoral labor demand, or employment multipliers of tax and spending policies • Distributional employment and consumption effects, differentiated by income, skill level, oc- cupation type and, spatially, by subnational province • Sectoral tax revenue 2 We used release 057 of the GLORIA global environmentally-extended multi-region input-output (MRIO) database (Lenzen et al. 2021), constructed in the Global MRIO Lab (Lenzen et al. 2017). 3 The model has been used in several climate-informed country analytics, such as the Bosnia and Herzegovina Country Economic Memorandum, CCDRs in the Philippines, Peru, and Bangladesh, as well as country operations, such as a Morocco Climate Op- eration PforR. 7 • Sectoral consumer price differences & consumption distributional effects • Competitiveness effects and differences in sectoral import and export volumes The main transmission mechanism translates climate policies into price and demand changes and sim- ulates the economy's response to different climate policy scenarios. The results then are typically given as scenario differences, such as percentage difference of employment, output, or other economic indica- tors. The approach needs assumptions regarding the responses of industries and households to the sce- nario’s price level and labor tax level. These responses are determined by the respective elasticities, with own- and cross-price elasticities for energy carriers as intermediate inputs to production and price elas- ticities of final demand from households and Governments. The same holds true for labor taxes. MINDSET combines the strengths of input-output analysis – namely, short- to medium-term economic responses to exogenous demand differences, accounting for all multiplier effects from intermediate demand along the value chains in a consistent framework – with responses to exogenous price im- pacts. It connects all 164 countries in the dataset to each other through bilateral trade flows, thereby capturing trade effects from intermediate and final demand changes in one country on its trading part- ners. It estimates indirect and supply-chain impacts, as well as price-induced differences in international trade. The model's core elements are the interindustry relations given by intermediate use at basic prices by economic sector and final demand vectors by country connected by respective bilateral trade. The model does not capture labor flows into or out of the labor force or unemployment, nor between countries. Introducing endogenous adjustments of production, consumption, and trade into the input-output framework, the model predicts policy-induced shifts across all sectors and all countries or regions. The model estimates indirect and supply-chain impacts, as well as price-induced changes in international trade. Coverage is disaggregated into 120 sectors, including 10 energy sectors, across 164 countries and regions, of which 133 are included here. The model is parametrized with the various sector-country- (fuel)-specific elasticities, including energy efficiency and fuel switching, household income and demand, trade, and output-induced investment (and converters), and output-employment elasticities. Augmenta- tion with national labor force microdata allows for distributional labor demand analyses across income strata, skill levels, occupation types, and subnational provinces, providing relevant information for active labor market and reskilling interventions, as well as sectoral or regional government support funds. Caveats MINDSET does not assume that all available economic capacity is utilized. Instead, the level of effec- tive demand determines the use of factors of production in the economy, including labor. In the input- output tradition, prices in the model are mostly fixed, assuming that firms draw upon available capacity. Thus, the supply level of material and labor inputs matches overall demand. This assumption is well aligned with empirical data. For example, European Union countries utilize just under 80 percent of their production capacity. This share drops to as low as just above 60 percent for Europe’s upper-middle- income countries (European Commission 2024), and can be expected to be even lower in low- and low- er-middle-income countries. The assumption of additionality of investments (as opposed to crowding out) and unconstrained labor supply (involuntary unemployment) may overstate the expansionary ef- fects of climate policies. The model, therefore, complements the Computable General Equilibrium (CGE) approach. That said, MINDSET may be more appropriate than GE models for identifying transition im- 8 pacts and informing targeted complementary policy measures that could reduce short-term frictions. It also serves well for studying employment impacts, where labor market frictions or idle resources could have important socioeconomic consequences. Depending on the structure of a country’s economy, the assumption of additionality of unconstrained labor supply (involuntary unemployment) and investments (as opposed to crowding out) may slightly overstate the expansionary effects of climate mitigation policies. However, in countries with a large share of informal labor or unemployment, the assumption of strong constraints from the labor supply side (such as those captured in CGE models) may understate the expansionary macro effects of tax cuts or spending increases. Suppose a model assumes that development investments, such as in the health sector, have to crowd out investments from other areas of the economy (as assumed in CGE models). In that case, this will likely lead to negative GDP effects of climate or development policies. Because MIND- SET does not assume that the economy is at full capacity in the baseline, i.e., not at the efficiency fron- tier, the model further does not automatically allow prices to adjust. Thus, when demand increases (e.g., in the health sector), prices are not immediately pushed up. One area that is not yet considered well in model-based analysis is the mobility of labor between sec- tors, regions, and countries (Spencer et al. 2018). Models either assume perfect labor mobility (equilib- rium tools) or constraints based on historical trends (macro-econometric models). However, the scale of structural change in the low-carbon transition could run up against skill- and location-based constraints that have not previously been seen. Methods to assess these constraints are being developed (e.g., Mealy and Coyle 2022) and will be integrated into a future version of MINDSET. However, the policy sce- narios analyzed in this paper do not trigger fundamental structural shifts in labor demand, such that la- bor supply to meet the demand is unlikely to be constrained by labor market dynamics. 9 Scenario description Table 2: Scenario overview Carbon Revenue re- Additional Description Tax4 cycling spending Gov $80/tCO2 Public None 65% of the revenues are used for public in- Spending vestment in education and healthcare (e.g., construction and machinery procurement of hospitals and schools) and 35% for public spending on education and healthcare (hiring more doctors, nurses, and teachers5). PR $80/tCO2 Payroll Tax None Revenues are used to reduce payroll taxes, Cut with half of the revenues reducing labor costs for employers and the other half going to households as transfers. PR+RE $80/tCO2 Payroll Tax Clean ener- 100% of revenues is used to reduce payroll Cut gy invest- taxes. Investment in renewable energy and ment energy efficiency needed to close the gap to net-zero GHG emission electricity generation is not funded by the carbon tax, but rather by exogenous sources. Three scenarios have been implemented in the model and are compared to a business-as-usual sce- nario without additional climate policy in the simulation horizon, i.e., 2030. The overarching driver of people’s and firms’ behavior is the introduction of a global carbon tax. The policy scenarios are identical across all countries to facilitate comparison across countries and to illustrate the drivers of different eco- nomic outcomes. The same holds for the suggested revenue recycling schemes. Scenarios are identical across countries in terms of the percentages of the revenue allocated to the different purposes. In terms of levels, however, they differ according to the respective revenues collected in each country. The payroll tax cut scenarios try to capture the double dividend hypothesis, taxing fossil fuels and un- taxing jobs. Although this is somewhat simplistic, it illustrates how reducing labor costs in combination with carbon taxes leads to a shift from energy to labor as input factors of production. Half of the reve- nues from the carbon tax are used to lower labor costs, the other half is transferred to households, who spend it according to their existing expenditure patterns. Alternatively, revenue can be spent investing in health and education infrastructure and people. The health and education sectors are important for human capital development and, in most countries, un- derfinanced. Infrastructure is lacking, including in high-income countries, and is only gradually adapting 4 The suggested level of $80/tCO2 lies within the corridor concluded by the High-Level Commission on Carbon Prices, Carbon Price Leadership Coalition, 2017. 5 Note that given the timeframe of the analysis, secondary effects from having a healthier and better educated workforce are not included in the reported analysis. 10 to climate change. Schools need to be climate-resilient so that education is not hampered by extreme weather events. Schools can be turned into shelters in their neighborhoods, too6 (Feinstein and Mach 2020). Hospitals and clinics need infrastructure upgrades to be resilient and able to cope with extreme weather events or slow-onset changes. The public spending scenario picks up this idea and uses 65% of the revenue from the carbon tax for investment in resilient hospitals and schools. The remaining 35% is spent on staff, which could, for instance, include hiring climate health and education experts. Additional mitigation measures are needed because taxing carbon does not induce changes away from fossil fuel use to the extent needed to achieve climate targets. The gap between a carbon tax rate ac- ceptable to many countries and implementable and the necessary push for moving out of fossil fuels must be bridged by other measures. The scenario PR+RE assumes additional investment in clean energy, which has been determined by identifying the gap in renewable energy installations for electricity gener- ation and efficiency increases under the IEA’s net zero and current policies scenarios. This gives a coun- try-specific investment pathway that will unfold the respective economic impacts under the respective country’s economic structure. The results are driven by the interplay between price increases stemming from the carbon tax for all economic agents, income increases from income support and additional jobs, and demand increases from additional household expenditures and investment in clean energy. Higher prices for fossil fuels induce technology shifts for industries, trying to replace expensive inputs with less expensive ones or cutting back production. Households’ budgets are similarly impacted by direct price increases of fossil fuels and indirect price increases of goods and services using fossil fuels as inputs. Payroll tax cuts reduce the price of labor and are expected to lead to more jobs across the economy. Higher employment levels yield higher income and more demand for domestically produced and imported goods. Investment in renewable energy yields additional demand for goods and services, both during the installation phase and later for operation and maintenance. If the carbon tax is applied across all countries, the terms of trade are not largely affected, and trade effects occur from increased demand for renewable energy goods, reduced consumption of fossil fuels, and increased consumption and investment goods from ad- ditional spending. 6Examples are the COOLSCHOOLS project https://www.uoc.edu/en/news/2022/072-climate-shelters-schools , or the activities of the World Bank, who has between 2014 and 2023, supported 35 countries through projects that have benefited over 121 million students in around 564,000 schools and informed the design and implementation of more than $3.1 billion in World Bank-financed school infrastructure operations. https://www.worldbank.org/en/results/2024/03/08/building-safer-and-more- resilient-schools-in-a-changing-climate 11 Results The impacts on economic and employment outcomes differ across regions, sectors, country income levels, and countries with different resource endowments. The results are presented as relative differ- ences to the baseline. Indicators reported are relative differences in total employment, GDP, and sectoral employment (for the 10 economic sectors listed inError! Not a valid bookmark self-reference.), and sec- toral output. The results are grouped into seven regions, four income levels, and being a net crude or refined oil importer or exporter. The sample contains 46 countries in HIC, 11 in LIC 7, 37 observations in LMIC, and the remaining 38 in UMIC, while 15 countries are in the Eastern Pacific region, 47 In Europe and Central Asia, 20 in Latin America and the Caribbean, 16 in MENA and 2 in North America, 2 in South Asia and 28 in Sub-Saharan Africa8. Table 3: Result dimensions9 Economic sectors Regions Resources Income Groups •A. Agriculture; forestry • East Asia & Pacific •Oil importer •High Income (HIC) and fishing • Europe & Central Asia •Oil exporter •Upper Middle Income •B. Mining and quarrying • Latin America & •Petroleum importer (UMIC) •C. Manufacturing Caribbean •Petroleum exporter •Lower Middle Income •D. +E. Utilities • Middle East & North (LMIC) •F. Construction Africa •Low Income (LIC) •G.-N. Services • North America •O. Public administration • South Asia and defense; compulsory • Sub-Saharan Africa social security •P. Education •Q. Human health and social work activities •R.+S. Other Source: Own compilation Overview global results The overall employment effect of all scenarios is small but positive compared to the reference. Under the PR Scenario, the median value across all countries for the relative difference in employment is 0.2%, meaning employment is at least 0.2% higher than in the baseline for half of all countries. Albeit small, this is significant because it indicates that positive employment effects prevail. In 85% of all countries the impact of the carbon tax plus payroll tax reduction scenario (PR) is positive. Combining this with addi- tional investment in renewable energy (under the PR+RE scenario) drives the median impact to 0.6%, and spending revenues on health and education (as posited under the Gov scenario) leads to 0.3% addi- tional jobs. Lower middle-income countries experience slightly larger positive employment effects under the combined Pr+RE scenario, and the GOV scenario, partly because productivity is still low, yielding many jobs from additional output. However, revenue recycling through payroll tax reductions alone has a 7 Venezuela is currently not assigned an income group 8 Note that MINDSET contains 163 countries. The simulation results in this report cover 133 countries. The difference in country count is mainly due to country size and FCV status. 9 The classification of sector has been aggregated based on main ISIC sectors to make the result dimensions more manageable. 12 low effect on LICs. High-income countries benefit similarly from additional spending in renewable energy and see the highest effect from government spending compared across country income classes. Table 4: Impacts of different scenarios on employment, relative difference to reference. PR PR+RE Gov All 0.2% 0.6% 0.3% LIC 0.3% 0.7% 0.1% LMIC 0.2% 0.7% 0.3% UMIC 0.2% 0.6% 0.3% HIC 0.2% 0.6% 0.4% Source: MINDSET model results. GDP globally and by income group is relatively higher under all scenarios. GDP effects are highest un- der the payroll tax reduction scenario combined with investment in renewable energy technologies. LICs benefit strongly under this scenario especially if they are dependent on fossil fuel imports, because in- creased RE deployment reduces these imports. Relative to jobs impacts, GDP increases more strongly in HICs. The payroll tax reduction alone yields a lower GDP difference than the scenario with government spending on health and education plus household income support. Table 5: Impacts of different scenarios on GDP, relative difference to reference. PR PR+RE Gov All 0.4% 1.3% 0.6% LIC 0.2% 1.0% 0.1% LMIC 0.3% 1.2% 0.3% UMIC 0.4% 1.3% 0.8% HIC 0.5% 1.3% 0.8% Source: MINDSET model results. Output effects fall somewhere in the middle between employment effects and GDP effects, except for high-income countries. In high-income countries, the output effect is highest and most detached from the impact on jobs. The next section takes a deep dive into disaggregated sector effects and provides a structural explanation for these observations. Under the government spending plus household income support scenario, LIC cannot benefit as much in output because a large share of the additional demand will be met by imports. Sectoral disaggregation will illustrate this point further. Table 6: Impacts of different scenarios on output, relative difference to baseline. PR PR+RE Gov All 0.2% 1.3% 0.6% LIC 0.3% 1.1% 0.1% LMIC 0.1% 1.1% 0.3% UMIC 0.2% 1.2% 0.8% HIC 0.4% 1.4% 0.8% Source: MINDSET model results. 13 Economic and employment effects by region and sector The differences in economic outcomes are equally dissimilar across regions as across income groups. Median employment impacts under the payroll tax reduction plus renewable energy investment scenar- io by region are small for East Asia Pacific and North America. The largest variation in outcomes is ob- served across MENA countries, while Sub-Saharan African countries cluster much narrower around their median value of a little more than .3%. Europe and Central Asia, too, show the results more concentrat- ed around over 1% relative difference in employment under this scenario (Error! Not a valid bookmark self-reference.). High-income countries show a narrower range of employment impacts, with a standard error of 0.11% around an employment impact of .2%, while low and lower-middle-income countries have a larger variation in their respective results. For countries in regions and income groups with fewer variations, turning to their respective peers gives more information regarding their expected perfor- mances. In MENA and in LICs and LMICs, peer countries from the same group can exhibit very different results and may lead to invalid expectations. Table 7: Median values and confidence intervals for the PR scenario. Median [95% conf. interval] Region East Asia & Pacific 0.31% 0.16% 0.42% Europe & Central Asia 0.24% 0.19% 0.29% Latin America & Caribbean 0.25% 0.18% 0.35% Middle East & North Africa 0.06% -0.02% 0.29% North America 0.21% 0.10% 0.32% South Asia 0.31% 0.12% 0.50% Sub-Saharan Africa 0.31% 0.13% 0.41% Income Group Low Income 0.34% 0.01% 0.50% Lower Middle Income 0.23% 0.15% 0.33% Upper Middle Income 0.22% 0.11% 0.32% High Income 0.22% 0.18% 0.29% Source: MINDSET model results. Recycling the carbon tax revenues into spending and investing in education and health as well as in income support households (Gov scenario) creates positive employment impacts in the targeted sec- tors and the whole economy. Error! Not a valid bookmark self-reference. shows the relative differences under this scenario compared to employment in the respective sector under the baseline, grouped by income group. In all income groups, targeted investment and spending for the education and health sec- tors are reflected in additional employment. For LICs, additional employment in the health and educa- tion sector amounts to 1% of baseline sectoral employment. Additional investment in these sectors trig- gers construction activities and is reflected in almost 1.2% more jobs in construction. The construction sector is one of the largest non-agricultural sectors in many LIC economies. Manufacturing benefits from additional demand because many households have more income due to transfers and higher employ- ment. Services benefit along the value chain but do not see strong effects under this scenario. The up- take of the mining sector is related to quarrying along the value chain of the construction sector. Moving to the other income groups, this effect is more than offset by reductions in fossil fuel mining because of 14 the carbon tax and the resulting impact on mining is negative; most prominently in UMICs with more than 1% job losses. Albeit these losses are compensated on average by all other positive impacts, the effect on mining regions can be quite severe. UMICs also see the largest negative impacts on electricity generation based on fossil fuels, whereas HIC already have parts of this industry devoted to cleaner en- ergy options and not being as much negatively affected by the carbon tax. Figure 1: Employment effects by sector and income group, in relative differences of sectoral employment to baseline under ‘Gov’ scenario Source: MINDSET model results. Under the payroll tax cut scenario, effects are smaller and more evenly distributed across sectors. In Figure 2, the manufacturing and service sectors benefit more from this scenario than from the previously discussed redistribution scheme. Health and education also see additional employment under this sce- nario, partly from improved labor costs and partly from increased incomes and people’s willingness to spend more on health and education. The construction sector sees fewer impacts, because the targeted investment in health and education facilities is missing. The pattern in energy and mining is comparable to the scenario previously discussed; negative impacts in electricity generation are more relevant in more developed countries due to price increases, loss of demand, and less additional demand from in- creased economic activities compared to the scenario Gov. The largest positive effects are observed in some services and hiring by households, as shown in Figure 2. 15 Figure 2: Employment effects by sector and income group, in relative differences of sectoral employment to baseline under sce- nario PR Source: MINDSET model results. Combining clean energy investment with the payroll tax reduction (scenario PR+RE) leads to the high- est effects. Additional employment is seen across most sectors, with the construction sector further benefiting from clean energy investment. Investment in renewable energy, as well as increased energy efficiency, creates additional demand for output from the construction sector. HICs and UMICs see addi- tional jobs in the manufacturing sector, which are above the average across all sectors. This may be due to the demand for clean energy equipment and technologies, which are meeting additional global de- mand. Services respond to additional intermediate demand and lower labor costs, showing above- average increases in labor demand. 16 Figure 3: Employment effects by sector and income group, in relative differences of sectoral employment to baseline under sce- nario PR+RE 17 Source: MINDSET model results. Summary of employment effects across countries The effect of the outlined revenue-neutral carbon tax schemes can differ across countries, depending on their economic structure and income level. Figure 4 visualizes the employment and GDP changes for payroll tax cut and government expenditure scenarios in selected countries. Middle-income countries such as in East Asia and Pacific (Indonesia, Vietnam) can benefit more from revenue-neutral carbon tax compared to high-income countries such as in European countries (Germany, Switzerland) for either sce- nario (also see Table 4 and 5). In more detailed investigation, certain middle-income countries with lower labor productivity, such as Indonesia and Ecuador, are benefitting more from payroll tax cut, where more services jobs can be created from the productivity increase. The payroll tax reduction alone can absorb more labor in these countries, although targeted government spending can increase the GDP since the healthcare and education sectors are high-value-added sectors. The global results from this rather standardized exercise would need to be complemented by a more detailed analysis of job quality, informality, skill needs and gender distribution to paint a more com- plete picture. Additional job impacts from consumption demand, for instance, often occur in sectors that are characterized by high shares of job informality, unwaged jobs, or family labor in the agriculture 18 sector. Construction, where large impacts are observed, likewise exhibits high shares of informality. Cli- mate policies need to be complemented by the respective labor oriented policies to have the green tran- sition lead to more and better jobs. In general, oil exporting countries (in form of crude oil or petroleum products) with higher income lev- el can benefit more from revenue-neutral carbon tax than net oil importers. In major oil producing countries in Middle East such as Saudi Arabia, Oman, and Iran, revenue-neutral carbon tax creates a pos- itive impact on employment and GDP. Meanwhile, oil producing countries in other regions such as Bru- nei, Turkmenistan, and Nigeria, revenue-neutral carbon tax does not generate as positive an impact as observed in the Middle East region. Since Middle Eastern countries have both more efficient technology and more diversified economy compared to other oil exporting countries, the impact of carbon tax on the economy is relatively lower and the high tax revenue obtained could then be used to develop the existing non-petroleum industries in these countries, particularly if they have more diverse industry structure. Figure 4: Impacts of revenue-neutral carbon tax in payroll tax cut (PR) and government expenditure (GOV) scenarios on GDP and employment, relative difference to reference. The color represents the country region based on World Bank classification. Payroll taxes plus renewable energy investment result in greater economic benefits than government spending with income support in most countries. This confirms the finding in the literature, as Tilimisa 2022 also points out in the literature review on the benefits of carbon taxation. Most countries fare bet- ter under the increased health and education spending and investment in making the health sector more resilient, by investing in buildings, early warning systems etc. The mere effects from recycling the reve- nue to reduce distortionary taxes, in this case payroll taxes, do not exceed the benefits from the targeted government spending. Figure 4 displays a map indicating where employment impacts under the PR sce- nario are larger than under the Gov scenario (“yes”, denoted in green) or not. Error! Reference source n ot found. poses – and answers – the same question for the PR+RE scenario. 19 Figure 5: Comparison of effects across countries. Comparing the PR scenario with the Gov scenario Source: MINDSET model results. Adding renewable energy investment changes the picture. Notably, relative differences between sce- nario effects are not represented in (Figure 5), and for several countries, the economic effects are of very similar size. However, the combined payroll tax reduction and investment scenario does better than the revenue recycling into targeted government spending for all countries. Conclusions and outlook This analysis provides very detailed results for countries to inform their understanding and expecta- tions for different carbon tax schemes. The results illustrate how different countries respond to a styl- ized and generalized set of climate policies. The combination of carbon taxes and revenue recycling through payroll tax reduction seems to be the most efficient and the most beneficial in terms of addi- tional jobs. It removes negative externalities of both emissions and distortionary taxes, and hence is the- oretically appealing and yielding a double dividend. To more fully understand the implications and fac- tors which are driving the results, a wider net on explaining factors could be cast. Potential indicators could be population or area size. This analysis is part of a broad research agenda on the employment implications of decarbonization and green transition (that is, to more climate-resilient and sustainable economies). Future research will explore emerging policy questions of relevance to policymakers, such as revenue recycling schemes with transfers to only lower-income quantiles, to better target the poor and vulnerable groups. The literature on fossil fuel subsidy reform might provide some compelling examples (IISD 2019; World Bank 2022). Deep dives into subsets of countries would allow more analysis of skills needed and existing, gender balance of results and other labor market aspects. Global results allow cross country comparison with peers or aspirational peers. Country deep dives then allow the analysis of drivers under different scenar- ios and a better understanding of the transmission pathways of employment effects in the economy. The 20 MINDSET model has been used in country specific analyses, such as (Lehr and Pollitt 2024). Further sce- narios, as outlined above would be of interest in the decision process for the urgently need more strin- gent climate policies for both mitigation and adaptation. References Ayaburi, John, Morgan Bazilian, Jacob Kincer, and Todd Moss. 2020. “Measuring ‘Reasonably Reliable’ Access to Electricity Services.” The Electricity Journal 33 (7): 106828. https://doi.org/10.1016/j.tej.2020.106828. Barker, Terry, Eva Alexandri, Jean-Francois Mercure, Yuki Ogawa, and Hector Pollitt. 2016. “GDP and Em- ployment Effects of Policies to Close the 2020 Emissions Gap.” Climate Policy 16 (4): 393–414. https://doi.org/10.1080/14693062.2014.1003774. Calì, Massimiliano, Nicola Cantore, Leonardo Iacovone, Mariana Pereira-López, and Giorgio Presidente. 2022. “Too Much Energy The Perverse Effect of Low Fuel Prices on Firms.” Journal of Environ- mental Economics and Management 111 (January): 102587. https://doi.org/10.1016/j.jeem.2021.102587. European Commission. 2024. “Current Level of Capacity Utilization in Manufacturing Industry.” https://doi.org/10.2908/TEIBS070. Feinstein, N. W., and K. J. Mach. 2020. “Three Roles for Education in Climate Change Adaptation.” Cli- mate Policy 20 (3): 317–22. https://doi.org/10.1080/14693062.2019.1701975. Fragkos, Panagiotis, Kostas Fragkiadakis, and Leonidas Paroussos. 2021. “Reducing the Decarbonisation Cost Burden for EU Energy-Intensive Industries.” Energies 14 (1): 236. https://doi.org/10.3390/en14010236. Goulder, Lawrence H. 2013. “Climate Change Policy’s Interactions with the Tax System.” Energy Econom- ics 40 (December): S3–11. https://doi.org/10.1016/j.eneco.2013.09.017. Green, Jessica F. 2021. “Does Carbon Pricing Reduce Emissions? A Review of Ex-Post Analyses.” Environ- mental Research Letters 16 (4): 043004. https://doi.org/10.1088/1748-9326/abdae9. IISD. 2019. “Story 1: Removing Subsidies for Gasoline and Diesel Consumption in Indonesia.” IPCC. 2018. Global Warming of 1.5°C. An IPCC Special Report on the Impacts of Global Warming of 1.5°C above Pre-Industrial Levels and Related Global Greenhouse Gas Emission Pathways, in the Con- text of Strengthening the Global Response to the Threat of Climate Change, Sustainable Devel- opment, and Efforts to Eradicate Poverty. Edited by V. Masson-Delmotte, P. Zhai, H. Pörtner, D. Roberts, J. Skea, P. Shukla, A. Pirani, et al. In Press. Lehr, Ulrike, and Hector Pollitt. 2024. “HEADING TOWARDS 1.50C – IMPACTS ON LABOR DEMAND IN SE- LECTED COUNTRIES.” Lenzen, Manfred, Arne Geschke, Muhammad Daaniyall Abd Rahman, Yanyan Xiao, Jacob Fry, Rachel Reyes, Erik Dietzenbacher, et al. 2017. “The Global MRIO Lab – Charting the World Economy.” Economic Systems Research 29 (2): 158–86. https://doi.org/10.1080/09535314.2017.1301887. Lenzen, Manfred, Arne Geschke, James West, Jacob Fry, Arunima Malik, Stefan Giljum, Llorenç Milà i Ca- nals, et al. 2021. “Implementing the Material Footprint to Measure Progress towards Sustainable Development Goals 8 and 12.” Nature Sustainability 5 (2): 157–66. https://doi.org/10.1038/s41893-021-00811-6. Lilliestam, Johan, Anthony Patt, and Germán Bersalli. 2021. “The Effect of Carbon Pricing on Technologi- cal Change for Full Energy Decarbonization: A Review of Empirical Ex-post Evidence.” WIREs Cli- mate Change 12 (1): e681. https://doi.org/10.1002/wcc.681. Marin, Giovanni, and Francesco Vona. 2019. “Climate Policies and Skill-Biased Employment Dynamics: Evidence from EU Countries.” Journal of Environmental Economics and Management 98 (No- vember): 102253. https://doi.org/10.1016/j.jeem.2019.102253. 21 Metcalf, Gilbert E., and James H. Stock. 2020. “Measuring the Macroeconomic Impact of Carbon Taxes.” AEA Papers and Proceedings 110 (May): 101–6. https://doi.org/10.1257/pandp.20201081. Nieto, Jaime, Óscar Carpintero, Luis Fernando Lobejón, and Luis Javier Miguel. 2020. “An Ecological Mac- roeconomics Model: The Energy Transition in the EU.” Energy Policy 145 (October): 111726. https://doi.org/10.1016/j.enpol.2020.111726. Pai, Sandeep, Johannes Emmerling, Laurent Drouet, Hisham Zeriffi, and Jessica Jewell. 2021. “Meeting Well-below 2°C Target Would Increase Energy Sector Jobs Globally | Elsevier Enhanced Reader.” 2021. https://doi.org/10.1016/j.oneear.2021.06.005. Sampedro, Jon, Anil Markandya, Clàudia Rodés-Bachs, and Dirk-Jan Van De Ven. 2023. “Short-Term Health Co-Benefits of Existing Climate Policies: The Need for More Ambitious and Integrated Pol- icy Action.” The Lancet Planetary Health 7 (7): e540–41. https://doi.org/10.1016/S2542- 5196(23)00126-2. Schoder, Christian. 2022. “Regime-Dependent Environmental Tax Multipliers: Evidence from 75 Coun- tries.” Journal of Environmental Economics and Policy, June, 1–44. https://doi.org/10.1080/21606544.2022.2089238. Stiglitz, Joseph, and Nicholas Stern. 2017. “Report of the High-Level Commission on Carbon Prices.” Washington, DC: World Bank Group: Carbon Pricing Leadership Coalition. Timilsina, Govinda R. 2022. “Carbon Taxes.” Journal of Economic Literature 60 (4): 1456–1502. https://doi.org/10.1257/jel.20211560. Vrontisi, Zoi, Kostas Fragkiadakis, Maria Kannavou, and Pantelis Capros. 2020. “Energy System Transition and Macroeconomic Impacts of a European Decarbonization Action towards a below 2 °C Cli- mate Stabilization.” Climatic Change 162 (4): 1857–75. https://doi.org/10.1007/s10584-019- 02440-7. World Bank. 2022. “Securing Energy Subsidy Reform: When Textbook Solutions Clash with Reality.” World Bank Blogs. June 2022. https://blogs.worldbank.org/en/eastasiapacific/securing-energy- subsidy-reform-when-textbook-solutions-clash-reality. UNFCC 2023, Submission by Saudi Arabia on behalf of the Arab Group On efforts related to addressing the social and economic consequences and impacts of Response Measures Pursuant to para.11 FCCC/PA/CMA/2022/L.7 22