With country-specific development objectives and constraints, multiple market failures and limited international transfers, carbon prices do not need to be uniform across countries, but must be part of broader policy packages.
There is broad agreement that achieving the objectives of the Paris Agreement and limiting warming to below 2 °C in an efficient manner will require the implementation of national carbon prices that increase throughout the twenty-first century1. According to the High-Level Commission on Carbon Prices, the explicit carbon-price level that is consistent with achieving this target, assuming a supportive policy environment, is at least US$40–80 tCO2e−1 by 2020 and US$50–100 tCO2e−1 by 20301.
The Commission’s report recommends carbon price levels that are tailored to a country’s characteristics, including its income level, the quality of its institutions, its endowment in renewable energy and other key resources, its economic structure, its social protection systems, its political situation and many other factors. This is at odds with basic economic theory, which argues that an equal price for all regions and sectors, whether through a tax or cap-and-trade, is the most effective and efficient tool to reduce emissions.
Here we discuss three reasons why carbon prices — consistent with the Commission’s recommendation — should (and will) differ across countries. First, political limits to financial transfers between countries, as well as differing national contexts and development levels, justify lower carbon prices in developing countries. Second, multiple market failures (such as those regarding innovation) mean that carbon pricing needs to be complemented with other policies, and their nature and ambition impact the final carbon price to achieve a given objective. Third, non-climate development objectives (and the policies to achieve them) interact with climate goals, and influence the adequate carbon price level.
We conclude that nations, when designing climate policy packages, should not start with a carbon price level. They should instead begin by defining their emissions objective in context of their other goals, and design comprehensive and integrated strategies to achieve these goals. Carbon pricing would be a key component of the policy package to implement these strategies, but its incidence and stringency would be tailored to the national context and objectives.
Carbon prices and country context
The standard conclusion that optimal climate policy implies a unique carbon price around the globe is valid only if unlimited transfers among countries are possible to compensate for differences in abatement costs and welfare effects2,3. In the presence of limits to international transfers — limits that are well illustrated by tensions on official development aid — the optimal distribution of efforts implies different price levels across countries.
These levels depend on preferences regarding global inequality and how well-being depends on consumption levels (the social welfare function in economic jargon) and multiple country characteristics. Countries differ by size, weather, intra- and interurban density, renewable energy endowment (for example, hydropower, geothermal energy, solar and wind), potential for geological storage of CO2, capacity for employing nuclear power, economic structure (including the importance and mix of energy-intensive industries) and social preferences — all of which can affect GHG intensity and the marginal cost of abatement4.
At the national level, the price of carbon needed to achieve a given national objective (as defined by Nationally Determined Contributions to the Paris Agreement, or a longer-term objective for 2050 and beyond, for instance) will be lower when the marginal cost of abatement is lower. In particular, many developing countries with large endowments of solar, wind, hydropower or geothermal energy may find it possible to achieve large emission reductions with lower carbon prices than other countries.
Developing countries also have different cost structures: poorer countries have larger shares of energy costs and lower shares of labour costs, making their productivity more vulnerable to increases in energy prices and correspondingly higher macroeconomic effects5. In addition, poorer people, especially when they are close to subsistence level, lose a lot of welfare when their consumption is reduced or grows more slowly. The high economic vulnerability of poor people to increases in energy prices could justify implementing lower carbon prices, unless one can ensure that these populations are compensated and protected through social protection systems, cash transfers or reductions in other taxes6.
Coverage failures exist where it is difficult to measure or price GHG emissions (for example, fugitive emissions due to pipeline leaks, emissions fluxes related to land-use management). Network failures can arise where a physical or institutional infrastructure must already exist to allow discrete mitigation systems to operate (such as land-use and transport-oriented urban planning and investment to facilitate mode shifting11; electricity grid planning, and market design and transmission investment to facilitate electrification and intermittent renewables. Knowledge spill-overs occur when private investors cannot capture all of the benefits of early research and development and commercialization investment in mitigation technologies, and therefore underinvest from a public point of view12. Failures also take place where behaviour makes carbon pricing ineffective, or when people fail to anticipate the long-term carbon-price pathways, decades in advance. For example, in the personal transport and housing sectors a myriad of factors can overwhelm available information on the GHG intensity of mode choices or personal investment in appliance or building efficiency13.
Addressing these issues requires the construction of packages of complementary policies. These policies can contribute to reducing emissions, but also improve the political acceptability of carbon pricing by reducing the carbon price needed to achieve climate objectives.
Market imperfections are particularly large in developing countries, which will affect the balance between pricing and non-pricing instruments14. Carbon price signals may be swamped in a myriad of contradictory information and incentives in the presence of incomplete markets, informal exchanges, lack of regulatory enforcement, instability of institutions and fast-evolving infrastructures affecting access to information and foresight stability. For instance, in some countries energy markets do not dispatch power generation options based on cost. In this case, a carbon price, while it reduces the relative cost of renewables, would not necessarily lead to a reduced use of fossil fuels. All of the above may lead policymakers, especially those in low-income and weak-institution contexts, to use instruments that are easier to implement and enforce non-price options, at least in the short term.
Countries differ in the social and political circumstances that affect the political economy of carbon pricing, such as income levels, poverty incidence, agreement on the distribution of efforts over time (that is, the discount rate) and the ability of government to protect and support the transition for vulnerable populations and industries. Because there are many market and government failures, externalities and biases in behaviours, climate mitigation will involve multiple instruments in all countries, including carbon pricing, innovation policies, regulations and performance standards, targeted subsidies, and education and training. The balance between these instruments will depend on the local political economy and the social and political acceptability of these instruments. Some countries — especially those with low income levels — will have a lower explicit carbon price, because that is what is realistically possible, and may have to do more using other tools.
These dynamics can be illustrated by 2 °C scenarios developed for the Canadian Deep Decarbonization Pathways Project15 to limit GHG emissions to 1.7 tCO2e per capita by 2050 (Fig. 1). These scenarios use either a carbon price alone or a package of carbon pricing and complementary policies to achieve the same ends. The latter uses roughly half the carbon price, but increases short- and long-term political and social acceptability while addressing coverage and behaviour failures in land use, fugitive emissions, transport and buildings.
The policy package included: (1) economy-wide carbon pricing starting at CAN$10 per ton of CO2-equivalent emissions (tCO2e; in 2015 Canadian dollars) and rising by CAN$10 per year steadily through time, recycled to reduce corporate and income taxes equally; (2) sector-specific performance standards that fall to net-zero emissions by 2025–2040 for new investments in personal transport, freight transport, residential buildings and commercial buildings; (3) an intensity-based, tradable performance standard for large emitters using output-based allocations that falls to 90% below 2015 GHG intensity per unit gross output by 2050 and could potentially be linked to other regional cap and trade systems16; and (4) methane and land-use regulations.
This policy package was included in a submission15 to the 2015 Alberta climate policy process and cited in its deliberations17. Supporting the development of recognizably similar policy in Alberta, the overall structure of the policy package also helped to inform carbon pricing policy at the Canadian federal level, as part of the Pan-Canadian Framework on Clean Growth and Climate18.
Non-climate development objectives
Every country, developed or developing, has a host of societal development goals (such as economic growth, equity, access to services, health and air quality, education). Climate is one policy goal among many that have strong interactions with energy supply and demand and land use. Policies to achieve any development goal can have consequences on GHG emissions — and vice versa for climate mitigation policies.
For example, the Indian Deep Decarbonization Pathway Project19 looked at two scenarios with the same cumulative emissions targets. One is ‘climate centric’ and considers only a carbon price that rises roughly linearly to US$130 tCO2e−1 by 2050. The other scenario covers broader development goals, such as local air pollution, energy security, urban planning, decentralized energy for rural areas and water management. It achieves those objectives with a stronger emphasis on the demand side, through behavioural change and infrastructure measures. Although a carbon price is included in the portfolio of instruments, it is only part of it. The final carbon price in 2050 in this scenario needs to reach only US$35 tCO2e−1, instead of US$130, thanks to the benefits of other non-price policies (Fig. 2).
Similar insights can be drawn from global integrated assessment models20, particularly from a comparison of scenarios with similar climate outcomes but varying underlying socio-economic assumptions (that is, similar climate objectives with different shared socio-economic pathways, or SSPs21). The carbon prices needed to achieve a 2 °C-consistent pathway are about 15–80% lower under sustainable development assumptions compared with middle-of-the-road assumptions (SSP1 versus SSP2), with most scenarios showing at least a 50% decrease in price. Correspondingly, the carbon prices needed to achieve a 2 °C target are projected to be 5–70% higher when assuming a world with high technological development but a strong focus on fossil fuel exploitation (SSP5 versus SSP2). This example illustrates the important effect that other societal objectives and policies can have on the carbon prices consistent with a climate target such as 2 °C. For instance, climate policies do not need to be as stringent (and the carbon price as high) in a world (or a country) where environmental conservation is high on the agenda as those in a world that prioritizes material consumption.
Carbon prices and policy packages
Carbon pricing will not be enough to trigger an efficient, just and acceptable decarbonization transition. Successful climate policies will take the form of nationally designed climate policy packages that include both carbon prices and complementary climate and non-climate policies tailored to national circumstances.
These packages can be designed to increase predictability and stability (which is important to trigger private-sector investment) and to reduce the needed explicit price signal, making them easier to implement and sustain. These packages can be implemented by reforming existing domestic fuel tax systems such that they are based on carbon intensity, as done by the Scandinavian countries starting in the 1990s. Complementary policies would address energy efficiency, coverage gaps, network effects, generate innovation and help in meeting other development goals.
The design of these packages cannot start with a carbon price level as the only entry point. Instead, it should start from the definition of a national climate goal (for example, net-zero GHG emissions by 2070 or some other year22,23, with sector-specific interim and long-term targets24) alongside a nation’s other development objectives, for which comprehensive and integrated strategies can be developed24. These strategies can be physically grounded on four pillars of action: (1) efficiency, (2) fuel switching, (3) decarbonized energy carriers; and (4) direct emissions reductions and offsets through land-use management and CO2 removal. National stakeholders and policymakers can then consider the multiple relevant dimensions in their entirety, to allow the design of a policy package that is internally consistent, addresses the many obstacles to emission reductions and considers the carbon price schedule that is needed as one component of this package.
Report of the High-Level Commission on Carbon Prices (Carbon Pricing Leadership Coalition, 2017).
Chichilnisky, G. & Heal, G. Environmental Markets: Equity and Efficiency (Columbia Univ. Press, 2000).
Sheeran, K. A. Environ. Resour. Econ. 35, 89–98 (2006).
Bataille, C., Rivers, N., Mau, P., Joseph, C. & Tu, J.-J. Energy J. 28, 145–169 (2007).
Waisman, H., Guivarch, C., Grazi, F. & Hourcade, J. C. Climatic Change 114, 101–120 (2012).
Hallegatte, S. et al. Shock Waves: Managing the Impacts of Climate Change on Poverty (World Bank, Washington DC, 2015).
Acemoglu, D., Aghion, P., Bursztyn, L. & Hemous, D. Am. Econ. Rev. 102, 131–166 (2011).
Fay, M. et al. Decarbonizing Development: Three Steps to a Zero-Carbon Future (World Bank, 2015).
Stern, N. Why are we Waiting? The Logic, Urgency, and Promise of Tackling Climate Change (MIT Press, Cambridge, MA, 2015).
Hallegatte, S., Fay, M. & Vogt-Schilb, A. Green Industrial Policies—When and How (World Bank, 2013).
Avner, P., Rentschler, J. & Hallegatte, S. Carbon Price Efficiency: Lock-in and Path Dependence in Urban Forms and Transport Infrastructure Policy Research Working Paper No. 6941 (World Bank, 2014).
Newell, R. G. Oxf. Rev. Econ. Policy 26, 253–269 (2010).
Gerarden, T. D. J. Econ. Lit. 55, 1486–1525 (2015).
World Development Report 2002 (World Bank, 2001); https://doi.org/10.1596/0-1952-1606–7
Bataille, C., Sawyer, D. & Melton, N. Pathways to Deep Decarbonization in Canada (Deep Decarbonization Pathways Project, 2015).
Burtraw, D., Munnings, C., Palmer, K. & Woerman, M. Linking Carbon Markets with Different Initial Conditions (Resources for the Future, 2017).
Leach, A., Adams, A., Cairns, S., Coady, L. & Lambert, G. Alberta Climate Leadership Plan: Report to Minister (Alberta Provincial Government, 2015).
Federal Carbon Pricing Backstop (Environment and Climate Change Canada, 2017).
Shukla, P. R., Dhar, S., Pathak, M., Mahadevia, D. & Garg, A. Pathways to Deep Decarbonization in India (SDSN & IDDRI, 2015).
McCollum, D. L., Krey, V. & Riahi, K. Nat. Clim. Change 1, 428–429 (2011).
Riahi, K. et al. Glob. Environ. Change 42, 153–168 (2017).
Rogelj, J. et al. Environ. Res. Lett. 10, 105007 (2015).
Pye, S., Li, F. G. N., Price, J. & Fais, B. Nat. Energy 2, 17024 (2017).
Bataille, C. et al. Clim. Policy 16, S7–S26 (2016).