Getting to a Carbon-Free Economy

The urgent is attainable, and at entirely affordable cost.


The Green New Deal refers to two distinct ideas. The narrower idea, which I discuss here, is a strategy to decarbonize the U.S. energy system in line with the Paris Agreement. The broader, urgent idea is an integrated program including renewable energy, infrastructure, health care, education, and jobs.

Contrary to some commentaries, decarbonization will not require a grand mobilization of the U.S. economy on par with World War II. The incremental costs of decarbonization above our normal energy costs will amount to 1 to 2 percent of U.S. GDP per year during the period to 2050. By contrast, during World War II, federal outlays soared to 43 percent of GDP from the prewar level of 10 percent of GDP in 1940.

The key today is to redirect outlays now spent on fossil fuel–based technologies toward zero-carbon technologies instead. That redirection will require a serious increase in federal and state public infrastructure spending, but most importantly will depend on new federal and state regulations to redirect the energy-related spending. Carbon pricing (such as a carbon tax) will be one useful tool for redirecting the spending, but will be of less importance than regulations.

Limiting Global Temperature Rise

The core goal of the Paris Agreement is to strengthen the global response to climate change by:

Holding the increase in the global average temperature to well below 2°C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5°C above pre-industrial levels, recognizing that this would significantly reduce the risks and impacts of climate change. (Article 2, Section 1a)

The Intergovernmental Panel on Climate Change (IPCC) followed the Paris Agreement with a crucial study (2018) highlighting the very grave risks of exceeding the 1.5 degree Celsius limit. There are three kinds of dire risks. First, the global damages from 2 degrees of warming would be significantly higher than from 1.5 degrees. Second, even 1.5 degrees may well be above the threshold for major irreversibilities such as a multi-meter rise in the sea level caused by the partial disintegration of the Antarctic and Greenland ice sheets. Third, increases in temperature above 1.5 degrees threaten to unleash powerful positive feedbacks that could lead to a spiraling of further warming, such as the massive release of CO2 and methane from the melting permafrost. A superb comprehensive scientific overview of several of these issues was written by James Hansen and co-authors. [1] (See here for an up-to-date account of possible large-scale positive feedback effects that could dramatically amplify climate change.)

Indeed, the long-term risks of even a 1.5 degree warming sustained over decades or centuries are so high that we must be more ambitious than merely stabilizing at 1.5 degrees. Hansen cogently argues that we should return CO2 concentrations to levels consistent with long-term warming of just 1 degree Celsius. That means that after reaching a peak CO2 concentration of perhaps 450 ppm (parts per million) or less at mid-century, we should enter a sustained phase of global net negative CO2 emissions so that CO2 concentrations gradually return to 350 ppm over the long term, or ideally even sooner. Besides phasing out carbon emissions, this goal can be accomplished through two kinds of carbon storage: geological storage (capturing CO2 in the air and pumping it into geologic reservoirs) and biological storage (afforestation, reforestation, and restoration of degraded lands, all capturing CO2 in vegetative cover and soils).

To stay below 1.5 degrees of warming, all regions of the world should reach net-zero emissions by 2050 at the very latest. The U.S. currently represents around 15 percent of global CO2 emissions, and that share will decline in the coming years. Therefore, decarbonization of the U.S. energy system must be complemented by decisive actions abroad, but the U.S. needs to lead. The Green New Deal should include a pillar for U.S. diplomacy and international economic policy designed to speed actions abroad, not only in the United States. At a minimum, the U.S. must remain a leader of the Paris Agreement.

Five Technological Pillars of Decarbonization

Jason Schneider

Hundreds of scholarly and policy studies have reached a broad consensus on the technology pathway to decarbonization. The consensus points to five pillars of decarbonization. [2] 

The first and most important single pillar is zero-carbon electricity. This is the most important measure since zero-carbon electricity can be deployed directly (in battery electric vehicles, for example) and indirectly in green chemistry to manufacture zero-carbon fuels (hydrogen, for example). Zero-carbon electricity involves a shift toward zero-carbon primary energy sources and a very significant overall expansion of electricity production for the electrification of transport, buildings, and industry.

Zero-carbon electricity worldwide will tap multiple primary energy sources, including renewables (broadly defined to include wind, solar, hydro, geothermal, ocean, and tidal); nuclear; biofuels; and carbon capture and storage of fossil fuel–generated electricity. Recent global studies have emphasized the core role of renewables in zero-carbon electricity. This is because the costs of renewable energy have plummeted (especially solar photovoltaics), while nonrenewable energy sources—nuclear, biofuels, and carbon capture—each pose major technical and social obstacles leading to public opposition. Nuclear power of course raises massive concerns over nuclear accidents and nuclear wastes. (See the companion article by Alexander Sammon.) Biofuel raises great concerns about ecosystem degradation and competition with food supplies. Carbon capture raises massive opposition over technological doubts (e.g., leakage of CO2), land use obstacles (e.g., pipelines to carry the CO2), and high costs.

The second pillar of decarbonization is the electrification of end uses. There are many sectors currently using fossil fuel energy that can be converted to direct use of (green) electricity. These include battery electric vehicles (BEVs), heat pumps for residential and commercial buildings, electric cooking (e.g., induction and microwave stoves), and direct reduction of ores in metallurgy. These pathways of electrification seem more likely today than just a few years ago. Major automakers are now making significant commitments to electric vehicles, for example, with dates set for the phaseout of conventional internal combustion engine vehicles.

The third pillar is green synthetic fuels for economic sectors not easily electrified. In aviation transport, there is continuing debate about the feasibility of electrification. It seems increasingly likely that electrification will cover short-haul flights (e.g., under one hour) but that longer-haul flights will continue to require liquid fuels of high energy density. These synthetic fuels can be manufactured using green electricity. (See the companion article by Mara Prentiss.)

The fourth pillar is a smart power grid, built on big data, artificial intelligence, and the Internet of Things. The idea of a smart grid is a self-regulating system that can shift among multiple sources of power generation and multiple uses to provide reliable and low-cost systems operations despite the variability of renewable energy. On the supply side, a smart grid will integrate variable renewable energy from many sources in order to smooth the variability of power generation. A larger connected grid, covering more geography and more sources of variable renewable energy, will reduce variation of power. Various storage options, including batteries, pumped hydro, compressed air, and conversion of renewable energy into synthetic fuels, will help to stabilize the supply side. The demand side will also show flexibility by enabling smart meters to turn on and off electricity consumption of users depending on temporal needs, urgency, and shifts in market prices that reflect supply-demand conditions.

The fifth pillar is energy and materials efficiency to economize on the use of primary energy, and on the plastics, metals, cement, and other industrial materials that emit CO2 in their production and use. Improved materials and material flows, popularly known as “reduce, reuse, and recycle” or “circular economy,” can significantly improve energy and materials efficiency, reduce the process emissions of CO2 (such as in the manufacture of clinker for cement), and slash energy inputs needed for industrial processes.

Easier and Harder Sectors

Decarbonization of electric power will be relatively straightforward, though there remain important challenges to managing power grids that are fully dependent on renewable energy. Low-carbon power sources such as photovoltaics and wind have already come down dramatically in cost so that their levelized costs (that is, their long-term costs including the costs of capital investments) are already competitive with fossil fuels.

The biggest operational challenges arise from the intermittency of the renewables and their limited dispatchability (the relative inability of renewable-energy sources to ramp up and down quickly in response to market needs). There are two complementary solutions to intermittency. One is power storage, for example in grid-scale batteries, synthetic fuels (e.g., hydrogen), and systems such as pumped hydro, in which excess renewable energy is used to pump water into an uphill reservoir for later use in hydroelectric power generation. The other solution is geographical diversification of renewable-energy sources through a large interconnected grid. A more extended grid reduces the variability of power generation relative to the average load and therefore reduces the need for energy storage as a percentage of the average load. As for dispatchability, storage solutions such as pumped hydro and synthetic fuels (e.g., hydrogen) offer the necessary dispatchability.

We need to enter a sustained phase of global net negative CO2 emissions so that these concentrations gradually return to 350 ppm over the long term.

Some downstream users of fossil fuels will be easily electrified, while others will be difficult or impossible to electrify. Light-duty vehicles and urban delivery vans, for example, will almost certainly shift soon from internal combustion engines to zero-emission battery electric vehicles or hydrogen fuel cell vehicles. The hydrogen will have to be produced in a zero-emission manner, such as by hydrolysis using renewable energy or by fossil fuels combined with carbon capture and storage. Other transport modes—long-distance trucking, ocean shipping, and aviation—are “hard sectors” in that electrification is much less straightforward or out of reach. Onboard batteries are heavy and take up considerable volume needed for freight. Other solutions, such as synthetic green fuels or biofuels for aviation, hydrogen fuel cells for ocean shipping, and catenary lines for electric trucks along major highway routes, will be necessary.

The recent report by the Energy Transitions Commission, “Mission Possible: Reaching Net-Zero Carbon Emissions From Harder-to-Abate Sectors by Mid-Century,” and the FEEM-SDSN report, “Roadmap to 2050: Power, Industry, Transport and Buildings,” discuss the technology possibilities for the harder sectors.

Most new buildings will be relatively easy to electrify using electric heat pumps and overall improvements in insulation and ventilation. Older buildings that currently rely on fossil fuel heating will have to be retrofitted for electric heat pumps. Retrofitting will require a long national effort and will be moderately expensive. Major challenges remain in decarbonizing certain industrial sectors, including metallurgy, petrochemicals, cement, and some other heavy industries that intensively use fossil fuels for process heating and feedstocks. There will be no one-size-fits-all strategy. Certain kinds of process heating can be electrified; others not, at least currently. Certain feedstocks can be replaced by non-carbon-emitting materials; others not, at least currently.

A Timeline for Decarbonization

The essence of decarbonization is the replacement of today’s fossil fuel–using capital stock with a new zero-carbon capital stock. Coal-fired power plants need to be phased out and replaced by renewable-energy power generation. Internal combustion engine vehicles need to be phased out and replaced by electric vehicles. Boilers and furnaces in buildings need to be phased out and replaced by electric heat pumps. And so forth.

The least-cost solution in each case is to retire the existing capital at the end of its normal life and replace it with zero-carbon capital. Cars and trucks last 15 to 20 years; power plants 30 to 50 years; buildings 50 to 100 years; and so forth. This means that with the natural life span of vehicles, we would require 15 to 20 years from the first date at which all new vehicles brought to the market are zero-emission vehicles. The alternative strategy, which is more costly, is to scrap the existing capital stock early, for example by removing even new internal combustion engine vehicles from the road. Of course, if we factor in the costs of the worsening climate, the cost-benefit calculus changes.

Gerry Broome/AP Images

Consider, for example, the challenge of decarbonizing the U.S. fleet of some 200 million light-duty vehicles. Suppose, as an illustration, that cars last for 20 years, and that ten million vehicles are currently retired each year and replaced with ten million newly produced vehicles. The industry’s production capacity is geared to ten million sales per year. In order to shift the U.S. automobile fleet to electric vehicles, the industry must be retooled.

Let us stipulate for purposes of illustration that converting the U.S. automotive industry to the production of ten million BEVs per year will require a decade, providing the time not only to retool existing production lines and to design the new vehicles, but also the time to build new supply chains for batteries and other components. As of 2030, all new U.S. vehicle production and sales will be electric. Between 2030 and 2050, the entire fleet of 200 million internal combustion vehicles will be phased out and replaced by electric ones.

Could this happen, instead, by 2040? That would require replacing 200 million vehicles during a ten-year period, 2030 to 2040. Annual production and sales of electric vehicles in the 2030s would have to average 20 million vehicles, twice the current industrial capacity. Yet after 2040, production and sales would fall for many years because of the young age of the BEV fleet. The production boom would be replaced by a bust. In the long run, production would revert to ten million per year on average. Moreover, the early conversion of the fleet would only reduce emissions to the extent that the U.S. power system also had been decarbonized and expanded by 2040 to accommodate the 200 million electric vehicles.

Perhaps the early replacement of internal combustion engine vehicles by 2040 could be accomplished, but the extraordinary costs of the early scrappage of existing vehicles, the massive rise in overall vehicle production to 20 million per year during the 2030s, and the subsequent closure of industrial capacity after 2040 would effectively require the nationalization of the automobile industry. And the emissions from the temporary boom of automobile production could easily overwhelm any emissions reductions achieved by the early scrappage of the internal combustion engine fleet.

This simple parable is meant to illustrate a point. We will need until 2050 to achieve full decarbonization. Even then, we will be incurring many extraordinary costs to meet the mid-century target, including the early closure of hundreds or even thousands of fossil fuel–based power plants. Yes, we should have started decarbonization in earnest in the 1990s, by adopting the Kyoto Protocol. Instead, it was spurned by the U.S. Senate. So here we are in 2020 with far too little time left for climate safety. Yet we must proceed.

Zero-carbon power generation can probably be achieved nationwide by 2040 to 2045. New York state has set 2040 and California has set 2045 as respective target dates. Given the older ages of most existing coal-fired power and many gas-fired plants, such a timeline would permit the natural retirement of most carbon-based power generation, with only a modest need for early closures. Market forces will also lead to some early closures, as the capital-inclusive costs of new renewable-power generation decline below the marginal operating costs of some existing thermal plants. The conversion of the automobile and trucking fleet will require most of the period to 2050. We should anticipate—and regulate—that as of 2030, all new sales of light-duty vehicles will be zero-emission vehicles, mostly battery electric vehicles. New buildings could become all-electric within the next few years, since the technologies to build all-electric heating and cooking systems are already at hand. The retrofitting of existing buildings, however, will take far more time, almost surely to 2050.

The Incremental Costs of Decarbonization

The best current estimates put the incremental cost of decarbonization at around 1 percent of GDP each year or perhaps even less. Decarbonization will neither break the budget nor massively spur the economy by itself. But yes, it can save the planet from environmental ruin.

The extra costs incurred to shift from fossil fuels to zero-carbon energy include the incremental costs of providing electricity with renewables rather than fossil fuels; of shifting to zero-emission vehicles; of buildings using green electricity for heating, ventilation, and cooking, rather than fossil fuels; and the incremental costs of decarbonizing several “hard” sectors, including aviation, ocean shipping, steelmaking, cement, and petrochemicals.

Smart meters turn electricity consumption on and off depending on the needs of users, and shifts in market prices that reflect supply and demand conditions.

The incremental costs also include all the complementary infrastructure needed to operate a zero-carbon energy system. For example, the U.S. will need additional grid transmission to connect high-quality renewable energy (Southwest solar energy, Midwest wind energy, Canadian hydropower, and offshore wind energy) with final users. New charging stations across the nation will be needed for electric vehicles. New catenary lines (overhead electric lines) may be added to interstate highway lanes for electric trucks. Extensive new investments in 5G capacity will be needed to manage smart grids. And so forth.

Several recent studies have nonetheless concluded that the overhaul of the energy system will not be large relative to the economy. My Columbia colleague Geoffrey Heal provided an especially insightful estimate of transition costs, showing that renewable energy plus storage can cut U.S. emissions by 80 percent by 2050 for less than 1 percent of GDP per year.[3] Another colleague, Jim Williams of the University of San Francisco, has similarly calculated that an 80 percent reduction of U.S. emissions by 2050 could be accomplished for around 0.8 percent of GDP per year.[4]  A recent study on global decarbonization by 2050 led by Christian Breyer at Lappeenranta University of Technology in Finland concludes that the levelized costs of a 100 percent renewable-energy system in 2050 will be less than the costs today (52 euro/MWh compared with 70 euro/MWh today). And a 2015 study by Alexander MacDonald and co-authors found that CO2 emissions from the U.S. electricity sector can be reduced by 80 percent relative to 1990 levels without any increase in the levelized cost of electricity. Moreover, the transition to renewable energy will create many more jobs than will be lost in the closure of the fossil fuel sector. (See the companion article by Harold Meyerson.)

A major ongoing study by the National Renewable Energy Laboratory (NREL) reportedly finds that a U.S. “super-grid” connecting high-quality renewable-energy resources with major population centers would cost around $80 billion in total, a mere 0.4 percent of GDP as a one-time investment, and would deliver economic benefits of at least twice that sum. There are fears that the NREL findings are being suppressed or delayed by the Trump administration.

I should underscore that most of the studies to date examine the incremental costs of 80 percent decarbonization rather than 100 percent decarbonization. It is likely true that the marginal costs of decarbonization will rise as we approach 100 percent. As one example, Jessika Trancik’s team at MIT has determined that the storage costs to back up 95 percent of an all-renewable power system would be roughly half of the costs to back up 100 percent of the system, since 100 percent backup requires vastly more storage to protect against an extremely rare shortfall of intermittent energy.

In summary, the gist of recent studies is that decarbonization is not hugely expensive as a proportion of the total economy. The estimates suggest that 80 percent decarbonization can be reached at a cost of 1 percent of GDP or less per year.  Complete decarbonization by 2050 could perhaps cost up to 2 percent of GDP per year, taking into account the higher marginal costs of decarbonizing the harder sectors, but could in fact end up being much less costly than 2 percent of GDP, and even below 1 percent of GDP. 

Would it be worth the extra costs to accelerate U.S. decarbonization to 2040 rather than 2050, assuming that it would be technically feasible? The answer is no. U.S. CO2 emissions from energy in 2020 will be around 5.3 billion tons. If we compare a linear ramp-down to zero by 2050 and a faster linear ramp-down by 2040, the faster ramp-down would reduce U.S. cumulative emissions of CO2 by around 27 billion tons. That in turn would reduce the atmospheric concentration of CO2 by around 1.8 ppm, with reduced global warming on the order of 0.01 degrees as of 2100—in other words, negligible. The short-term effect on temperature would be even less. And we must remember that an accelerated decarbonization might not actually cut cumulative emissions as assumed by a steeper linear ramp-down because of the larger buildup of new industrial capacity (e.g., electric vehicle production) needed in the accelerated transition.

Even if the entire world were somehow to decarbonize on a linear ramp-down by 2040 rather than 2050, the reduction of global warming as of 2100 would be on the order of 0.1 degrees, with a smaller short-term effect. Yes, such a reduction in warming would definitely be beneficial; each 0.1 degree warming hurts. Yet the practical time needed to phase out the world’s existing fossil fuel capital stock, build massive new infrastructure, retool major global industries such as automotive production, and solve countless technological challenges, all while attending to the urgent sustainable-development needs of the more than six billion people in developing countries, argues overwhelmingly for a timeline to 2050 rather than 2040. Decarbonizing globally by 2050 will itself be a near-miracle, one that requires a global breakthrough in politics and policies starting imminently and carried forward for decades.

A National Policy Framework for Decarbonization

The national policy framework for decarbonization should aim at three overriding objectives: (1) the end of U.S. energy-based emissions of CO2 by 2050; (2) a low-cost pathway for the transition; and (3) a fair transition to address vulnerable groups and regions, including workers in fossil fuel–related sectors who will lose their jobs in the transition; regions currently dependent on fossil fuel production; and low-income households.

The Green New Deal should involve all levels of government, including state and local, rather than put all responsibility at the federal level.

An effective plan should involve distinct strategies for each of the four major energy sectors: power generation, transport, buildings, and industry, with special attention paid to key subsectors, such as light-duty vehicles, intercity trucking, aviation, and ocean shipping. The plan should distinguish between the easy and hard sectors, including R&D for the hard sectors. The plan should address several dimensions of public policy, including:

Carbon regulation and pricing. There are two basic policy approaches to pollution control that also apply to decarbonization: (1) quantity regulation and (2) corrective pricing. Under quantity regulation, the government sets quantitative limits on pollution (either in absolute amounts or per unit of output) and timelines to abide by the limits. Under corrective pricing, the government taxes the pollutant and/or subsidizes the green alternatives. In the case of decarbonization, both quantity regulation and corrective pricing will be needed, but the relative balance between the two should vary according to the sector of the economy.

Quantity regulation is the preferred option when the technological alternatives are well known, easy to monitor, and already cost-effective. Corrective pricing makes sense when the alternatives are uncertain, difficult to monitor, and with highly varying cost-effectiveness depending on the context. In those cases, a corrective price (e.g., a tax on the pollutant) allows markets to search for low-cost solutions and promote innovations.

The most important international pollution abatement to date has been the highly successful global phaseout of ozone-depleting chlorofluorocarbons (CFCs). This phaseout was governed by the 1987 Montreal Protocol and subsequent amendments. In this case, the pollutant (CFCs) had a known and cost-effective alternative, the hydrofluorocarbons (HFCs). The Montreal Protocol thereby relied on quantity regulation. All signatory governments to the Montreal Protocol agreed to timelines for the phaseout of CFCs. The U.S. also deployed a bit of corrective pricing mainly in the form of import levies to block the importation of CFCs while U.S. production of CFCs was being curtailed. Another notable success story of quantity regulation is the removal of lead from gasoline, a process mandated by federal law rather than carried out by corrective pricing.

John Minchillo/AP Images

In the case of decarbonization, quantity regulation has been used for power plants both at the federal level (such as President Barack Obama’s Clean Power Plan to phase out coal-fired power plants) and at the state level (for example, in state renewable portfolio standards that require utilities within the state to phase out fossil fuel–based power generation on a timeline). Quantity regulations have also been used to raise the fuel efficiency of the U.S. vehicle fleet under Corporate Average Fuel Economy (CAFE) standards. Corrective pricing, by contrast, has also been used in a variety of ways to induce profit-oriented utilities to choose zero-carbon power solutions. These include tradable emissions permits, where the market price of the permit creates the incentive to shift technologies; feed-in tariffs, whereby utilities are given a bonus per unit of low-cost energy; and reverse auctions for the supply of zero-carbon power (the low bidder wins the supply contract). No state has yet adopted a carbon tax, but it is under consideration in several states

I would suggest that quantity regulations for decarbonization will be most effective for four sectors: power generation, light-duty vehicles, heavy-duty vehicles, and buildings. For these sectors, the federal government should implement a timeline for phasing out the fossil fuel–based capital stock. For example, the federal government might mandate that all new power generation capacity in the U.S. should be zero-carbon after a certain date, say 2022, and that remaining fossil fuel–based power generation should be phased out by a certain date, say 2040. Similarly, the federal government should mandate that all new light-duty vehicles sold from 2030 onward must be zero-emission vehicles, and that internal combustion light-duty vehicles must be phased out no later than 2050. A somewhat extended timeline might be given for trucks, e.g., all new trucks must be zero-emitting as of 2035. For buildings, a federal building code for new construction might mandate zero-emission buildings (electric heating, ventilation, cooking) from 2030 onward, and retrofits on existing buildings by 2050. Such quantity regulations can be supplemented by a gradually rising carbon tax that will accelerate the transition process and ensure the retirement of old carbon-emitting capital.

Corrective pricing makes more sense as a primary instrument for sectors in which the best alternative technologies and incremental costs are still highly uncertain. Thus, we should consider a gradually rising carbon tax on aviation, ocean shipping, steelmaking, cement, and petrochemicals. Since the capital stock in these sectors rolls over gradually, the carbon pricing should be introduced with enough lead time to enable companies to make changes in planned future investments. Thus, the carbon tax on aviation, ocean shipping, and other sectors might be introduced into law but begin only as of 2025 or 2030. This will avoid an arbitrary rise in taxes on current capital, but will incentivize firms and industrial sectors to replace that capital in the future.

The Green New Deal should avoid the social eruptions that have recently hit France, Chile, and elsewhere where governments introduced a new levy on transport that essentially taxed the existing capital (e.g., automobile use), often in a regressive manner, without any real opportunity for those hit by the levies to change their behavior in the short term. Carbon taxes should be phased in gradually, and focus on future decisions, rather than tax existing carbon-using capital that only serves to redistribute income regressively without affecting short-term emissions to any significant extent. (See the companion article by James Boyce.)

Federal, state, and local responsibilities. The regulation of the U.S. energy system is shared among the federal, state, and local governments. The Green New Deal should tap into this intergovernmental structure rather than put all responsibility at the federal level. Bypassing the states would create huge inefficiencies, political backlashes, and serious obstacles for ongoing state-level regulation of the utilities.

The federal government should lead on nine dimensions of the Green New Deal:

State renewable portfolio standards. Even with a federal mandate on all utilities to decarbonize by 2050, the states will still have responsibilities for regulating the power utilities within each state, including the licensing of sites, the regulation of transmission and distribution lines, system safety and reliability, environmental protections, and of course electricity pricing. Each state should be directed to adopt and implement a renewable portfolio standard (RPS) program that is consistent with the federal mandate to decarbonize by 2050.

To defeat Big Oil, advocates need a specific plan that demonstrates how renewable-energy alternatives will benefit every part of the country.

The Lawrence Berkeley National Laboratory (LBNL) reports that 29 states have RPS requirements to raise the share of renewable energy in overall retail sales. Some non-RPS states, such as Indiana, North Dakota, and Wyoming, have also increased their renewable-energy capacity to serve RPS demands in nearby states. The 29 RPS programs cover 56 percent of total U.S. retail electricity sales. Three states have recently set dates for 100 percent zero-carbon power in their RPS: California (2045), New Mexico (2045), and New York (2040). In total, ten states plus the District of Columbia and Puerto Rico have set zero-by-2050 targets into law or executive order: California (2045), Hawaii (2045), Maine (2050), Minnesota (2050), Nevada (2050), New Jersey (2050), New Mexico (2045), New York (2040), Washington (2045), and Wisconsin (2050).

The RPS requirements provide a very important institutional mechanism for implementing federal zero-emission power standards. RPS programs to date have mandated 45 percent of the increased delivery of renewable-power generation since 2000. In 2018, 37 percent of solar-capacity additions and 19 percent of wind additions were to meet RPS requirements.

Some of these renewable-energy investments would have happened without RPS, but RPS surely played a significant role, not only in directing the utilities toward zero-carbon energy, but supporting them to do so in an efficient manner consistent with overall objectives of power plant siting, low costs to consumers, system reliability, and other objectives. At the same time, the current power crisis of California’s PG&E points to the reality of significant underinvestments in systems infrastructure when private-sector utilities pursue short-term profits at the expense of long-term standards.

Financing Decarbonization

There are two basic ways to finance the energy transition. The first is through market transactions. The government mandates zero-carbon technology, and private-sector producers invest and sell goods and services to the public. For example, utilities are required to invest in renewables, and they recoup their costs by their sales to households and businesses. Automobile producers are mandated to sell zero-emission vehicles, and they recoup their costs through vehicle sales. In this case, the direct role of government financing is very limited. The second is through direct government provision of zero-carbon energy. In this case, the federal or state governments would directly own and operate power generation facilities and transmission and distribution grids. They would recoup some or all of their costs through market sales by public enterprises, or alternatively finance their operations through general government revenues.

There will be a mix of the two. Outlays for certain public infrastructure will require government financing, for example the expansion of the interstate transmission grid for renewable energy. As a natural monopoly, the expanded transmission grid would be unsuitable for private ownership. Another case will be the federal outlays for a fair energy transition, which by design will not recoup a flow of earnings. Federal support for building retrofits will also likely be grants rather than loans.

Most of the transition costs, therefore, will be borne not by the federal and state governments but by energy users: household and commercial buyers of electricity, car owners, and others. Since the zero-emission technologies are already close to the costs of fossil fuel–based technologies, there is no reason to anticipate any major hardships on energy users. As noted earlier, the total national costs might come to around 1 to 2 percent of national income per year, roughly $200 to 400 billion, with most of that borne by the private sector. There are of course many technological uncertainties, and the total costs could end up higher or lower. Indeed, as the costs of clean energy continue to fall and as conservation measures improve (improved building insulation, more efficient appliances), many users could experience net savings on energy bills.

Total federal outlays could reach up to $400 billion per year, or roughly 2 percent of GDP, if we factor in the total costs of new infrastructure spending on roads, fast rail, protected coasts and waterways, restored bridges, expanded transmission grids, and other infrastructure that is needed in any event given the decrepit state of U.S. infrastructure. This added spending is not the incremental cost of decarbonization per se, but the cost of restoring and modernizing America’s overall infrastructure, a worthy and much-needed objective. Research and development outlays should also rise significantly. We may estimate that research outlays for renewable energy should be of roughly the same scale as the biomedical research budget of the National Institutes of Health, roughly $30 billion per year.

It is notable that Senator Bernie Sanders has presented a multiyear Green New Deal plan with a headline price tag of $16 trillion. The main reason for this enormous sum is Sanders’s call for the federal government to build and operate the new renewable-energy system, essentially displacing the existing utility industry in the process. In the plan that I have sketched, I assume that the utility sector, not the federal government, will bear the investment costs. Sanders also includes generous outlays for retrofitting buildings and for a highly accelerated replacement of the existing vehicle fleet. My cost estimates, and those of the studies I have quoted, are based on decarbonization by 2050 in line with the IPCC 1.5 degree scenario.

Getting Started in 2020 for the Next Presidency

Proponents of the Green New Deal will need to do their homework if they are to triumph in Congress over the Big Oil lobby. Public opinion is in favor of climate action, but that doesn’t matter in Congress. What counts is who pays the campaign bills. Unfortunately, the answer is Big Oil. In the 2016 federal election cycle, the oil and gas industry gave $56 million in campaign financing to Republican candidates compared with just $8 million to Democratic candidates. In the 2018 election cycle, the Republicans received $43 million compared with just $6 million for the Democrats.

To beat Big Oil, the Green New Deal advocates need a specific plan: one that demonstrates how decarbonization will work, and how it will benefit every part of the country. With such a plan, Green New Deal advocates will be able to turn public opinion into votes in Congress. Every congressman, indeed every voter, should have a specific idea of what the Green New Deal would mean for their district and region. Without such a plan, climate activists will continue to win the battle over climate science but still lose the war over climate action.

To date, advocates of decarbonization have tended to focus their advocacy on pricing policies (cap-and-trade, carbon tax, feed-in tariff, etc.). Yet such pricing proposals only serve to raise suspicions and opposition to new taxes. Such policy proposals fail to win the hearts and minds of the general public, and the public fails to press the Congress for action.

A historical analogy may be useful. When President Dwight Eisenhower first proposed the Interstate Highway program in 1953, conservatives in Congress were reluctant to support the program. The congressmen could not see what was in it for them and the proposal was stalemated. Then, in September 1955, the Bureau of Public Roads of the Department of Commerce put out a book of maps showing the “general location” of the proposed highway system in dozens of major metropolitan areas. Suddenly, congressmen could see the benefits of a new highway system passing through their district. The vividness of the proposed plan helped to carry the day in Congress. (See the article by Robert Paaswell.)

The good news is that the specifics of a Green New Deal to decarbonize the energy system are finally coming into focus. We now understand clearly the key pillars of decarbonization. We now understand that today’s technologies can get most of the transition accomplished, and that even the “hard” sectors can be decarbonized after a bit more R&D. And the truly great news is that all of this can be achieved at very low cost. Decarbonization to save the planet is actually the greatest bargain of our time.


[1] James Hansen et al. “Young People’s Burden: Requirement of Negative CO2 Emissions,” Earth System Dynamics 8, 577–616, 2017.

[2] There is, in fact, a sixth pillar that will be needed in a comprehensive national climate policy: sustainable land use, mainly involving the agriculture sector. Agriculture contributes CO2 from deforestation and land degradation, and also non-CO2 greenhouse gas emissions, including methane released by ruminant animals (especially cattle) and flooded rice paddy, and nitrous oxide emissions from nitrogen-based fertilizers. Agriculture must be part of an integrated strategy, but is beyond the scope of this article.

[3] Geoffrey Heal, “Reflections—What Would It Take to Reduce U.S. Greenhouse Gas Emissions 80 Percent by 2050?” Review of Environmental Economics and Policy, Volume 11, Issue 2, Summer 2017, 319–335.

[4]Jim Williams, “Decarbonizing the United States: Challenges of Scale, Scope, and Rate” (lecture, National Academy of Sciences) July 22, 2019.

Back to topbutton