Smoke is released into the sky at a refinery in Wilmington, California March 24, 2012. Picture taken March 24, 2012. REUTERS/Bret Hartman  (UNITED STATES - Tags: ENERGY ENVIRONMENT BUSINESS INDUSTRIAL COMMODITIES) - RTR2ZXLT

Growth, carbon, and Trump: State progress and drift on economic growth and emissions ‘decoupling’

Devashree Saha and Mark Muro

“Decoupling” economic growth from the growth of carbon emissions is critical for the world, nations, and states alike because it represents the “squaring of the circle” necessary to decarbonize the global economy while maintaining economic growth.


In recent years, at least 35 countries, including the United States, have increased their real gross domestic product (GDP) while reducing their carbon dioxide (CO2) emissions. This success is an encouraging juncture in the campaign to limit global warming, and would seem to license cautious optimism.

Yet now all of that is in question. With the stunning election of Donald Trump to the presidency, every aspect of the low-carbon paradigm for national and world progress has been thrown into doubt, starting with the federal government’s support of the quest to “decarbonize” the economy by decoupling economic growth from emissions growth.

All of which raises the question of how resilient the decarbonization paradigm is at the state and local level. Given their substantial powers to encourage emissions decoupling, states and cities are crucial players in the carbon drama.  Therefore, it is worth assessing whether states’ and localities’ momentum on decoupling is strong enough to maintain recent progress.

And so this brief takes a look at state-level decoupling trends by matching data on real GDP growth between 2000 and 2014 for all 50 states and the District of Columbia with data on energy-related carbon dioxide emissions for the same years and locations. In doing so, the brief provides an initial baseline look at the pace and geography of state-side decoupling and decarbonization—with an eye to assessing state-level momentum on the brink of federal pull-back.

What do these data show? Overall, they show that more than 30 states have delinked their growth and carbon emissions, confirming that economic growth does not inevitably require emissions growth.   At the same time, the findings of this report provide a layered assessment that runs as follows:

  • Decoupling is occurring in most U.S. states, confirming that cleaning up the economy doesn’t necessarily put an end to growth
  • However, the pace and extent of decoupling varies greatly
  • Market forces and changes in states’ economic structures are one set of influences on carbon emissions and decoupling
  • Region and fuel mix matter a lot
  • Despite significant progress, all states need to do more to decouple emissions from growth and decarbonize their economies in keeping with the goal of holding global mean temperature increases to 2 degrees Celsius

As to the future, states and cities are going to need to do much more to maintain and accelerate the pace of local, national, and global decarbonization. With responsibility for U.S. decarbonization now devolving to the states and cities, state and regional actors will need to fill the vacuum created by Washington’s abdication of leadership with new energy and resolve.

Read the full brief here »


In April 2016, when the presidential primaries were just warming up and Donald Trump’s candidacy was a longshot, separate analyses by the World Resources Institute (WRI) and the UK-based Carbon Brief concluded that at least 35 countries, including the United States, had increased their real gross domestic product (GDP) over the last 15 years while actually reducing their carbon dioxide (CO2) emissions.1

Earlier, the International Energy Agency (IEA) had found that the world’s emissions remained flat in 2014 and 2015 even as GDP continued to grow by more than 3 percent in each year.2

In sum, an important juncture—anticipated for years—had finally been reached. For the first time, the globe’s growth and emissions had “decoupled.” For a moment, the data warranted a small measure of encouragement that the world’s nations might be able to reduce greenhouse gas (GHG) emissions enough to limit global mean surface temperature increases to less than 2 degrees Celsius above preindustrial levels, an objective agreed to at the Paris climate summit.3

Yet now all of that is in question.

With the stunning election of Donald Trump to the presidency, every aspect of the low-carbon paradigm for national and world progress has been thrown into doubt, starting with the government’s support of the push to  “decarbonize” the economy by decoupling economic growth from emissions growth.4

Trump, after all, openly disavows the concept of human-caused climate change; has threatened to pull the United States out of the Paris climate agreement, through which 175 nations have pledged to cut back on emissions; and promises to scrap major regulations like the Clean Power Plan, which aims to reduce CO2 emissions from coal-fired power plants. More broadly, the president-elect does not subscribe to the paradigm of decarbonization for the world or the nation. All in all, it seems fair to say that federal climate policy will all but disappear in the coming years.

All of which raises the question of how resilient the decarbonization paradigm is at the state and local level at a moment of federal retreat.

States and cities are crucial players in the carbon drama because global and national emissions reductions originate in specific locations.

States and cities matter, in this regard, because they control much of the legal and policy power that can control emissions.5 State public utility commissions regulate investor-owned electric utilities. State legislatures set and update renewable energy and energy-efficiency targets and portfolio standards. And states shape land-use rules, building codes, and transportation systems that are implemented and delivered by municipalities and metropolitan planning organizations. In fact, it’s safe to say that the largest share of America’s past progress on cleaning carbon out of the economy in recent decades has owed to state and local efforts.6

Which is why state and local progress on emissions matters more than ever as Donald Trump takes office. With the likely coming abdication of federal leadership on energy and climate issues, cities and states—which are already important laboratories for clean energy problem-solving—represent a possible counterbalance to a climate-skeptic federal government.

More than 30 states have delinked their growth and carbon emissions, confirming that economic growth does not inevitably require emissions growth

And so this brief looks at state-level decoupling trends by matching data on real GDP growth between 2000 and 2014 for all 50 states and the District of Columbia with data on energy-related carbon dioxide emissions for the same years and locations.7 In doing so, the brief provides an initial baseline look at the pace and geography of state-side decoupling and decarbonization—with an eye to assessing state-level momentum on the brink of federal pull-back.

What do these data show? Overall, more than 30 states have delinked their growth and carbon emissions, showing that economic growth does not inevitably require emissions growth. At the same time, the pace and degree of states’ decoupling vary widely, with distinct regional dynamics.

Pinpointing the precise factors that are influencing these outcomes is mostly beyond the scope of this analysis. However, it is clear that many states have made progress in delinking emissions from growth through the replacement of coal-burning power plants with natural gas-fired plants or, in some cases, renewables.8 Nuclear generation and changes in states’ industrial structure have played a role as well.9 And while formal statistical analyses of the role of clean energy policy in decarbonization are also beyond the scope of this study, it is fair to say that state- and city-level policy choices have also contributed to decoupling and decarbonization.10

In short, while decarbonization is not happening nearly fast enough in most states, as a final analysis here shows, state-level policy choices and economic trends have fostered solid momentum that in many places could proceed even without federal leadership.

Yet even still, the outlook is more challenging than ever. Continued progress on decoupling will require heroic, expanded efforts by more states—and hopefully some degree of support from Washington.

How and where growth and carbon have decoupled

Why does carbon decoupling matter? Decoupling is important because it represents the squaring of the circle—the accomplishment of a task some have said was impossible—necessary to decarbonize the global economy.

Deep reductions in carbon emissions will be necessary to transition the world to a low-carbon economy consistent with the internationally agreed goal of limiting human-associated global warming. Holding the world’s climate to 2 degrees Celsius of warming, as the world community has agreed is imperative, will require cutting the world’s greenhouse gas emissions 40 to 70 percent from their 2010 levels by 2050 and to zero by the century’s end.11 Such an accomplishment will entail a profound transformation of the economy.

At the same time, governments across the globe—especially in the developing world—are faced with the challenge of promoting faster economic development. Even in the United States an often torpid recovery from the Great Recession has prompted calls for faster growth.12

These twin demands combine to form the challenge of the present (one that Donald Trump rejects): Can the world decarbonize the economy and erase emissions by the end of the century while maintaining or accelerating growth?

For most of the 20th century, the possibility of preserving growth and erasing emissions remained theoretical and was most commonly presented as an unsatisfactory, divisive face-off between the “growth imperative” and the “climate imperative.” Consequently, conventional views on the relationship between economic growth on the one hand and energy consumption and greenhouse gas emissions on the other generally claimed that economic growth would lead to an increase in production and therefore to an increase in energy use and emissions.13

Yet in recent decades, more and more data have confirmed that it is possible to address global climate challenges while preserving economic growth and prosperity. (See Figure 1.)

WRI’s analysis of 67 countries, for instance, shows that in the United Kingdom carbon emissions declined even as real GDP grew in five different years between 2000 and 2014.14 Over the full 14-year period, the United Kingdom reduced its carbon emissions while growing its economy by 27 percent. Similarly, Carbon Brief’s global analysis found that, while 45 nations reduced their emissions between 2000 and 2014, 35 did so while increasing their real GDP.15 Intensely urban Singapore, for example, scored the most dramatic decoupling, as it doubled its real GDP while slashing its CO2 emissions by 46 percent.


The United States first decoupled its economic growth and emissions in 2001, when it achieved a modest 2 percent reduction in carbon emissions while growing its economy by 1 percent.16 That was a recession year, but decoupling occurred again in 2006, a year of solid growth, before relapsing. More recently, the nation’s emissions and growth decoupled from 2010 to 2012, and then again in 2015. Emissions decoupling has clearly become more frequent amid the ongoing large-scale switch from coal to natural gas—driven by the hydraulic fracturing (“fracking”) boom. At the same time, numerous other factors are clearly influencing outcomes, ranging from changes in the structure and growth of the national economy to investment decisions and technology change to land-use change and the availability of clean new energy resources, including renewables.17

Hence the present assessment: To examine decoupling trends at the state level, this analysis links state-level GDP data to state-level carbon emissions information to provide a simple look at states’ progress on decarbonization. Secondarily, some observations are made about some of the state-side influences on those trends, especially those involving states’ fuel mix and industry structure.18 Finally, in concluding, a few observations are offered about the prospects for maintaining progress in the current policy environment.


  • Decoupling is occurring in most states, confirming that cleaning up the economy doesn’t necessarily put an end to growth

The Trump-style climate view frequently depicts efforts to curb emissions as a drag on economic growth. However, between 2000 and 2015 (the latest year available for national data), the United States expanded its GDP by 30 percent while cutting its emissions by 10 percent—making it the largest country that has had multiple years in which economic growth has been decoupled from growth in carbon emissions (see Figure 2).19 Many states have managed similar feats, which in turn confirms the compatibility of economic growth and emissions reductions and counters the Trump-style skepticism.

Metro_20161206_decoupling fig 2

Altogether 33 states and the District of Columbia managed to expand their economies between 2000 and 2014 while reducing their carbon emissions (see Figure 3 and Appendix Table A). As a group, these jurisdictions expanded their economies by 22 percent while reducing their emissions by nearly 12 percent.20 Maine achieved the largest CO2 decline among the 50 states, at 25 percent, while growing its economy by 9 percent.21 Among the larger states (in terms of GDP), Massachusetts, New York, and Georgia have seen some of the largest reductions in emissions since 2000. Massachusetts managed to cut its emissions by 22 percent even as its GDP grew 21 percent. New York and Georgia decreased their emissions by 20 percent and 17 percent while growing their GDP by 24 percent and 15 percent, respectively.

The pace of decoupling, meanwhile, has accelerated over time, with more and more states breaking the historically tight link between GDP growth and increased carbon emissions. For instance, only 14 states and the District of Columbia managed to sever the link between growth and emissions between 2000 and 2007. Among these pioneers were five New England states—Connecticut, Maine, Massachusetts, Rhode Island, and Vermont—as well as New York. As a group, these early decouplers reduced their aggregate carbon emissions by nearly 4 percent between 2000 and 2007 while expanding their economies by 18 percent. However, since 2008, coinciding with the onset of the Great Recession, the number of decoupled states has doubled, with new states such as California, Georgia, New Hampshire, South Carolina, and Virginia joining the ranks of decoupled states (while a few from the earlier group, such as Connecticut, Maine, and Nevada, fell off the list).22 Across this 36-state group of 2008-2014 decouplers, the average reduction in carbon emissions was 10 percent and GDP growth was 6 percent.

Overall, as discussed in more detail below, the largest reductions in energy-related carbon emissions, especially after 2007, can be attributed to the fuel use changes in the electric power sector.

More broadly, President-elect Trump’s notion of an opposition between economic growth and environmental stewardship appears to be a false one.

  • The pace and extent of decoupling vary greatly

metro_2016120_emissions_GDP_martix-1Decoupling, meanwhile, has spread widely in the last decade, but not evenly. The trend is strongest among 11 states—mainly in the Northeast—and the District of Columbia that all reduced their carbon emissions by more than 15 percent during the years 2000–2014. Collectively these states have reduced their emissions by 19 percent while expanding their GDP by a robust 22 percent. These strong decouplers, except for Alaska, also tend to have the lowest per capita carbon emissions. New York, for instance, had the lowest per capita carbon emissions at 8.6 metric tons (mt) per capita in 2014, followed by Massachusetts, Connecticut, and Maryland, which emitted around 10 mt per capita (see Appendix Table B).23 In general, the states with higher per capita emissions reduction also achieved the greatest overall reductions in carbon emissions during the study period.

At the same time, about eight states have experienced a weaker form of decoupling, with emissions reductions of less than 8 percent during the 2000 to 2014 period. These states have seen average emissions reductions of about 6 percent accompanied by GDP growth of about 25 percent. California falls in this category: despite its record of adopting stringent climate action policies, California reduced its emissions by a modest 6 percent during the study period while expanding its economy by 28 percent.24 California remains something of a unique case here: the initial low presence of coal-fired power generation in the state deprived it of the leading mechanism for reducing carbon emissions: fuel switch from coal plants. Thus, labeling California a “weak” decoupler during the study period does the state something of a disservice given its past progress, history of low-carbon policy, and low carbon emissions per capita, at 9.3 mt per capita. By contrast, states like Kansas, Kentucky, and Wisconsin, which generate an average of about 17.4 mt per capita—close to the national per capita emissions level of 17.0 mt per capita—are also exhibiting “weak” decoupling.

In between the strong and weak decouplers, meanwhile, lie about 14 states such as Indiana, Ohio, Pennsylvania, and Washington that have each reduced their emissions by 8 percent to 15 percent while increasing the size of their economy by an average of 19 percent. As a group, these states have average per capita carbon emissions of 19.8 mt per capita.

Finally, the data reveal that 16 states, including Arizona, Colorado, Iowa, and Oklahoma, have not decoupled and instead experienced rising emissions (an average of 4 percent) along with rising GDP (32 percent) between 2000 and 2014. A few of these states have shown dramatic increases in emissions—for example, 26 percent in Nebraska and 16 percent in North Dakota. Not surprisingly, these 16 states also have some of the highest per capita carbon emissions in the nation. Wyoming’s carbon emissions, at 111.6 mt per capita, are the highest, followed by North Dakota’s, at 74.8 mt per capita. Decarbonization, while prevalent, is a work in progress as Donald Trump begins his presidency.

  • Market forces and changes in states’ economic structures represent a significant influence on carbon emissions and decoupling

Multiple factors are influencing the pace of decoupling across states, and some of the most important factors are market trends and the shifting nature of state economies. Driven by technology change, global value chain decisions, and evolving consumer preferences, the nation’s economy has been shifting steadily from “dirtier,” more carbon-intensive goods production to less energy-intensive high-tech goods and services provision, and this transition may be contributing to decoupling in some states.

To be sure, a statistical analysis of the impact of changing industry structure on states’ carbon emissions did not find a highly visible relationship between state industry mix and emissions.25 Nevertheless, copious research highlights the link of sector change and emissions declines. These analyses document that traditional manufacturing is relatively energy- and carbon-intensive while higher-value, high-tech manufacturing is much less so. Services are even less energy- and carbon-intensive.26

Regardless of exactly how large the role of industry structure has been in reducing emissions to date, the present analysis does surface some interesting trends. States that have seen their economies shift significantly toward service delivery, for example, have tended to see reductions in their carbon emissions. In fact, almost all of the states that experienced the largest shift toward services industries also registered large declines in their carbon emissions during 2000–2014. For example, as Maine’s service sector’s share of real GDP (in millions of chained 2009 dollars) expanded from 75 percent in 2000 to 83 percent in 2014, its carbon emissions declined by 25 percent. Similarly, Delaware, Georgia, North Carolina, and Virginia all experienced some of the largest relative expansions of their service sectors among states and likewise achieved substantial carbon emissions declines of 20 percent, 17 percent, 15 percent, and 15 percent, respectively.27

Parallel trends have followed many states’ shifts away from energy-intensive manufacturing to cleaner forms of production, such as computer chip and electronic component manufacturing.28 Nevada’s carbon emissions, for example, declined by 18 percent between 2000 and 2014 as the state’s output from energy-intensive manufacturing industries dropped from 44 percent of manufacturing output to 24 percent. Connecticut, Delaware, Maryland, New Jersey, and Oregon witnessed similar trends as they shifted from commodity manufacturing to advanced manufacturing during the 15-year period. The role played by changes in state and local industrial structure in decoupling warrants further investigation.

In any event, significant decarbonization will continue to take place regardless of policy changes in Washington thanks to state and regional economic restructuring.

  • Region and fuel mix matter a lot

Determining even more of the varied state-by-state decoupling stories are significant differences in energy sourcing across regions, these often also driven by market-ordained price changes, such as the decline of natural gas prices made possible by the “fracking” revolution.

Northeastern and many Southern states, for example, have achieved some of the most impressive feats of decoupling, and their achievement owes to mostly favorable changes in the states’ fuel mixes, such as through the substitution of natural gas for coal in power plants. (See Figure 4 and Appendix Table C.)

Carbon Emissions by State_map

Overall, Northeastern states have achieved the largest declines in carbon emissions—a collective 15 percent—even as they significantly expanded their economies by 19 percent between 2000 and 2014. Heavily influencing these trends have been changes in the region’s energy systems.

The Northeastern states, for instance, have been generating more electricity from natural gas and importing more hydroelectric power from Canada. Since 2000, the region has witnessed a dramatic shift away from petroleum and coal-fired generation to natural gas-fired output. This shift has raised concerns about overreliance on a single fuel, but it has clearly reinforced decoupling in the region.29 Natural gas accounted for 95 percent of Rhode Island’s net electricity generation in 2014, for example, and enabled significant delinking of growth and emissions. In Connecticut the share of net generation from natural gas jumped from 12 percent in 2000 to 44 percent in 2014. In Massachusetts it jumped from 28 percent to 60 percent.

New England states have also maintained or expanded the capacity of multiple zero-emissions nuclear plants, although the Vermont Yankee plant was taken out of service late in 2014.30 Nuclear’s share of net electricity generation went up from 23 percent in 2000 to 31 percent in 2014 in New York and from 14 percent to 19 percent in Massachusetts. Nuclear plants accounted for no less than 35 percent of Northeastern states’ electricity generation in 2014—the highest share among all regions. These choices have allowed these states to substantially reduce the use of coal and oil as generation fuels even while expanding their economies. In addition, the New England states along with New York and Delaware are members of the nation’s first regional cap-and-trade program, the Regional Greenhouse Gas Initiative, which has capped carbon dioxide emissions from the power sector. It is not surprising, therefore, that a number of these states have led the national decoupling trend while driving their per capita CO2 emissions to some of the lowest levels among states.

In similar fashion, power-sourcing choices across the South have allowed multiple states there to sever the growth/emissions link. While not as impressive as the Northeastern states, the Southern states reduced their carbon emissions by 7 percent while growing their economies by 30 percent.

Most notably, significant coal-based power generation has been replaced by cleaner natural gas plants, as the share of net electricity generation from natural gas in the South increased from less than 20 percent in 2000 to 35 percent in 2014. Alabama, Delaware, Florida, Georgia, and Virginia have all shifted significant power generation to natural gas.

Southern states reduced their carbon emissions by 7 percent while growing their economies by 30 percent

In addition to natural gas, nuclear plants, which operate in 12 of the 17 Southern states, accounted for 19 percent of net electricity generation in the South in 2014. Maryland’s sole nuclear power plant supplies 38 percent of the state’s net electricity generation—increasing its share from 27 percent in 2000—and the state has curbed its emissions by 20 percent while expanding its economy by 33 percent. Tennessee has reduced its emissions by 19 percent while expanding its economy by 23 percent in part because its two nuclear facilities account for more than one-third of the state’s net electricity generation. In fact, nuclear power increased its share of Tennessee generation from 27 percent to 35 percent between 2000 and 2014. In short, the combination of nuclear plants and coal’s replacement by natural gas has enabled many Southern states to pull off some of the most dramatic decouplings in the country. Like Georgia, North Carolina has supported its double-digit growth and emissions reductions in large part by sourcing nearly 32 percent of its electricity from nuclear and 22 percent from natural gas. Coal-based generation has fallen to just 38 percent.31 That four new nuclear reactors are under construction in Georgia and South Carolina (with another due to begin commercial operation soon in Tennessee) suggest that multiple Southern states will continue their recent progress and have zero-carbon power at their disposal as they continue to grow.

Decoupling trends in the Midwest and West, by contrast, reflect a far less optimal set of fuel sourcing trends. Both Midwestern and Western states managed carbon reductions of just 5 percent each, while economic growth came in at 14 percent and 30 percent, respectively. Behind these trends lie distinctive energy-sourcing patterns. Most notably, Midwestern states have been able to switch far fewer coal-burning power plants over to natural gas than states in the Northeast and South. In 2014, for example, Midwestern states sourced just 7 percent of their electricity from natural gas generation, compared to 35 percent each for Northeastern and Southern states. That leaves the Midwest far more dependent on coal for generating power—61 percent in 2014 compared to 17 percent in the Northeast, 38 percent in the South, and 27 percent in the West. For example, in 2014 Iowa, Missouri, and Nebraska depended on carbon-intensive coal for 60 percent, 82 percent, and 63 percent of their electricity production, respectively. For its part, while the West burns more natural gas (enough to account for 30 percent of electricity generation) than the Midwest, it runs far fewer nuclear plants. At present only Arizona, California, and Washington possess nuclear plants in the West, and the share of net electricity generated by nuclear declined from 10.5 percent in 2000 to 7.7 percent in 2014. (This drop may reflect the closure of the San Onofre plant in California in 2013).

A note about renewables is in order, meanwhile: wind and solar generation have yet to register as broad an impact on decoupling as might be expected—even in the green West. In this regard, while solar and wind’s share of electricity generation has been on the rise, its large-scale growth in some states dates only to the last decade, and so this analysis does not find a strong statistical relationship between states’ emissions reductions and solar and wind’s share of power generation.32 With that said, however, wind’s share of electricity generation grew from less than 2 percent in 2008 to 19 percent in 2014 in Idaho, and similar trends can be seen in Kansas, North Dakota, South Dakota, Texas, and Wyoming. Texas leads the nation in wind-powered electricity generation, producing more than one-fifth of the U.S. total in 2014 and supplying 9.3 percent of its own power needs from that source.33 Similarly, solar generation has increased in a few states. California, for instance, became the first state in 2014 to get five percent of its electricity generation from solar. In 2014 solar contributed less than 3 percent to Arizona’s net electricity generation but had increased almost 50 percent from the previous year. However, in most of these states carbon emissions rose between 2001 and 2014, as well as in the period since 2008 when wind and solar development has taken place in the country. The positive impact of solar and wind energy deployment is probably being undercut by other factors. But it is likely that renewables will soon contribute to decoupling and decarbonization.

In sum, it is impossible to overestimate the importance of decisions, both market-ordained and otherwise, about electricity sourcing in states’ decarbonization. Overall, the Energy Information Administration concludes that changes in the national mix of electricity production—especially the shift toward cleaner-burning natural gas—accounted for more than two-thirds of the country’s and states’ emissions reductions between 2005 and 2015.34 And that link is extremely visible here. On the one hand, the rapid switch from coal to natural gas in scores of power plants has been a principal driver of state-level decoupling and emissions reductions.35  Altogether, coal’s share of state electricity generation declined in no less than 43 states and contributed heavily to emissions reductions in most states. (See Figure 5).36 On the other hand, nuclear power has helped a number of Northeastern and Southern states limit their carbon emissions since 2000.37 That Massachusetts, New York, Vermont, Maryland, Tennessee, and Michigan all increased the share of their electricity drawn from nuclear plants allowed these states to power growth with plentiful zero-carbon energy and so decouple strongly. At a time when more plants are being shut down than opened, however, it is important that Maryland, Tennessee, and New York are all increasing the nuclear share of their total generation. With that said, many more reactors will likely need to come on line in the next 30 years to enable the next round of deeper decarbonization.38

Going forward, policy uncertainty is going to be a variable in determining whether market dynamics and state self-determination can maintain adequate progress on decarbonization, particularly through stable renewables markets and nuclear expansion. With that said, significant progress has taken place during both supportive and challenging federal administrations. And to the extent that natural gas prices remain low, significant decarbonization will continue at least in the next few years by dint of the market-based substitution of coal by gas in power plants.

Metro_20161206_decoupling scatter plot_fig5

  • Despite significant progress, all states need to do more to decouple emissions from growth and decarbonize their economies.

Many states, in short, have made impressive progress in reducing carbon emissions and decarbonizing their economies. These gains have been achieved during both sympathetic and skeptical administrations, and much of that progress has been aided by the gas boom. In fact, in 2016 natural gas became the number one fuel source for electricity generation.39 However, much more needs to be done for the United States to come anywhere near to meeting the Paris goal of long-term decarbonization, which will remain an important benchmark whether or not Donald Trump withdraws from the global agreement.

To understand the challenge, it is helpful to consider the needed pace of decarbonization that will be required to prevent warming in excess of 2 degrees Celsius. Analysis from PricewaterhouseCoopers calculates that the global economy will need to cut its carbon intensity (meaning its emissions of CO2 per dollar of GDP) by a rapid 6.3 percent every year from now to 2100 to achieve that outcome.40 For its part, the United States will need to decarbonize its economy by 4.3 percent a year from now till 2030.41 What does that say about the current pace of decarbonization in the United States? It says that the nation and its states—notwithstanding recent progress made possible by the current windfall of the switch from coal to gas in many states’ power sectors—are falling far short of the goal.

At the national level, for example, the United States decarbonized its economy at a rate of 2.1 percent a year between 2000 and 2014—a pace just at half of the needed pace.

As to the states, some of them have reduced their carbon intensity faster than others and could be better positioned than others to achieve the needed 4.3 percent a year benchmark. (See Figure 6.) The decarbonization of state economies has been driven by two forces: changes in the energy intensity of the economy and changes in the carbon intensity of the energy supply.42 In keeping with those trends, at least 27 states and the District of Columbia have been decarbonizing their economies at rates above the national rate and sometimes exceeding the needed rate of 4.3 percent. At 5.1 percent and 4.4 percent, for example, North Dakota and the District of Columbia surpassed the needed 4.3 percent mark. Alaska, Maryland, Massachusetts, New York, and Oregon accomplished more than 3.0 percent average annual reductions in carbon intensity between 2000 and 2014.

The District of Columbia, for example, a city, has been able to lead the nation in decarbonization thanks to a combination of density, progressive energy policies, and fast growth across its relatively low-carbon service industries.43 North Dakota’s decarbonization is in some ways deceptive and cautionary, as it has been driven more by large oil and gas shipments from its fracking rigs than by a modest shift in its electricity mix from coal to wind energy. Maryland has achieved its rapid rate of decarbonization thanks to significant change in its energy intensity. The state’s economy is not energy intensive: service industries contribute two-thirds to state GDP while manufacturing, including the manufacture of chemicals and electronics, contributes less than 6 percent.44 In a similar fashion, Massachusetts and New York’s economies rely on less energy-intensive industries, such as financial services, information technology, health care, and professional, technical, and scientific services. Meanwhile, Missouri (0.4 percent a year decarbonization pace), Mississippi (0.7 percent), Nebraska (0.7 percent), and Illinois (0.7 percent) have barely decarbonized their economies at all. In short, all states will need to do significantly more to decarbonize their economies so that they can contribute to the national decarbonization target of 4.3 percent per year going forward.

metro_20161216_50 states_bargraph_fig6

It does not bode well that decarbonization has proceeded only half as fast as needed during a period encouraged by the pro-decarbonization Obama administration and empowered by a glut of cheap natural gas that allowed plentiful switches from coal.


Overall, the data on decoupling across U.S. states highlight some troubling dynamics but also an encouraging fact: that places can sever the historical link between economic growth and carbon emissions, and that local factors (not just federal ones) matter a lot to how this happens. Moreover, the trends depicted here suggest that while federal policy reversals could be traumatic, progress on decarbonizing the nation’s economy will likely continue regardless of Donald Trump, driven by technology advances, market dynamics, and state policy.

Those three trends—technological progress, market dynamics, and state policy—warrant confidence that the decarbonization of the U.S. economy will continue.

And yet, such confidence must be tempered given the coming sea change in federal energy and environmental policy.

Neither the world, the nation, nor the states are making enough progress at decarbonizing the world economy, even acknowledging the genuine progress of recent years. And now, the arrival of Trump in the White House promises vast uncertainty and likely setbacks.

States and cities, in short, are going to need to do much more.  Which raises the question of policy. With responsibility for U.S. decarbonization now devolving to the states and cities, state and regional actors must fill the vacuum created by Washington’s abdication of leadership with new energy and resolve.

Policy choices at the state level are going to matter even more than in the past. Smart oversight of the utility sector, for example, to ensure the continued rapid switch from coal to gas in power plants should remain a top priority. Likewise, renewable portfolio standards—which have played an important role in increasing the production of energy from renewable sources in many states—must continue to be maintained.45 And then, states can double down on their recent experimentation in the area of green finance, in which some states have developed innovative uses of clean energy funds, green banks, and green bonds.46

Progress on decarbonizing the nation’s economy will likely continue regardless of Donald Trump, driven by technology advances, market dynamics, and state policy

Looking forward, the standards should be strengthened. States can regulate emissions from power plants within their own borders, push for deeper investments in energy efficiency, spur the development of new technology, and find other ways of cutting carbon emissions.47 Moreover, the important role of nuclear power in decarbonization underscores the importance of state-level regulatory support. More states need to expressly include nuclear energy in their renewable energy plans, and more states should consider, as New York has done recently, ways to subsidize and otherwise support struggling nuclear plants to keep them running.48 Enlightened renewables development will also be crucial.

At the same time, states will need to push back against the most disruptive policy assaults from Washington and demand a minimum viable platform of basic policy supports.

What might that platform look like? Investment in clean energy research and development may provide one element of it, since clean energy R&D has long enjoyed broad support in both parties. Likewise, Congress should leave in place key wind and solar tax breaks, which were extended in a bipartisan deal through 2020 and 2021 and enjoy pockets of GOP support. There also exists the opportunity to leverage Trump’s promise to invest over $500 billion in infrastructure to modernize and expand the electrical grid to eliminate impediments to the long-term growth of clean energy resources.

Beyond that, Trump-circle statements suggest the president-elect may support actions to expand the nation’s nuclear capacity, which will be critical to developing a clean, modern power system. To that end, state leaders should impress upon the new administration the urgent need to take steps—in partnership with state and local regulatory authorities—to use production payments, loans, efficient licensing, and other supports to ensure the continued operation of existing plants and the possibility of new plants coming on line.49 Research into safer, cheaper, more compact reactor designs would seem a viable priority for the new administration.

In any event, states are going to need to lead in the next years. Whether they can be successful or not, their goal must be a bottom-up brand of American decarbonization strong enough to mitigate the worst costs of federal abdication, if that is what occurs.


Ultimately, the reality of emissions decoupling across nations and states confirms that the transition to a modern energy system can occur without sacrificing growth, contrary to the false notion in command in Washington that these are opposing forces. In fact, the decoupling of economic growth from carbon emissions in so many states demonstrates that states and cities alike now have the opportunity—with or without federal partnership—to craft a new sort of growth that at once widens the circle of prosperity and achieves environmental sustainability.

The challenge now is to spread the decoupling of growth from emissions to all states, increase the pace of decarbonization, and maintain it for decades to come with or without the help of Washington.

That the main responsibility for this urgent project has fallen to America’s states and regions may seem discouraging, but it is nothing new—and it does not rule out success.


  1. WRI found that 21 countries, including the United States, have fully “decoupled” their economic growth from carbon emissions over the last 15 years. Carbon Brief, which extended the data to include all 216 countries, not just the 67 used in the WRI study, found that 35 countries increased their real GDP at the same time they cut their carbon emissions. For more information, see Nate Aden, “The Roads to Decoupling: 21 Countries Are Reducing Carbon Emissions While Growing GDP” (Washington: World Resources Institute, 2016); and Sophie Yeo and Simon Evans, “The 35 Countries Cutting the Link Between Economic Growth and Emissions” (London: Carbon Brief, 2016).
  2. See International Energy Agency (IEA), “Global Energy-Related Emissions of Carbon Dioxide Stalled in 2014,” March 13, 2015 and IEA, “Decoupling of Global Emissions and Economic Growth Confirmed,” March 16, 2016.
  3. The 2-degree Celsius warming target has been a point of scientific consensus for years and gained substantial backing among nations this year at the COP 21 climate change conference. For more details about COP 21, see United Nations Framework Convention on Climate Change, “Paris Agreement,” The United States and China, the two top emitters of greenhouse gases, formally ratified the Paris Agreement on September 3, 2016; see Jean Chemnick, “U.S. and China Formally Commit to Paris Climate Accord,” Scientific American, September 6, 2016.
  4. David Roberts and Brad Plumer, “Most People Are Wildly Underestimating What Trump’s Win Will Mean for the Environment.” Vox, November, 14, 2016.
  5. Barry Rabe, “Statehouse and Greenhouse: The States Are Taking the Lead on Climate Change” (Washington: Brookings Institution, 2002).
  6. Mark Muro, “Climate, Energy, Trump: Progress Is Still Possible,” The Avenue, Brookings Institution, Nov. 15, 2016. See also Don Grant et al., “Effectiveness of State Policies in Reducing CO2 Emissions From Power Plants,” Nature Climate Change 4 (2014): 977-82.
  7. Energy-related carbon dioxide emissions data at the state level are available at the EPA website The data include state CO2 emission inventories from fossil fuel combustion, by end-use sector (commercial, industrial, residential, transportation, and electric power), in million metric tons of carbon dioxide (MMTCO2) from 1990 through 2014. EPA developed these state-level CO2 estimates using (1) fuel consumption data from the Department of Energy (DOE)/Energy Information Administration (EIA) State Energy Data 2014 Consumption tables and (2) emission factors from the Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2014. State GDP has been obtained from the Bureau of Economic Analysis and is in millions of chained 2009 dollars. See
  8.  As an exploratory analysis, Brookings researchers employed ordinary least squares (OLS) and fixed effects (by state) estimation techniques to evaluate the impact of a number of factors on state change in carbon emissions between 2000 and 2014. The dependent variable in all the models is the percent change of CO2 emissions by state from year to year. The independent variables employed include: economic structure variables relating to the percent change in share of private industries output of goods-producing sectors, service-producing sectors, manufacturing sectors, and energy-intensive manufacturing sectors; the percent change of fuel sources as a share of total electricity generation with sources of coal, natural gas, nuclear, hydro, wind and solar combined; the percent change in state GDP from year to year; region of the state; the percent change in population density from year to year; the percent change in population from year to year; and a series of year dummy variables for 2001 to 2014 (with 2001 as the reference group). With 50 states and the District of Columbia, there were 714 observations across all the models. A total of six models was run and the R-squared was between 0.415 and 0.421 across the models, indicating a relatively strong model fit.
  9. While the literature discussing the energy intensities of industries in the services sector, such as information, finance, and health care, is limited, it is generally believed that these industries emit comparatively less carbon than the manufacturing sector. See Brynhildur Davidsdottir and M. Fisher, “The Odd Couple: The Relationship Between State Economic Performance and Carbon Emissions Economic Intensity,” Energy Policy 39, no. 8 (2011): 4551-62; B.W Ang., “Is the Energy Intensity a Less Useful Indicator Than the Carbon Factor in the Study of Climate Change?” Energy Policy 27, no. 15 (1999): 943-46.
  10. Renewable portfolio standards (RPS), for instance, have been found to have a negative and significant impact on state carbon intensities through their influence on state electricity prices. Additionally, the adoption of RPS was found to reduce overall U.S. carbon emissions by 4 percent by 2010. See Samantha Sekar and Brent Sohngen, “The Effects of Renewable Portfolio Standards on Carbon Intensity in the United States” (Washington: Resources for the Future, 2014). Another paper found significant and robust decreases in carbon emissions associated with the introduction of public benefit funds, a form of carbon tax, adopted by various states. See Monica Prasad and Steven Munch, “State-Level Renewable Electricity Policies and Reduction in Carbon Emissions,” Energy Policy 45 (2012): 237-42; and Grant et al. (2014).
  11. International Panel on Climate Change (IPCC), “Climate Change 2014: Mitigation of Climate Change” (Geneva: IPCC, 2014).
  12. Terence Jeffrey, “U.S. Has Record 10th Straight Year Without 3% Growth in GDP,” CBS News, February 26, 2016.
  13. The issue of whether GDP growth can be delinked from carbon emissions is usually framed in terms of the Carbon Kuznets Curve (CKC). The CKC hypothesis holds that carbon emissions initially increase with increasing growth (due to industrialization), but later peak and decline after a threshold level of per capita GDP as countries become more energy efficient, more technologically sophisticated, and more inclined to reduce emissions by corresponding legislation. The empirical work to test the CKC hypothesis has produced varying results. Some studies have found that CO2 and other greenhouse gases do not decline after a country reaches a higher stage of economic development; rather CO2 emissions increase monotonically with income. Other studies have found that CO2 emissions continue to increase only for developing countries. For a good discussion of the literature, see Kris Aaron Beck and Pratibha Joshi, “An Analysis of the Environmental Kuznets Curve for Carbon Dioxide Emissions: Evidence From OECD and Non-OECD Countries,” European Journal of Sustainable Development 4, no. 3 (2015): 33-45.
  14. Aden (2016).
  15. Yeo and Evans (2016).
  16. Looking at GDP and emissions data from 1990 to 2015.
  17. David Victor et al., “Introductory Chapter,” in Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge, England: Cambridge University Press, 2014).
  18. Data on state industry structure have been obtained from Bureau of Economic Analysis, “State Output by Industry,” Data on state fuel mix are from EIA, “Net Generation by State by Type of Producer by Energy Source (EIA-906, EIA-920, and EIA-923), 1990-2014,”
  19. The decline over the years can be mainly attributed to changes in the electric power sector—arising from decreased use of coal and increased use of natural gas for electricity generation. See EIA, “U.S. Energy-Related Carbon Dioxide Emissions in 2015 Are 12% Below Their 2005 Levels,” May 9, 2016.
  20. Michigan is an outlier, having reduced its carbon emissions by 16 percent while also experiencing a negative GDP growth of 2 percent.
  21. The District of Columbia reduced its carbon emissions by 30 percent while growing its economy by 32 percent during the same period.
  22. During the 2008-2014 period, for instance, Connecticut, Maine, and Nevada decreased their carbon emissions by 7 percent, 13 percent, and 9 percent, respectively, but their GDP declined significantly following the Great Recession, by 4 percent, 2 percent, and 7 percent, respectively. Feng et al. found that 83 percent of decrease in carbon emissions between 2007 and 2009 can be attributed to economic decline and the remaining 17 percent to changing energy sources; see Kuishuang Feng et al., “Drivers of the US CO2 Emissions 1997-2013,” Nature Communications 6 (2015).
  23. Brookings analysis of EPA’s “CO2 Emissions From Fossil Fuel Combustion—Million Metric Tons CO2 (MMTCO2)” and state population data for 2014.
  24. California is an interesting example for a variety of reasons. The state has some of the strictest policies in place to curb greenhouse gas emissions. It also has one of the lowest carbon intensities as well as per capita carbon emissions among all states. And yet, compared to other climate-friendly states like Massachusetts and New York, it experienced a slower reduction in carbon emissions between 2000 and 2014. In fact, the state increased its carbon emissions by nearly 6 percent between 2000 and 2007. Carbon emissions dropped by 7 percent from 2008 to 2014.
  25. In all regressions, variables measuring the economic structure of a state were found to be statistically insignificant. It possible that the impact of this variable is weaker, slower moving, or more diffused than other factors. More data and better model specification in future research will likely uncover the impact of states’ economic structure on change in carbon emissions over time. Four variables were used to measure state economic structure: percent change in the share of goods-producing sectors’ output to all private industry output, percent change in the share of services-producing sectors’ output to all private industry output, percent change in the share of manufacturing sectors’ output to all private industry output, and percent change in the share of energy-intensive manufacturing sectors’ output to all private industry output.
  26. See, about manufacturing, Lee Schipper et al., “Carbon Emissions From Manufacturing Energy Use in 13 IEA Countries: Long Term Trends Through 1995,” Energy Policy 29, no. 9 (2001): 667-88; and Oak Ridge National Laboratory, “U.S. Manufacturing Energy Use and Greenhouse Gas Emissions Analysis” (Oak Ridge, Tenn.: 2012). For discussion that includes the energy and emissions performance of services industries see Davidsdottir and Fisher (2011) and Ang (1999).
  27. Georgia’s service sector’s share of real GDP grew from 75 percent in 2000 to 82 percent in 2014, North Carolina’s from 63 percent to 72 percent, and Virginia’s from 76 percent to 83 percent.
  28. EIA, “Chapter 7. Industrial Sector Energy Consumption,” in International Energy Outlook 2016 (Washington: EIA, 2016). Energy-intensive industries emit large quantities of CO2, related to both their energy consumption (combustion emissions) and their production processes (process emissions). For purposes of this analysis, Brookings researchers have placed in the energy-intensive category food manufacturing (NAICS 311); paper manufacturing (NAICS 322); petroleum and coal products manufacturing (NAICS 324); chemical manufacturing (NAICS 325); primary metal manufacturing (NAICS 331), which includes the iron and steel and alumina and aluminum industries; and non-metallic mineral product manufacturing (NAICS 327), which includes the cement and glass industries.
  29. EIA, “Northeast Grows Increasingly Reliant on Natural Gas for Power Generation,” November 12, 2013. As increasing amounts of natural gas are being used for electricity generation in the New England states, assurance of natural gas supply has become a critical energy issue for the region.
  30. Vermont is an exception. At the end of 2014, Vermont shut down the Vermont Yankee Nuclear Plant and with that the state lost 55 percent of its electricity generating capacity and the source of more than 70 percent of its net generation in recent years. See Vermont’s energy profile,
  31. For more details, see North Carolina’s energy profile,
  32. Both OLS and fixed-effects (by state) estimation techniques were used with cluster-robust standard errors (states being the clusters) to estimate the impact of solar and wind combined on carbon emission change across states. The dependent variable is percent change of CO2 emissions by state from year to year. In all of these regressions, the coefficients on the percent change of wind and solar sources as a share of total electricity generation were statistically insignificant, with p-values all below 0.10. The magnitudes were all negative, which conforms with expectations and suggests that further study may possibly show statistically significant decreases in CO2 emissions due to larger shares of solar and wind electricity generation.
  33. For more details, see Texas’ energy profile,
  34. EIA, “U.S. Energy-Related Carbon Dioxide Emissions in 2015 Are 12% Below Their 2005 Levels,” May 9, 2016.
  35. Coal still represents 34 percent of electricity generation in the United States, but its use is in decline. Several factors have contributed to the gradual decline in coal consumption, including the availability and price of other electricity-generating fuels like natural gas and cleaner energy generation options like wind and solar. While coal’s share of electricity generation has fallen from 54 percent in 1990 to 34 percent in 2015, the natural gas share of electricity generation has grown tremendously, from 11 percent in 1990 to 32 percent in 2015. At the same time, renewables other than hydropower have been growing their share of electricity generation in recent years. Wind and solar’s share of generation has increased from a mere 1 percent in 2005 to 6 percent in 2015. Hydropower contributed 6 percent in 2015 as well. See Brookings analysis of EIA’s “Share of Three Fossil Fuels and of Non-Fossil Fuel Generation 1990-2014,” The 2015 data are available at
  36. In all regressions, the coefficients on the percent changes of coal as a share of total electricity generation were statistically significant. For each 1 percentage point increase in the share of total generation by coal, CO2 emissions increased by 0.044 to 0.045 of a percentage point.
  37. In all regressions, the coefficients on the percent changes of nuclear as a share of total electricity generation were statistically significant. For each 1 percentage point increase in the share of total generation by nuclear, CO2 emissions decreased by 0.094 to 0.097 of a percentage point.
  38. Peter Behr, “Nuclear Closures Magnify U.S. Climate Challenge for Trump,” E&E News, November 17, 2016.
  39. EIA, “Natural Gas Expected to Surpass Coal in Mix of Fuel Used for U.S. Power Generation in 2016,” March 16, 2016,
  40. PricewaterhouseCoopers (PwC), “Conscious Coupling? Low Carbon Economy Index 2015” (London: PwC, 2015).
  41. Ibid.
  42. The energy intensity of a state is measured by the amount of energy consumed per unit of economic output. The carbon intensity of energy supply is reflective of the energy fuel mix within a state. The product of these measures gives the carbon intensity of the economy.
  43. The District of Columbia has a very different carbon emissions breakdown compared to other cities. Buildings—including the administrative hubs for massive federal agencies, foreign embassies, lobbying groups, and think tanks—generate 75 percent of all greenhouse gas emissions in the city. DC, therefore, has one of the most progressive green building policy and implementation frameworks in the nation. It was the first U.S. city to pass a green building law. It is the leading metro area for number of Energy Star certified buildings. For more information see
  44. Brookings analysis of U.S. Bureau of Economic Analysis, Interactive Data, GDP and Personal Income, Regional Data, Annual Gross Domestic Product (GDP) by State, GDP in current dollars, All industries, Maryland, 2013.
  45. Galen Barbose, “Renewable Portfolio Standards in the United States: A Status Update” (Berkeley, Calif.: Lawrence Berkeley National Laboratory, 2013).
  46. See, for example, Lewis Milford et al., “Leveraging State Clean Energy Funds for Economic Development” (Washington: Brookings Institution, 2012); Ken Berlin et al., “State Clean Energy Finance Banks: New Investment Facilities for Clean Energy Deployment” (Washington: Brookings Institution, 2012); and Lew Milford et al., “Clean Energy Finance Through the Bond Market: A New Option for Progress” (Washington: Brookings Institution, 2014).
  47. Northeastern states have created their own cap-and-trade program called Regional Greenhouse Gas Initiative, California has its own cap-and-trade program along with one of the nation’s most ambitious climate goals—to reduce its greenhouse gas emissions to 40 percent below 1990 levels by 2030. And Hawaii plans to go 100 percent renewable by 2045.
  48. See, for example, Richard Martin, “New York State Has a Plan to Rescue Nuclear Power,” MIT Technology Review, August 2, 2016, and Vivian Yee, “Nuclear Subsidies Are Part of New York’s Clean-Energy Plan,” New York Times, July 20, 2016.
  49. Secretary of Energy Advisory Board (EAB), “Task Force on the Future of Nuclear Power: Draft Report” (Washington: EAB, 2016).