A drop in U.S. carbon dioxide pollution in recent years stems from burning natural gas instead of coal. Or does it? Given that the U.S. bid to combat climate change through actions like the Clean Power Plan relies on more burning of gas than coal in power plants, that answer is both politically and scientifically important.

Compared with coal, burning natural gas results in roughly half the amount of CO2 per megawatt-hour of electricity. Yet even half the CO2, when spread over hundreds of power plants,is too much to achieve such goals as a CO2-emission reduction of 80 percent by 2050 or 100 percent by the end of this century, in order to avoid more than 2 degree Celsius of global warming, more acidic oceans, inexorable sea level rise and extreme weather, among other unpleasant impacts predicted by scientists. Under the terms of the Clean Power Plan, the most advanced natural gas burning power plants can still emit 771 pounds of CO2 per megawatt-hour of electricity produced. So is natural gas a bridge to a cleaner energy future or a slightly longer route to climate catastrophe?

To answer that question the past may provide a rough guide. At roughly the same time after the turn of the 21st century the U.S. underwent a recession, an energy transition to more natural gas and a move away from producing highly polluting products such as steel. So which of these factors deserves the most credit for the accompanying drop in the nation’s global warming pollution? It's an important puzzle to solve because recessions are not widely viewed as a policy option (advocates of “degrowth” notwithstanding) whereas the export of fracking—the process by which natural gas is extracted from shale rock—to other countries to help deliver a cleaner fossil fuel habit is.

To disentangle all these competing explanations or at least find out their relative importance, scientists have used a mathematical technique known as structural decomposition analysis. Here's how it works: There is a big number to look at, say, total consumption in an economy. This big number can be broken down into contributing factors, such as population size and consumption per person. Hold population steady through time as total consumption changes and you derive the change in total consumption caused by a change in consumption per person. Then hold consumption per person steady while changing population size and you derive how much of total consumption comes from each of these factors.

A group of scientists and economists used such a mathematical analysis to look at U.S. CO2 emissions between 1997 and 2013, a period that saw total pollution drop by nearly 800 million metric tons, or roughly the annual pollution of Germany. The group looked at six different factors: population and consumption per person, but in addition shifts in consumption patterns; shifts in industry; the energy intensity of the economy; and changes in the fuel mix.

The study found that prior to 2007 (and, hence, the start of the recession), U.S. CO2 pollution continued to grow, largely because the economy continued to expand; people bought more and more. After 2007 U.S. CO2 pollution dropped but roughly 80 percent of that decline was because people and companies bought and built less stuff, supplemented by the shift away from heavy industry. This finding matches a slew of previous analyses by the U.S. Energy Information Administration, Pacific Northwest National Laboratory and CO2 Scorecard, among others, that have also concluded that the 2008 great recession was largely responsible for the observed emission reductions. In fact, one warm winter in 2012 alone played an outsize role in recent CO2 reductions.

Natural gas did play a significant supporting role in reducing pollution from the energy sector, however—a role that has increased over time as people and companies have started buying more stuff. The shift away from burning coal has counterbalanced population growth, according to this new analysis. In fact, cheap and abundant natural gas appears to have helped keep some 160 new coal-fired power plants from being built, which would have spewed hundreds of millions of metric tons of CO2 over the years.

Coal's share of electricity generation in the U.S. has been dropping since 2009 and more than 180 gigawatts of power plants that burn natural gas have been built since 1990. The electricity from a one-gigawatt coal-fired power plant can be replaced by burning one billion cubic meters of natural gas instead, resulting in an annual savings of roughly three million metric tons of CO2 in addition to reductions in other air pollution, like the sulfur dioxide that causes acid rain or the nitrogen oxides that create smog. There is now 1.5 times more potential electricity generation from burning natural gas than from burning coal in the U.S., and coal-fired power plants representing roughly 7 percent of this country's electricity generation are retiring this year, mostly in the eastern half of the country. An analysis by the National Renewable Energy Laboratory suggests that natural gas and renewables like wind turbines and solar panels have picked up the slack produced by missing coal—the beginnings it seems of a long-term energy transition. More simply put, natural gas may be keeping a lid on growth in CO2 emissions from generating electricity in the U.S. at present.

But natural gas hasn't just killed coal. From Florida to Wisconsin, gas-fired power plants are replacing nuclear ones. That fuel switch actually increases CO2 pollution, however. And, in the absence of mandates like renewable portfolio standards—mandates for a certain percentage of electricity to derive from renewable resources—natural gas could also prevent the building of wind and solar farms or geothermal power plants.

Furthermore, all those power plants that burn natural gas will still spew CO2, albeit less than the equivalent coal-fired power plant. In a world aimed at zero emissions, that reduction is not good enough ultimately. In fact, the more than 1,000 gigawatts of natural gas–fired power plants built around the world would spew roughly 300 billion metric tons of CO2 if operated over the next 50 years—or more than half of the world's remaining carbon budget. Exceeding that budget may lock in the worst of climate change, whether fast sea level rise or extreme weather. Cheap natural gas may even slow the shift away from heavy industry in the U.S.: New fertilizer plants and chemical plants have already been built as a result of cheap and abundant natural gas and new steel plants may not be far behind. Finally, natural gas can leak, adding methane to the atmosphere, which also exacerbates global warming.

In the context of an energy transition that may take decades the U.S. does not have 20 years for natural gas to kill coal and replace oil in power and transportation, respectively, followed by another 50 years needed to replace now-entrenched natural gas with renewables and/or nuclear power plants powering electric cars and trucks. The U.S. Environmental Protection Agency expects natural gas to be producing one-third of U.S. electricity in 2030 and technologies that might make natural gas near-zero carbon, like those that capture and store CO2, have yet to be tried or even tested on the gaseous fossil fuel.

For all these reasons, natural gas makes for a weak bridge to a zero-pollution future and truly clean power—one that cannot span more than a few decades. Still, a bridge made of gas is better than none at all.