The Trump administration is currently facing a major decision—whether to withdraw the United States from the Paris Accords on climate change. The huge multi-national agreement was finalized in the closing weeks of the Obama administration, just days before Trump’s surprise victory in the presidential election. The key commitment made by the United States under the accords is to reduce carbon dioxide emissions in the next decade by about a quarter of their 2005 rate, with further reductions to come thereafter. But during his campaign, Donald Trump promised to pull out of the accords, and, at the recent meeting of the G-7, was the lone holdout against a ringing endorsement of the agreement. German Chancellor Angela Merkel has been insisting that the United States stay the course, but it appears as if Trump is inclined to honor his campaign promise to pull out of the accords, a position in line with that of Scott Pruitt, the new head of the Environmental Protection Agency.

The President’s instincts are spot on here. He should withdraw the United States from the accords and be prepared to stoutly defend his decision on both political and scientific grounds. Ironically, the best reasons for getting out of the accords are the evident weaknesses in the reasons that a wide range of businesses and environmental groups offer for staying in.

One constant refrain of both large American corporations and environmental groups is that by withdrawing from the Paris Accords, the United States will suffer a “huge missed opportunity” to work on the cutting-edge technologies of wind and solar energy. But why? At this point, solar and wind energy, as the indefatigable Matt Ridley points out, amount to at most a trivial portion of the global energy supply, less than one percent in total. Indeed, most of that production comes from state-subsidized ventures that could never survive on their own. And while firms race to collect government subsidies to develop so-called cleaner energy, none of their research is likely to solve the intractable problem of how to store wind or solar energy efficiently.

Further, to label wind and solar as “green” energy simply ignores the substantial environmental costs associated over their life-cycle of development, fabrication, installation, and maintenance. Covering the ground with huge solar panels is a form of thermal pollution; wind turbines emit a low hum injurious to people and are notorious for killing birds; and mining the materials required for the manufacture of each form of energy results in more environmental harm.

On the other hand, it is a given that coal, oil, and natural gas will remain central pillars of the world’s energy supply for the indefinite future, given their energy richness and operational reliability. Research that reduces harmful emissions from these widely used sources has a far higher rate of social return than any improvements in wind and solar, which are large enterprises that require a huge number of workers to generate a tiny amount of energy: 374,000 people work in solar and 100,000 in wind, compared to 160,000 for coal and 398,000 for natural gas. These “green” energy sources are clearly inefficient, which is why so much labor is wasted on so little output.

On the other hand, there have been major technological advances in the production of natural gas, oil, and coal (the cost of fracking dropped 30 percent in less than two years), and these sources remain as vital as ever to global energy production. It is mere propaganda to insist that “The End of Coal is Near” now that China has decided to scrap the construction of 103 new power plants. China’s five-year plan still calls for a nearly 20 percent expansion of its coal production capacity by 2020. India and Japan are following suit. Germany, pace Merkel, produces huge quantities of coal annually, virtually all of it lignite, or brown coal, the dirtiest variety. Coal from these sources not only produce all the traditional forms of pollution, but also causes ailments ranging from asthma to radiation diseases (from mining rare earths)—making the issue of carbon dioxide emissions an afterthought. Improving the efficiency of coal production will thus yield far greater returns than fine-tuning the production of wind and solar energy.

A further reason to get out of the Paris Accords is to make sure that the United States does not get dragooned into researching areas that only promise, at best, low rates of return. If wind and solar were worth their salt, private capital would flow into their research and development, just as it does for fossil fuels. The accord’s supporters have the basic economics exactly backwards, for there is no reason to think that superior American technology will be spurned by countries that can use it to increase energy yields while reducing pollution.

The defenders of the Paris Accords are as dogmatic on the science as they are on the economics. To them, it is an axiomatic truth that carbon dioxide emissions pose a grave threat to the environment, even though the putative causal chain is filled with missing links. The current practice is to assume that every adverse climate event is somehow the result of the rather smallish increases in carbon dioxide levels over the past 65 years. In order to reach that result, however, it is necessary to exclude other explanations for the adverse events. An observed rise in sea level in Florida, for example, is more likely attributable to the draining of local aquifers than to increases in global temperature. Indeed, sea level rises have, if anything, slowed down in recent years, notwithstanding increases in carbon dioxide levels. In a similar vein, the rapid melting of ice on the western part of the Antarctic is more likely attributable to underground volcanic activity, particularly given that the overall ice levels in the Antarctic are up and not down. And highly variable adverse events are probably more closely associated with changes in water vapor patterns, the recent El Niño, active sunspots, aerosol levels, and a host of other factors, some of which are well known and others of which are only dimly understood.

The situation is even more complex if one looks to the long run. Climate variability has been a constant long before human beings inhabited this earth. Of course, carbon dioxide is a greenhouse gas that can trap energy. But so is water vapor, and its levels are far harder to track because its amount and distribution are not constant across the earth’s surface. Most crucially, observed cyclical patterns of temperature change do not correlate with slow but steady increases in carbon dioxide. Recent work by climate scientists Richard Lindzen and others shows that during the so-called Holocene period (roughly covering the last 11,000 years), there was a negative correlation between temperatures and carbon dioxide concentrations—strongly suggesting that carbon dioxide levels cannot be the main driver of temperature changes. It should, therefore, come as no surprise that recent climate models that have predicted sharp temperature increases have consistently run “hot,” so much so that observed increases are less than 50 percent of those predicted. As climate scientist Judith Curry points out, the uncertainties involved are large and the role of natural forces in driving temperature change are systematically underestimated.

More specifically, the over-hyped climate models have ignored two key constraints that undercut the usual doomsday projections. First, changes in temperature occur much more slowly than changes in carbon dioxide concentration. At the same time, the increase in plant growth on land has vastly outstripped temperature changes, contributing powerfully to the greening of the earth in the last ten years, and suggesting that the social “cost” of carbon dioxide could be positive. Second, recent work suggests that “doubling sensitivity”— which adds the multiplier effect needed to determine the ultimate impact of carbon dioxide changes on temperature—is far lower than the previous estimates put forward by the Intergovernmental Panel on Climate Change ten years ago.

There is, therefore, no good reason to think that carbon dioxide could bring about any short-term crisis in temperature, let alone a crisis that can only be abated by instituting very costly (and likely ineffective) changes through the Paris Accords. The agreement could easily result in trillions of dollars in wasteful expenditures. What’s worse, the agreement seems to set its targets on autopilot, without accounting for new data that might require revisions to the initial figures. Right now there is little reason to believe that putting all the accord’s provisions into place would lower global temperatures by even a fraction of a degree.

I see no gain in having the United States participate in a treaty that combines bad science with bad economics. In the long run, the United States will gain in both wealth and influence if it adopts a more restrained approach to climate change. President Trump should not let himself be scolded to move in a fashionable but unsound direction.

© 2017 by the Board of Trustees of Leland Stanford Junior University