It doesn't take much imagination to see that pumping one pollutant into the atmosphere in an attempt to offset the effects of another could backfire. It may also be impossible to demonstrate which adverse climate events were caused by which single geoengineering intervention. That throws a wrench into the traditional research model: It's one thing to study the effects of a past volcanic eruption or to fiddle in a lab with self-contained experiments. It's quite another to devise an experiment that could be conducted in the real world. It would be all too easy to blur the line between experiment and deployment. That and many other questions need to be answered, lest we enter wholly unchartered territory when it comes to playing with the atmosphere of our shared home.
Seat belts are good, but there's no avoiding speed limits
Talk of geoengineering inevitably leads to the question of "moral hazard." Will the exploration of these technologies lull humanity into thinking that it need not act responsibly and cut carbon emissions? Perhaps. Seat belt laws may make some drivers feel so safe that they drive more recklessly. Still, that is hardly an argument against those laws.
While the international community should not abandon efforts to limit carbon emissions, scientists must also be prepared to take geoengineering seriously. Humanity may already have passed so many global warming tipping points that - even with radical action to cut emissions - it may be important to have some form of geoengineering in our toolkit.
The worst we can do is fall into the trap of thinking geoengineering is a panacea to our climate change problem. While its initial costs may be seductively low, no one knows the unintended consequences of trying to alter the planet's atmosphere. Just as it seems to cost almost nothing to emit carbon - leading all of us to emit more than we ought to - geoengineering may appear cheap at first, only to leave humanity and nature to foot a much larger bill later on. "Free riding" turns out not to be cheap after all. "Free driving" may face the same conclusion.
Nor is it likely that everyone will face the same costs. Climate change does not affect all areas of the globe evenly. Neither will geoengineering. What if it leads to a further drying out of the southwestern United States or of sub-Saharan Africa, or to flooding elsewhere?
While the risks cannot be ignored, not even considering geoengineering research is clearly not an option. Desertification and flooding are also among the many consequences of unchecked global warming. The benefit-cost calculation of geoengineering must take these effects into account.
The fact that climate change's effects are distributed unevenly around the globe may also lead some nations to experiment with geoengineering on their own. India's national security advisor wouldn't be doing his job if he didn't at least consider countering the monsoon effects of carbon with relatively small amounts of extra sulfur. And Bangladesh's finance minister would be remiss if he didn't weigh the all-too-real costs of moving tens of millions of people against the benefits of cloud-brightening (another possible way to deflect more sunlight back into space).
In short, it will not just be up to U.S. scientists or a handful of technologically advanced countries to weigh the pros and cons of geoengineering. These technologies will be available to many countries - and as we see today, world leaders don't always succeed in working together to combat the threat of climate change.
All it takes is a single actor willing to focus on the purported benefits to his country or her region to pull the geoengineering trigger. The task with geoengineering is to coordinate international inaction while the international community considers what steps should be taken. The fate of the planet cannot be left in the hands of one leader, one nation, one billionaire.
Fortunately, we are still many years off from the full "free driver" effect taking hold. There's some time to engage in a serious global governance debate and careful research: building coalitions, guiding countries and perhaps even individuals lest they take global matters into their own hands. In fact, that is where the discussion stands at the moment, with a governance initiative convened by the British Royal Society, the Academy of Sciences for the Developing World, and the Environmental Defense Fund, among other deliberations guiding how geoengineering research should be pursued.
With time come the "free drivers"
The clock, however, is ticking. A single dramatic climate-related event anywhere in the world - think Hurricane Katrina on steroids - could trigger the "free driver" effect. That event need not be global and it need not even be conclusively linked to global warming. A nervous leader of a frightened nation might well race past the point of debate to deployment. The "free driver" effect will all but guarantee that we will face this choice at some point.
"Free riding" and "free driving" occupy opposite poles of the spectrum of climate action: One ensures that individuals won't supply enough of a public good. The other creates an incentive to engage in potentially reckless geoengineering and supply a global bad. It's tough to say which one is more dangerous. Together, these powerful forces could push the globe to the brink.