For some time in the energy policy world, there’s been a trend toward trying to use subtle techniques, often termed “nudges,” to change people’s behavior. For instance, if you learn that your neighbors use less electricity than you do, you might be promoted to act similarly, due to the strong power of social norms.

Writ large, behavior changes like this could have a substantial impact on overall U.S. energy use and greenhouse gas emissions. But there’s a problem: What if some attempted behavioral nudges actually backfire, and don’t produce the desired change at all — or produce unexpected and undesirable changes?

Humans are, after all, notoriously quirky and weird beasts. There is even some research suggesting so-called “negative spillover” can occur in the wake of certain pro-environment behavioral changes. Thus, to give a hypothetical example, getting people to turn the lights out when they leave a room may make them feel like they’ve shown their environmentalist cred already and don’t have to do anything else – an effect that scholars call “moral licensing.”

So how often do such backfires happen, and why?

Psychologist Heather Truelove of the University of North Florida and her colleagues set up an intriguing experiment to test when virtuous, and vicious, cycles of environmental behavior occur. Their study involved a sample of 231 college students — fair-sized for these types of studies, though some research has questioned how representative college students are of the population as a whole. The research was recently published in the Journal of Environmental Psychology.

In the experiment, study subjects visited a psychology lab setting — but the actual test wasn’t what they thought. When they arrived, the desk where they were supposed to sit was a mess. Hurrying to clean it up, and apologizing, the experimenter asked each student to help out — to “toss” a plastic water bottle, which had been sitting on the desk, in the bin outside.

In different conditions of the experiment, the hallway featured either a prominent recycling bin alongside a trash bin, with a sign telling people to recycle, or an easy to reach trash bin, filled with recyclable objects (including a plastic water bottle), very near the lab and closer than the recycling. Thus, one experimental condition strongly nudged the students to recycle the bottle, while the other condition strongly nudged them to throw it in the trash. (In the “control” condition, the desk was messy but there was no bottle and students weren’t asked to do anything.)

The researchers also subtly labeled the bottles, so they could keep track of what happened to each one — where each student chose to dispose of it.

Then came a more standard battery of psychology measures in which the students were asked, among other things, about their political beliefs, environmental identities, views on global warming, and whether they supported the idea of a “campus green fund” in which each student would pay $ 20 extra to help maintain a nature preserve on campus.

And that’s where the very surprising finding came in: having recycled the bottle led Democrats, as a group, to be somewhat less likely to support the green fund. “Democrats who recycled the water bottle had lower environmental identities and were less supportive of the green fund than those in the control condition,” the study found.

In other words, at least in this sample, environmentalist-leaning Democrats showed an apparent “backfire” effect after recycling a water bottle.

Republicans, meanwhile, looked very different. Republicans who had just recycled the bottle were more supportive of the green fund than those who had just thrown a bottle in the trash – more like what you might expect to see. The researchers didn’t term this “spillover,” but noted that the difference was statistically significant.

As noted above, researchers also found that the act of recycling had lessened Democrats’ tendency to project an environmentalist identity (as a group), which presumably explained the result.

“Somehow the act of recycling among Democrats actually reduced their environmental identities, and that’s what led to the lessened support for the proposal on campus,” Truelove said.

At this point, you’re right to be scratching your head. Why on Earth would an environmentally friendly act make presumably environmentally friendly Democrats lessen their commitment?

Truelove suspects the key is that recycling — especially in this instance — is a pretty easy, even effortless, behavior. It doesn’t require that much sacrifice. And she thinks this, in turn, reminds people of how they fall short in terms of grander, tougher environmental behaviors.

“That easy behavior is somehow signaling to them that they are not living up to their part of the bargain, in terms of their environmental identity,’ she says. “So instead of feeling guilty….the recycling is somehow priming them to say, ‘eh, I don’t really care about the environment that much.’”

However, it’s also certainly possible that something else could explain these results. Ruth Greenspan Bell, a public policy scholar at the Woodrow Wilson Institute who focuses on changing human behavior with respect to energy use, notes that people are often very confused about what is, and isn’t, recyclable. “I have to question whether people in this study knew whether a plastic water bottle was actually recyclable, for example,” she noted by email. Bell also said that students may have rejected the idea of donating $ 20 to the fund because of their distrust of their university and its priorities, rather than any loss of environmental sentiment.

Certainly this one study, on its own, can’t be taken as proof that recycling as a whole is triggering backfire effects. It’s too surprising and small-scale to justify such generalizations. “Our findings are contrary to previous theorizing and show that, at the very least, the picture is more complicated than what has been proposed. Additional research is clearly needed,” the study concludes. That would include work that broadens beyond student samples.

Still, the paper raises questions about just what happens when we attempt to “nudge” people to be greener. The implication is that you’d better be careful that you know the psychological consequences before undertaking widespread nudging initiatives — lest they have the opposite of the intended effect.

“Nudging Republicans to do easy behaviors might help promote future environmental policy support, whereas nudging Democrats to do easy behaviors might backfire,” Truelove said.

Read more at Energy & Environment:

What Florida’s ancient past tells us about sea level rise today

Bad news for bats: Lethal white-nosed fungus is now in the Pacific Northwest

We’ve been getting these key details about Greenland’s melting all wrong

For more, you can sign up for our weekly newsletter here, and follow us on Twitter here.