Climate change is shaping up to be one of the great challenges of the century. Though its effects can be felt today, the most worrisome impacts lie ahead. It is therefore only fitting that young people are especially passionate about how to tame carbon dioxide and the other gases that cause warming. And they are turning to their educational institutions, particularly colleges and universities, to do just that.

But how? What can American universities actually do about the climate crisis? Across campuses, pressure is mounting for two answers to this question: divestment and refusing grants and other support from big carbon.

Divestment has been more visible. As in the good old days of the anti-apartheid movement, students and faculty are protesting to get endowment funds yanked from offending companies—now, those in the carbon business. It is an expansive list, including firms that produce and distribute hydrocarbons as well as those, such as many power companies, that burn prodigious quantities of the stuff. Administrations have signed on in different ways at Brevard College, Green Mountain College, College of the Atlantic, the New School, and Stanford; overseas institutions, such as the Australian National University in Canberra, are also onboard. Some public-interest foundations have also followed suit. At the Climate summit in Paris last December, hundreds of institutions made vague pledges to cut back on fossil fuel investments.

The best strategy is not symbolic action but research and teaching.

Divestment sounds like a clean and powerful response: starve big carbon of capital. In reality, though, it is a symbolic sideshow because capital is fungible. The total value of all American university endowments is about $430 billion, a tiny fraction of the roughly $200 trillion in global assets. When universities exit, many other investors stand at the ready to buy their shares.

Divestment enthusiasts like to claim, nonetheless, that symbolic acts matter—they mobilize and focus pressure. The South African case offers some evidence that symbolic campus action worked. But one data point doesn’t prove a theory. In many other instances—from tobacco companies to firms doing business in Israel and Sudan—pressure for divestment, along with some actual divestment, has come and gone with no practical impact.

What was effective in South Africa rarely works in other contexts. Divestment may have played a role in ending apartheid, but as part of a larger historical dynamic that doesn’t apply to other cases. The collapse of the Soviet Union also made South Africa less indispensable to the West. If divestment helped, it did so in concert with government-instituted travel bans and widespread social pressure against easily identified firms.

In contrast, identifying carbon targets is much harder because fossil fuels pervade the global economy and because the “bad guys” themselves are often the best hope for mitigating the emissions that cause climate warming. For example, it is fashionable to go after big oil, but all of the major Western oil companies also have vast holdings in natural gas—the cleanest of the fossil fuels. Indeed, the largest single reduction in emissions worldwide to date is rooted in the shift from coal to gas in the U.S. electric power fleet. When it comes to cutting carbon in business, the villains and the heroes get all mixed up.

And while pretty much everyone agrees that coal is the most polluting of the fossil fuels, even there, divestment faces practical problems. In some countries, cheap coal is the backbone of cheap electricity—a boon for the poor even as it harms the planet overall. But the harm must be kept in perspective. If the more than 1 billion people currently without access to electricity were connected to a coal-powered grid, and if they drew moderately from that grid, global emissions of CO2 would rise about 1 percent. Today, many Western leaders are trying to starve coal projects of new public financing, but they haven’t thought as much about the potential damage if their scheme actually worked.

Even if we ignore such tradeoffs, the reality is that almost all long-term value investors have already left. With the collapse in global commodities and lots of regulatory pressure already on coal, stocks in big coal companies, such as Peabody Energy, have fallen dramatically from their peaks, some by as much as 99 percent. I suspect that some university leaders have agreed to leave coal because doing so will have no actual impact on their fund managers plans.

In most universities and foundations, the endowment is a “fund of funds.” Most of the money is managed outside the university—in private equity pools, hedge funds, and the like. Even if they wanted to, in-house endowment managers couldn’t readily add or subtract individual firms and industries from their university’s portfolio. When Stanford decided to remove coal from its portfolio, the decision applied to less than 25 percent of the university’s endowment—the portion it manages itself. Harvard, exceptional in part because of its sheer wealth, controls just 40 percent of its endowment in-house.

So divestment is probably a dead end. Perhaps recognizing this, some campus activists are pushing the next best thing: refusing research funding from big carbon. Perhaps no university has faced this question more intensely than MIT, which has raised prodigious quantities of money from the energy industry. Much of MIT’s effort has centered on its highly successful MIT Energy Initiative (MITei), which has attracted $600 million in funding commitments since it was announced in 2006. (Full disclosure: I earned my Ph.D. at MIT and was present at the creation of MIT’s Joint Program on the Science and Policy of Global Change, which also takes money from the fossil-fuel industry and is now a semi-autonomous part of MITei.)

MITei has taken money from almost everyone. This diversity provides another reminder that, when it comes to energy-industry patrons, the white hats are hard to distinguish from the black. In the United States, the electric companies with the biggest coal fleets often have the most innovative programs aimed at modernizing the grid and cutting emissions. My university, the University of California, San Diego, is partnering with Texas-based NRG Energy to test new electric vehicle charging equipment. Is that relationship acceptable even though NRG, in addition to its whizz-bang work in vehicle electrification and solar power, has a large coal fleet?

The refuseniks seem to be worried about two dangers that flow with money from the fossil-fuel industry. One is that firms will use funding at elite institutions to replace their dirty crowns with haloes of virtue. This charge has most often been leveled at ExxonMobil, which has been an active funder of clean energy research—notably when it anchored the ongoing Global Climate and Energy Project (GCEP) at Stanford University. (I was a professor at Stanford through most of GCEP’s early years but didn’t draw funding from the effort.)

The halo argument has had many flavors. One is that the university’s sparkling reputation becomes tarnished by bedding down with big carbon. Money alone may tarnish; distorted research agendas even more. Another concern is that research funding offers a relatively cheap way to launder big carbon’s more questionable practices. ExxonMobil, for example, bought ads on the op-ed page of the New York Times announcing its investment in GCEP.

But these costs need to be weighed against those of refusing funding—beyond the loss of the funding itself. For one thing, energy companies have useful data and access to terrain that researchers need to investigate. Many energy systems are hard to understand without an inside look at the real frontier of technology and business practices. GCEP learned as much when, for example, it studied how CO2 injected underground would move and whether it would actually stay below the surface. Real industry data and experience were essential. Academics don’t and can’t, on their own, run projects that sequester a million tons of CO2 deep underground per year—that is something industry, such as Statoil in the North Sea or BP in Algeria, does.

What is really hard in decarbonizing the world’s energy system isn’t imagining the many cool ways it could be done if money were no object or the details of large-scale engineering were irrelevant. It is learning which of those ways can scale up cost-effectively. To that end, university researchers need to be able to work with firms such as SaskPower, Summit Power, and Southern Company, which are building the world’s first full-scale power plants that sequester carbon underground.

Refusing big carbon’s money also constrains relationships with the industry’s senior management. Some environmentalists don’t realize how important these relationships are in the grand scheme of climate change mitigation. They see big carbon as an implacable enemy—firms run by fat cats who know the science of climate warming and willfully refuse to acknowledge that their business model is doomed. Only by crushing them is it possible to open the corporate space to new industries and make progress in cutting emissions.

But the reality is usually different. The leaders of big carbon firms are normal humans. They connect the dots between disparate pieces of scientific information in self-interested ways. What really matters is that they learn to see the climate challenge as something new and fundamental, not simply a marginal problem to be managed. Changing how senior management thinks about these issues—helping them connect the dots in different ways—requires engagement based on trust. On several university campuses, I have observed that process as firms turned to their trusted academic partners to brief senior management. Most strikingly I saw that at BP, working closely with much of senior management over several years as they funded work in my lab that sometimes contradicted the firm’s priorities. Engagement doesn’t always work, but shutting the door almost always ensures that nothing changes.

Keeping that door open to senior management isn’t, of course, a license to take money from anyone under any circumstance. The standards that have always applied to interested funding apply to money from big carbon as well. Relationships should be at arm’s length, transparency is essential, and results must be published openly.

So what should the campus do about carbon? The best answer is not symbolic action but research and teaching—what universities are designed to do, after all.

That is how we will fundamentally rethink energy and agriculture systems, without prejudice toward politically popular but potentially unwise choices. Campus voices call for a massive increase in renewable power, which in turn requires integrating variable and intermittent energy into a stable grid. But, ironically, decarbonization might be better achieved with fossil fuels—as long as sequestration can keep up with CO2 output. Another option one hears little about from divestment advocates is nuclear power. This need not come only from the big reactors that have already proven cost-effective in many markets, but also smaller, “modular” models that could be much safer to operate and are less easily exploited for dangerous fissile material if their cores fall into the wrong hands.

Researchers at universities across the United States are developing and studying all of these technologies and more. Nearly always, the best research is being done with industry funding and engagement. If some of these options are removed from consideration because the ideas or the funding sources are not palatable, we will be fighting carbon with our hands tied behind our backs. Indeed, the latest report from the Intergovernmental Panel on Climate Change shows that, if all options are pursued, deep decarbonization over the coming five or so decades is feasible and affordable relative to the costs of humanity’s other existential missions. But eliminating options such as carbon sequestration or nuclear power (or both) drives up costs many times over and makes it impossible to cut emissions quickly.

We need unfettered research in order to assess decarbonization from many angles and disciplinary perspectives. Research shouldn’t focus only on engineering, although that is essential. Funding should be set aside for “red teams”—researchers who actively look for holes and flaws in sponsored work. And the social sciences and humanities must be engaged on issues of justice and social change: society shapes, and is shaped by, the appearance of radical new technologies, which bring with them risks and opportunities. Had there been more social research on nuclear power in the 1960s or early renewables in the ’80s, we might have learned, earlier, how these technologies can be difficult to site and scale.

As university researchers work on these ideas, they can experiment with their own physical plant. Modern universities are huge—in fifteen states, the university system (including university hospitals) is the state’s largest employer—and their campus complexes are well suited to testing microgrid and renewable power technologies, for example. Living laboratories are also good for teaching about both futuristic technology options and the real-world decisions that come into play when deploying technology for a customer who cares about cost and reliability.

It takes time to do this sort of interdisciplinary work. That is what I learned from Steve Schneider, a mentor of mine, physicist, climatologist, pioneer in politically popular areas of research—on the impacts of climate change—and in unpopular ones, too, such as the climate effects of nuclear war. Steve worked across disciplines, and he maintained that it took two to three years of solid effort—attending seminars in unfamiliar fields, learning how faculty in those fields ask and answer questions—before the first fruits of truly interdisciplinary research and teaching would bear.

There are few places where one has the luxury to make that multi-year investment with no immediate return in sight. A well-funded, freethinking university is one.