MY ARGUMENT is that classic resource wars-hot conflicts driven by a struggle to grab resources-are increasingly rare. Even where resources play a role, they are rarely the root cause of bloodshed. Rather, the root cause usually lies in various failures of governance. That argument-in both its classic form and in its more nuanced incarnation-is hardly a straw man, as Thomas Homer-Dixon asserts. Setting aside hyperbole, the punditry increasingly points to resources as a cause of war. And so do social scientists and policy analysts, even with their more nuanced views. I've triggered this debate because conventional wisdom puts too much emphasis on resources as a cause of conflict. Getting the story right has big implications for social scientists trying to unravel cause-and-effect and often even larger implications for public policy.

Michael Klare is right to underscore Saddam Hussein's invasion of Kuwait, the only classic resource conflict in recent memory. That episode highlights two of the reasons why classic resource wars are becoming rare-they're expensive and rarely work. (And even in Kuwait's case, many other forces also spurred the invasion. Notably, Iraq felt insecure with its only access to the sea a narrow strip of land sandwiched between Kuwait on one side and its archenemy Iran on the other.) In the end, Saddam lost resources on the order of $100 billion (plus his country and then his head) in his quest for Kuwait's 1.5 million barrels per day of combined oil and gas output. By contrast, Exxon paid $80 billion to get Mobil's 1.7 million barrels per day of oil and gas production-a merger that has held and flourished. As the bulging sovereign wealth funds are discovering, it is easier to get resources through the stock exchange than the gun barrel.

Klare takes me to task for failing to acknowledge the role of "lootable" resources as a motive for war. My point is that looters loot what they can-not just natural resources, but also foreign aid and anything else that passes within reach. (Paul Collier's research, which Klare cites for support, finds that a sizeable share of African military budgets is, in effect, aid money that is looted and redirected from foreign aid.) I suspect that we don't differ much in our assessment of the effects of lootable resources within weak and failed states, but where we do part company is in the implication for policy. Fixing the problems in the Niger River Delta-the case he uses-requires a stronger and more accountable government. That means making it harder to loot resources, taming official corruption, lending a hand with law enforcement in places where oil is produced and stolen, and engaging reformist forces in the Nigerian government. Resource looting and misallocation are severe, but they are symptoms whose cures require focusing on governance.

The realities of global resource depletion are somewhat different from Klare's story. It is true that primary resources, such as oil in the ground, are now more concentrated in "armpit" countries because more readily available resources are being depleted. That fact, though, only serves to further support my conclusion: That we must redouble our efforts to improve governance because all oil-consuming countries have a stake in the good governance of their oil producers. What really matters is not theoretical oil thousands of feet underground but actual oil produced and delivered to markets. And on that front, the armpit-country story isn't so bad because those countries tend to put themselves out of business. Witness Venezuela, where production is declining even though the country is one of the world's richest in untapped resources. High prices soon follow. And with those higher prices, a spate of "new" resources becomes viable-oil sands in Canada and shale in the western United States, for example. Moreover, many oil-rich countries actually have good governance systems (at least concerning their oil), such as Saudi Arabia, the United Arab Emirates and notably the bright new star among oil-majors, Brazil. Nonetheless, I echo a conclusion from my original article-one that Klare surely shares as well-that current patterns of oil consumption are not sustainable, and urgent efforts to tame demand are also needed.

I find it striking that none of the three have attacked my characterization of China's behavior in Africa, for it is the Chinese resource scramble that most animates fear among the punditry and threat industry of a coming resource war. My original article makes a strong argument for why the conventional wisdom about China is wrong and why all oil consumers (including China and the United States) actually have strong common interests. If, indeed, that argument is widely shared among experts then some radically different policy strategies would follow.

Nobody can disagree with Paul Kern and Sherri Goodman's maxim that "wars are best avoided by preparing ahead of time for potential threats. . . ." My concern isn't with the principle but, rather, putting this bromide into practice-exactly what they accuse me of ignoring. Just as Eisenhower warned of the industrial threat industry at the end of his administration, so too must we be concerned about the arrival of military planning to the problem of natural resources. These are broader concerns I have flagged, not specifically directed at CNA (Kern and Goodman's organization), and they merit careful attention because the generous instinct of environmentalists is to welcome all who share concern about resource depletion and stress. Yet, the threat industry is notoriously bad at setting priorities for interventions that involve the broader society and economy.

The CNA report they cite (and co-authored) rightly says that climate change is a threat multiplier, but all stresses on governance systems are threat multipliers, and real security policy is about setting priorities and matching responses to threats. I have a feeling that we agree on the implications for policy, although for different reasons. I am not much worried about climate change triggering hot conflict, but I am deeply concerned about the unequal impacts on poor societies and the severe impacts on fragile ecosystems. The solutions include deep cuts in emissions (exactly how that can be done is a subject I have addressed for most of my professional career) and also much better governance systems, so that societies do a better job of coping with the changes in climate that are inevitable. Thinking about climate change as a security problem inspires a logic of hardening, securing and protecting. What's really needed is flexibility, adaptiveness and fair systems of governance-all conclusions that are broadly consistent with the CNA report.

Homer-Dixon's critique is unabashedly misleading and wrong. Like Kern and Goodman he makes much of the "hard security" dangers in climate change-imagining all manner of ways that climate disturbances can ripple into hot conflicts. Such thinking is pernicious. Good imagination can find threat multipliers everywhere. Good policy is about setting priorities where leverage is greatest. Looking only at the disturbances-weakened rural economies, increased unemployment, dislocated people-just perverts policy. The world is already destined to face a lot of climate change that won't be reversed for a century or more. So now, we must look not only at avoiding disturbances but also at improving governance. (At the same time, we must be modest in realizing that outsiders often have little useful leverage on how countries govern themselves.)

He claims that I have ignored complex causation when, in fact, that's the centerpiece of my argument, and it is precisely the voluminous work in this area that gives me trouble. The problems in Darfur and Rwanda stem from many factors, but what matters for social science, and especially for policy, is getting cause and effect right. Indeed, radical Hutus pointed to resource inequities to mobilize support, but radicals, like looters, make do with whatever irredentism is available. Had someone reallocated the cropland, the problem would not have disappeared. Hutu grievance was rooted in Rwanda's system of governance-by-minority, the inability of that governance system to make decisions that commanded broad respect, and ultimately, the inability of those in power to provide security.

Nearly all of the vast literature that Homer-Dixon applauds suffers from the affliction of severe selection bias and failure to assign proper weights to causal factors. Put a microscope on any big conflict looking for resources, and you're sure to find exactly what you're looking for. Nobody doubts that causation is complex; the dispute is on the central forces. And to Klare's point about methodology, my article focuses narrowly on hot conflict-that is, "war"-because the best way to get causation right usually requires starting narrowly. However, technological change and economic shifts away from resource-intensive industries and the globalization of most resources into commodities implies that a broader version of my hypothesis probably also holds-natural resources matter less and thus are less important for conflict, except where lootable resources coincide with exceptionally poor governance.

Homer-Dixon is particularly upset that I seem to diminish the multitude of causes for conflict. My essay does not portray a Candide of peace and endless prosperity. The world is a dangerous place; conflicts are rife; important problems, such as climate change, are mounting for lack of more serious policy attention. Looking at all these problems through conflict-colored lenses is equally dangerous, for it focuses attention on weak symptoms rather than root causes.

David G. Victor is a professor of law at Stanford Law School, the director of the Program on Energy and Sustainable Development at Stanford University and an adjunct senior fellow at the Council on Foreign Relations. He is writing a book on new strategies for managing global warming, parts of which appear in the December issue of Scientific American.