Thank you Chris for that very kind introduction. It has been a great honor to get to know Chris in the last few years. I feel that I first met Chris the old-fashioned intellectual way, before e-mail and websites and blogs and cable t.v.--by actually reading an article of his cover to cover. It was the classic 1986 article in the Harvard Law Review he wrote with Judge Douglas Ginsburg. It introduced the idea that a President could require his administration to evaluate the costs and benefits of the regulations it issued. It is a sign of the success of Chris and others at AEI that this once-radical idea (and it is certainly still radical at the Yale Law School, where I first learned of it) seems like common sense today.

It was my good fortune to be able to meet Chris DeMuth, the radical balancer of costs and benefits, years later as I was leaving government. Chris kindly provided me with a home here at the American Enterprise Institute, where I have had the opportunity to think about the subject of my talk, and he has urged me to keep my work focused not just on the purely academic realm of ideas, but on the practical world of policy. I am grateful for his confidence in my work and the invitation to speak tonight.

I also would like to thank the Lynde and Harry Bradley Foundation for funding this year’s lecture series. I am proud to be invited in connection with a foundation whose programs support “limited, competent government; a dynamic marketplace for economic, intellectual, and cultural activity; and a vigorous defense at home and abroad of American ideas and institutions.” Living in the People’s Republic of Berkeley only makes me envious of the rest of you here, who get to live in a place with all of those neat things.

Tonight I hope to shed some light on the third of leg of the Bradley Foundation’s interests, that of defending American ideas and institutions at home and abroad. I want to share some thoughts about the ongoing war on terrorism, not just as someone who served in the government on the day of the September 11, 2001 attacks, and helped develop policies to respond to al Qaeda, but as an academic who has spent the years since studying how we might finish the job we started.

By many measures, the war on terrorism has gone well. There has been no second attack on the American homeland. America’s October, 2001 intervention in Afghanistan destroyed al Qaeda’s infrastructure of bases, training camps, and supply and recruitment chains. It removed the Taliban regime that had harbored and supported al Qaeda and installed in its place a regime friendly to the United States. With al Qaeda flushed from Afghanistan, our intelligence agencies and military have succeeded in degrading al Qaeda’s organization and leadership. The United States has killed thousands of al Qaeda operatives in the field, and its killed or captured about two-thirds of al Qaeda’s leadership as of September 11, 2001. Among them, most notably, are al Qaeda’s operational leaders who planned the attacks: Abu Zubaydah, Khalid Sheik Mohammed, and Ramzi Binalshibh. Enhanced intelligence gathering and tenacious law enforcement domestically have led to the arrest, trial, and conviction of cells of al Qaeda supporters within the United States.

Nonetheless, the war continues. Al Qaeda and its allies have launched attacks against America’s allies in places ranging from Madrid, Spain, to Bali. It appears to be supporting the deadly insurgency in Iraq through Abu Musab Zarqawi, the head of al Qaeda in Iraq. It replaces killed or captured operatives with new recruits, and continues to send operatives to the United States to launch attacks. It searches for new havens in the world’s ungoverned areas, such as western Pakistan or Yemen. Al Qaeda has suffered devastating losses at the hands of the CIA and the U.S. armed forces, but it has not suffered a knock-out blow and may even be recovering.

Two ways of thinking about terrorism may move us closer toward victory. I wish to discuss, first, our strategy toward failed states, and, second, a tactical approach toward disrupting the al Qaeda organization itself. Dealing more effectively with failed and rogue states will allow us to deny terrorists the infrastructure and staging bases to organize. Launching an effort to destroy al Qaeda based on network strategies will help make sure that the terrorists will wither away once cut off from their roots.

Failed and rogue states.

Failed states constitute one of the great challenges to international peace and security. Failed states breed violence, starvation, and often a flow of refugees that impose heavy burdens on their neighbors. Failed states become anarchic areas where terrorists can build resources, train operatives, and create bases from which to launch attacks. Afghanistan witnessed the free operation of the al Qaeda terrorist organization within its territory. A more conventional central government could have been expected to prevent the large-scale, open operation of international terrorist organizations from its own soil. The Bush administration clearly recognizes the dangers presented by failed states. As it said in its 2002 National Security Strategy, “America is threatened less by conquering states than we are by failing ones.”

The response of the West and of the international legal order has been to restore or impose a nation-state framework on these troubled areas. It maintains that international security depends on the exclusive existence of independent nation-states capable of controlling their own territories, policing their populations, and discharging their international obligations. States remain the most effective means to control or prevent conduct that threatens international order and global welfare. As Michael Howard has observed, “it is not clear what alternative creators and guarantors of peaceful order could or would take the place of the state in a wholly globalized world. The state remains the only effective mechanism through which people can govern themselves.”[1] Jeremy Rabkin similarly has argued in his recent AEI book, The Case for Sovereignty: Why the World Should Welcome American Independence, that nation-states remain the only serious actors who can solve today’s international security problems.

This is undoubtedly correct--nation-states remain the only reliable means to restore and maintain international peace and security. Where the United States, its allies, and the United Nations have erred, however, is to assume that because strong nation-states are the guarantors of international stability, every territory must have a nation-state. Hence, the United States and its NATO allies have set as their goal in Afghanistan the reconstruction of state institutions, rather than a trusteeship or colonial arrangement. We have promoted Afghani political parties, elections, a parliamentary democracy, and the rule of law.

Collapse of state institutions, however, may not be a sign of a failed state, which implies that a properly functioning state is the norm for that territory. Rather, state failure may indicate that a territory or people may not be best governed by the nation-state structure. Both the League of Nations (in the form of the mandate system) and the United Nations (the trusteeship system) once recognized that certain forms of quasi-sovereignty. But even these systems were considered temporary: a developed nation would take charge of a territory and lead it toward full sovereignty and independence. In recent interventions, however, such as in Somalia, the Balkans, and Afghanistan, the United States and its allies have not considered alternate forms of governance that might be better suited to the facts on the ground. For the United States, it is either the nation-state or nothing.

The United States and its allies do more than seek to restore the nation-state structure; they also consider pre-existing borders to be sacrosanct. In Iraq, for example, the United States and its coalition do not seem to have ever seriously considered the division of the country to more closely alight with its distinct ethnic and religious groups. Indeed, the status of the United States and the United Kingdom as “occupying powers” under the relevant U.N. Security Council resolutions signals that Iraqi sovereignty and borders are to remain intact. This is despite the fact that Iraq is composed of three fairly discrete groups: the Kurds to the north, the Shiites to the south, and the Sunnis in the middle. The only significant outlier case has been the former Yugoslavia, where NATO appears to be encouraging fragmentation by protecting Kosovo from Serbian control.

Demanding the imposition of a nation-state structure and maintaining nations within their current borders is mistaken. We can see this more clearly if we compare the international system to a market. Students of alliance politics, for example, compare nations to firms in a market deciding whether to merge with each other or engage in long-range contractual relationships instead. In the case of failed states too, the United States seeks a partner to enhance its security. The current U.S. policy of establishing nation-states everywhere it can fits this way of thinking, because it is an attempt, essentially, to create other firms with which the United States can contract.

Failed states, however, do not fit because there is no other nation-state for the United States to contract with. Instead, we might compare the relationship between the United States and a failed state as one of a firm seeking to develop a natural resource. In such cases, the firm would prefer that a third party with expertise in development of the resources come in--similarly, in the context of failed states, this would lower the intervention and governance costs for the United States. However, alternative mechanisms, including the United Nations, NATO, and others have not proven successful in restoring order, and after September 11, 2001, the expected costs of allowing failed states to remain ungoverned has only increased. As a result, the United States has had to take on the expensive and difficult task of intervening directly in failed states.

The problem is that the current rules will discourage states from intervening in failed states. The United States and its allies appear to believe that international law requires that intervention must be followed by a transition to independence and full sovereignty as a nation-state. This forces the United States to bear high governance costs, all with the prospect of losing control over the territory once it reaches independence. It is as if a firm was forced to create a subsidiary to develop a natural resource, and then the legal system required it to give up control over that subsidiary without compensation.

Intervention carries with it an extremely high governance cost: the responsibility to guide the territory toward full sovereignty, often in conditions where no fully functional state has ever existed. If one agrees that failed states impose negative externalities on the international system as a whole, public choice theory predicts that interventions in failed states will be under-supplied by individual nation-states. The international legal system only makes this situation worse by imposing a rule that increases the costs of intervention, which will further depress actual interventions below its optimal amount.

A similar phenomenon occurs with regard to the maintenance of state borders. A failure of state government can signal that the pre-existing nation-state was unstable, either because of internal conflict or weakness against external threats. Devoting substantial resources to maintain an unstable state may prove futile. Again, however, American policy so far has followed the international law rule that the borders of nation-states should remain intact. Although Kosovo is a counter-example, the United States has not attempted to redraw the borders of Afghanistan or Iraq. The international law rule, by forcing intervening nations to bear high governance costs--in this case forcing together an unmanageable country--will discourage the use of force to prevent failed states from serving as breeding grounds of terrorism.

In many cases, a more effective solution might permit the division of a failed state into more governable parts. Why not divide Rwanda, for example, into two countries, or Iraq into three? We are already living through a period of fragmentation--in 1945, for example, there were 74 independent nations; today, there are 195 (counting Taiwan and Palestine). There is no natural reason why the borders of failed states ought to be accepted as a given. Economists Alberto Alesina and Enrico Spolaore have proposed that the size of nations are determined by the trade-off between benefits of larger scale--such as defense, law and order, and free trade markets--and the costs created by heterogeneity of preferences within a nation. While larger nations can provide public goods at lower per capita costs, at some point they will become so large that heterogenous groups within the nation would rather split apart than live under policies which they oppose. They also note that autocratic nations are likely to be larger, so that dictators have larger populations from which to take rents.

Alesina and Spolaore argue that as the threat of war recedes and free trade agreements expand, nations will become smaller because the benefits of size are provided by the international system. Security and free trade are precisely the international public goods that the United States has supplied for more than a half-century. By playing a hegemonic role, the United States has created the exact conditions to encourage fragmentation; a development it opposes by keeping nations intact. All of this goes to show that there is no natural reason why the United States and its allies should bear the heavy costs of maintaining failed states within their current borders.

American policy--quite unconsciously--discourages intervention and thus makes it more likely that terrorists will find more parts of the world a safe haven. What should the United States do to reverse this course? First, it should reject the notion that it must bring a failed state to independence and sovereignty. It should explore other forms of governance, perhaps in partnership with other nations, which seek primarily to guarantee internal security and the functioning of effective markets. Striving first for the grail of constitutional democracy, political parties, and the rule of law may not only be getting ahead of ourselves, it also raises the costs of intervention and hence reduces its amount.

Rather, the United States should seek to achieve more minimal goals of restoring sufficient order in a failed state to prevent it from becoming a base for terrorism. While the United Nations has proven a poor resource for achieving these ends, the United States might benefit from working with nations with large militaries and populations that could specialize in providing security in failed nations.

Second, the United States and its allies should consider more seriously the prospect of fragmenting failed states into more governable parts. Requiring that national borders remain intact increases governance costs, which reduces the amount of the public good produced by preventing terrorists from operating in failed states. If the great powers continue to supply international peace and security, and current efforts to promote free trade succeed, existing national borders may represent a suboptimal result. The United States should reject efforts to keep failed states together, and instead allow them to break up along ethnic or religious lines. This will reduce the costs of intervention, which should allow for the removal of more territory from the grasp of terrorists.

Network warfare.

Dealing more effectively with failed states addresses a long-term strategy toward cutting off terrorists from their support networks. Let me now turn to tactics. One issue that continues to confuse our society is how to understand terrorist organizations. Al Qaeda is not a nation-state--it has no territory, population, or cities to defend, no armed forces in the field, and, as we saw on September 11 and since, no intention to obey the rules of civilized warfare. While we have understood these facts in regard to some policies--most notably the decision to deny al Qaeda operatives the status of prisoners of war under the Geneva Conventions--we have not in others. We still tend to conceive of al Qaeda as organized along the lines of a national military or government, with a top-down hierarchy, chains of command, and officers and foot soldiers. Hence, the administration in the first few years after September 11 placed great importance on killing or capturing Osama bin Laden and his deputy Ayman al Zawahiri, and has made much of its successes in capturing other high-ranking al Qaeda leaders.

Al Qaeda, however, is better conceived of as a network, rather than as a classic hierarchical organization. Most hierarchies would have collapsed after suffering the kinds of losses inflicting by the armed forces and the CIA--thousands of operatives killed, two thirds of its leadership killed or captured, and its open bases and infrastructure destroyed in Afghanistan. Nevertheless, al Qaeda has shown a remarkable resiliency. A killed or captured leader seems to be quickly replaced by the promotion of more junior member, other parts of al Qaeda spring to the fore (as in Iraq), and attacks on the west continue.

These are the typical characteristics of what is known as a free-scale network. A free-scale network is not created at random--it is made up of nodes connected to each other for some purpose, whether it be communication through the internet, social and business contacts, or even transportation of goods and services. In part, what distinguishes these networks is that they do not have command-and-control hierarchies, but instead are organized around hubs, which are nodes that have a high number of connections to other nodes. They are also dynamic; if a node or hub disappears, nodes may simply move their connections to others. Free-scale networks remain remarkably immune to attack; randomly destroying its nodes will not cause it to collapse, even destroying a hub will not bring down the whole network. Decentralization is a network’s great attribute. It can quickly collect and process information from a myriad of sources, it can gather the collective efforts of thousands located in different places and connected only by a shared ideology. Since it has no real single leader, it can function even after suffering severe losses. As some parts of the network lose their effectiveness, others can suddenly grow in importance.

Al Qaeda is just such a network; each node is a terrorist brought together through a desire to promote Islamic fundamentalism in the Middle East through violent means, its hubs are particularly important facilitators, and cells are collections of nodes that are linked to the hubs. While Osama bin Laden is its symbolic or spiritual leader, it appears to operate as a loose network in which different hubs and cells have authority to plan and launch terrorist attacks on their own. Capturing or killing an al Qaeda member such as a Khalid Sheikh Mohammed or Ramzi bin al Shibh is important, but it will only lead to the discovery of the cells and plots to which he is connected; other parts of the network can continue to function.

Two tactics can prove effective in destroying such a network, only one of which it appears that the United States has pursued. Both would involve changes in our legal approaches to permissible covert action in the war on terrorism. The first and perhaps easiest is destroying the hubs of the network. Random or individual attacks on a freescale network will not work. Only a coordinated, simultaneous attack on some of its major hubs can cripple a network, leaving its pieces isolated and relatively harmless. Achieving this result requires a change in the way we think about the executive order banning assassination, which has been with us since the 1970s. In general, to quote Hays Parks’ memorandum on the subject, “assassination involves murder of a targeted individual for political purposes.” The killings of Presidents Lincoln, Garfield, McKinley, and Kennedy are regarded as assassinations.

In wartime, however, the military may legally kill members of the enemy’s armed forces, and any civilians who take part in hostilities. A nation at war may use force against members of the enemy at any time regardless of their proximity to hostilities or their activity at the time of attack. If we are at war, then the United States may carry out selective attacks on al Qaeda hubs, such as the 2002 Predator missile attack in Yemen that reportedly killed an al Qaeda organizer. It is important that we continue to consider the war against al Qaeda a war, and not, as many in the mainstream liberal media and academia argue, as a form of persistent social condition like crime. Moving the struggle against terrorism out of the category of war would render the United States incapable of carrying out the targeted strikes necessary to eliminate the hubs of the al Qaeda network.

Second, the United States could attempt to create a competitor network to al Qaeda. Al Qaeda nodes provide recruits for terrorist operations, planners to bring them together, and financiers who raise funds for training and attacks. One way to destroy a network is to cause those nodes to switch to another network, much in the way that competitors in the market for computer operating systems seek to convince users to change over. Creating a competitor network can have both positive and negative dimensions. When one hears arguments that the United States needs to bring democracy and capitalism to the Middle East, in order to provide a productive alternative to terrorism for young Arab men trapped in poverty, we are really hearing arguments for creating a western-style social network which will compete with al Qaeda. Such efforts, however, are extremely difficult. In order to succeed, the Western network must be exclusive--the new, competitor system must demand incompatibility with al Qaeda; otherwise, terrorists could simply remain members of both networks.

This would require some change in the way the United States deals with religion--under the Constitution’s Establishment and Free Exercise clauses, many believe that the government can have no involvement with supporting or interfering with religion. Creating an alternative, more positive network would require taking positions on religion – the United States would have to act more aggressively to discredit the extreme fundamentalist version of Islam promoted by al Qaeda, and it would have to more actively support more moderate forms of Islam compatible with democracy and markets. The United States should be able to treat fundamentalist Islam as it could any hostile ideology.

A different tactic would be for our intelligence agencies to create false flag terrorist organizations that would compete with al Qaeda. It could have its own websites, recruitment centers, training camps, and fundraising operations. It could launch fake terrorist operations, and it could claim credit for real terrorist strikes, helping to sow confusion within al Qaeda’s ranks. Our false terrorist network would seek to be “inter-operable” with al Qaeda, but with the ultimate aim of disrupting the functioning of its network. An important way to do this is to cause nodes to doubt the identity of others, and to question the validity of communications and contacts. Our false terrorist group would be designed to draw recruits and money away from al Qaeda, and hopefully allow the United States to gain access to its pathways and methods through agents who are tricked into switching to the fake terrorist network.

This approach too requires a different understanding of some of our laws. Some of the acts that might be required to give our fake terrorist group authenticity, such as putting up false websites, publishing fake media reports about alleged successes, diverting funds and contributions from the accounts of Islamic front groups, and recruiting individuals with questionable criminal backgrounds, may fall afoul of different federal laws. Rules prohibiting fraud, interference with the security of the internet and banking systems, and working with former terrorists might have to be changed or interpreted to not apply to operations designed to undermine terrorist groups hostile to the United States.

I hope that these ideas, drawn from different trends of thought in international economics and organization, transaction cost economics, and network theory, can prove useful in re-energizing the offensive against al Qaeda and its allies. By solving the failed state problem, we can deny terrorists safe havens and recruiting grounds, and by understanding the nature of al Qaeda as a network, we can sharpen our tactics to defeat them. Thank you for coming tonight and I look forward to your comments and questions.

John Yoo is a visiting scholar at AEI and professor of law at the University of California at Berkeley School of Law.

Note

[1] Michael Howard, The Invention of Peace 103-04 (2000).