When Generals and Ambassadors Feud

Take it from this former ambassador: Disagreements over the war in Afghanistan may do more long-term harm than short-term good.

In 2007 in Iraq, Gen. David Petraeus and Amb. Ryan Crocker set a model for civil-military collaboration: They never let daylight show between their positions -- not to outsiders, not to official Washington, not even to their own staffs. In providing differing advice to Washington over troop levels in Afghanistan, General McChrystal and Ambassador Karl Eikenberry have diverged from this model.

Ambassador Crocker wisely recognized that the U.S. president, the congress, and the American people were looking primarily to Gen. Petraeus and his 160,000 troops to secure Iraq, and only secondarily to Crocker and his 1,000 diplomats and aid workers. Crocker chose to fight his policy battles not in Washington, but in Iraq. Petraeus for his part, was very sensitive to the need to secure unity of effort with his civilian partner, and to harness the expertise of his large and competent staff.  McCrystal and Eikenberry don't seem to have established the same chemistry.

Ambassador Eikenberry's reported recommendation -- that troop reinforcements be withheld until Afghan President Hamid Karzai demonstrates unmistakable signs government reform -- has a clear logic, and an equally clear limitation. Of course, the United States and its allies want Karzai to crack down on corruption, to appoint competent officials, and then to back them up. But are they willing to put their own mission, and the lives of their own troops, at greater risk should Karzai remain recalcitrant?

The dilemma mirrors one that I saw play out as a young Foreign Service officer serving under Averell Harriman, who was then heading the American delegation to the Vietnam peace talks. At one point early on in that multi-year effort, several members of our delegation expressed frustration at the South Vietnamese government's resistance to a Washington proposal for the North. Why, they asked Harriman, couldn't the United States successfully pressure South Vietnamese President Nguyen Van Thieu to go along? 

Harriman responded that client regimes held one card that trumped any pressure their much more powerful sponsors could bring to bear: They could threaten to collapse.

Unfortunately, this pretty well describes the dilemma Barack Obama faces in dealing with Karzai. The United States can threaten Karzai's political survival, and he can respond by threatening the success of the U.S. endeavor. Of course the fate of Karzai's own regime should mean more to him than it does to the United States. But what if it doesn't?

McChrystal and Eikenberry both have impressive credentials. Their selection for their current posts reflected Obama's determination to field his A team in Afghanistan. But Eikenberry's main qualification is also a potential source of tension, as he was one of McChrystal's predecessors commanding U.S. and NATO forces in Afghanistan in 2005 and 2006 (as pictured above). Gen.McChrystal has, with some justification, voiced criticism of the American and allied efforts in Afghanistan in the past. Gen. (now Amb.) Eikenberry would be less than human if he did not bridle at such criticism, also with some justification since Eikenberry lacked anything like the force levels and other resources that McChrystal already enjoys.

In the short term, President Obama has probably profited from getting candid, if differing, advice from his two principal on-the-scene representatives. But sustained divisions of this sort are likely to have a pernicious effect on his administration's prospects for success in Afghanistan. Secretary of Defense Robert Gates and Secretary of State Hillary Clinton have so far proved remarkably collegial and collaborative. Like the Petraeus-Crocker partnership, the relationship between Gates and Clinton would seem a model for their lieutenants in the field to emulate.



The Dependent Colossus

Although globalization today reinforces American power, over time it promises to have the opposite effect.

The United States plays a central role in all dimensions of contem­porary globalization, which at its core refers to worldwide networks of interdependence. A network is simply a series of connec­tions of points in a system. But networks can take a surprising num­ber of shapes and architectures and vary enormously in centralization and complexity of connections: a spider web, an electricity grid, a metropolitan bus system, and the In­ternet, for example. Theorists of networks argue that un­der most conditions, centrality in networks conveys power -- that is, the hub controls the spokes.

Some see globalism as a network with an American hub and spokes reaching out to the rest of the world. There is some truth in this picture, as the United States is central to four forms of globalization: economic (the United States has the largest capital market), military (it is the only country with global reach), social (it is the heart of pop culture), and environmental (the United States is the biggest polluter, and its political support is neces­sary for effective action on environmental issues).

Those who advocate a hegemonic or unilateralist foreign policy are attracted to this image of global networks. Yet there are at least four reasons it would be a mistake to envisage contemporary net­works of globalism simply as the hub and spokes of an American empire that creates dependency for smaller countries. First, the architecture of networks of interdependence varies ac­cording to the different dimensions of globalization. The hub-and-spokes metaphor fits military globalism more closely than economic, environmental, or social globalism because U.S. dominance is so much greater in that domain. Even in the military area, most states are more concerned about threats from neighbors than from the United States, a fact that leads many to call in U.S. global power to redress local imbalances. For instance, a U.S. presence is welcome in most of East Asia as a balance to rising Chinese power.

At the same time, in economic networks a hub-and-spokes image is inaccurate. In trade, for example, Europe and Japan are significant alternative nodes in the global network. Environmental globalization -- the future of endangered species in Africa or the Amazonian rain forest in Brazil -- is also less centered on the United States. And where the United States is viewed as a major ecological threat, as in the production of carbon dioxide, there is often resistance to U.S. policies.

Second, the hub-and-spokes image may be misleading in that it fails to take into account reciprocity or two-way vulnerability. Even militar­ily, the ability of the United States to strike any place in the world does not make it invulnerable, as Americans learned at high cost on September 11, 2001. And while the United States has the largest economy, it is both sensitive and poten­tially defenseless to the spread of contagions in global capital mar­kets, as Americans discovered in the 1997 financial crises. In the social dimension, the United States may export more popular culture than any other country, but it also imports more ideas and immigrants than most countries. Finally, the United States is environmentally sensitive to actions abroad that it cannot control. Even if the United States took costly measures to reduce emissions of carbon dioxide at home, it would still be vulnerable to climate change induced by coal-fired power plants in China.

A third problem with the simple hub-and-spokes dependency im­age is that it fails to identify other important connections and nodes in global networks. New York City is important in the flows of capital to emerging markets but so are London, Frankfurt, and Tokyo. In social and political globalization, Paris is more important to Gabon than Washington, D.C.; Moscow is still more important in Central Asia. Maldives, only a few feet above sea level in the Indian Ocean, is particularly sensitive to the poten­tial effects of carbon dioxide produced in the rest of the world. It is also completely vulnerable, since its sensitivity has to do with geography, not policy. At some time in the future, Beijing will become more relevant to Maldives than Washington, because China will eventually outstrip the United States in the production of greenhouse gases. For many countries, the United States will not be the center of the world.

Finally, as the prior example suggests, the hub-and-spokes model may fail to take into account changes that are taking place in the architecture of global networks. Network theorists argue that central players gain power most when there are structural holes -- gaps in commu­nications -- between other participants. When the spokes cannot communicate with each other without going through the hub, the central position of the hub provides power. When the spokes can communicate and coordinate directly with each other, the hub be­comes less powerful. The growth of the Internet provides these inex­pensive alternative connections that fill the gaps. As the architecture of global networks evolves from a hub-and-spokes model to a widely distributed form like that of the Internet, the structural holes shrink and the power of the central state is reduced. It is true, for now, that Americans are central to the Internet; at the start of the 21st century, they composed by far the largest contingent of all Internet users. But by 2003, projections suggest that the United States will have 180 million Internet users and that there will be 240 million abroad. This gap will be even more pronounced two decades hence, as Internet usage continues to spread.

English is the most prevalent language on the Internet today, but by 2010, Chinese Internet users are likely to outnumber American users. Chinese Web sites will be read primarily by ethnic Chinese na­tionals and expatriates, and Chinese will not dethrone English as the Web's lingua franca. However, Chinese power in Asia will increase because Beijing will be able, in the words of Harvard China scholar Tony Saich, "to shape a Chinese political culture that stretches well beyond its physical boundaries." And China will not be alone. With the in­evitable spread of technological capabilities, more widely distributed network architectures will evolve. At some time in the future, when there are a billion Internet users in Asia and 250 million in the United States, more Web sites, capital, entrepreneurs, and advertisers will be attracted to the Asian market.

The United States has been described as bestriding the world like a colossus. Looking more closely, we see that U.S. dominance varies across realms and that many relationships of interdependence go both ways. Large states such as the United States -- or, to a lesser extent, China -- have more freedom than do small states, but they are rarely exempt from the effects of globalization. And states are not alone. For better and worse, technology is putting capabilities within the reach of individuals that were solely the preserve of government in the past. Falling costs are increasing the thickness and complexity of global networks of interdependence. The United States promotes and benefits from economic globalization. But over the longer term, we can expect globalization itself to spread technological and economic capabilities and thus reduce the extent of U.S. dominance.