As Brookings Institution scholar Jeremy Shapiro has argued, the United States and Europe see the war on terrorism in much different ways. Americans view it as their top national security problem. Europeans, by contrast, tend to question the definition of the enterprise as a war. The former think of it primarily as a series of military and intelligence operations abroad, the latter largely as a law enforcement enterprise.

These differences are serious and significantly complicate the transatlantic security agenda. Yet given the way the threat presents itself in their respective societies, both Europe and the United States are largely right in their differing diagnoses. Rather than argue endlessly over which perspective is in the end more accurate, scholars and policymakers and citizens on each side of the Atlantic would do better to avoid blatant disagreement except when it is unavoidable, and to cooperate when feasible.

On one issue, at least, Americans and Europeans do not disagree: homeland security, or what Europeans call the defence of national territory, is the domestic front of a much broader struggle. The rising threat of radical sub-state actors using the Internet and asymmetrical means to mount spectacular attacks cannot be addressed by any one country acting in isolation. Stopping future terrorist attacks against domestic targets must involve coordination among the intelligence and law-enforcement agencies of numerous nations.

For that reason alone, NATO deserves to play a significant role in homeland security efforts. Suicide bombings in Istanbul in November 2003, March 2004 train bombings in Madrid, and July 2005 attacks against the public transportation system in London have brought the struggle against terrorism to the heart of the Alliance. NATO's original role, after all, was to protect the national territories of its members and the final objective of terrorists is to bring the war to us here at home.

This essay takes an admittedly American perspective in focusing on the remaining US homeland security agenda, recognising that on many matters Europeans will prefer a different approach. Both Europe and America would benefit from a more nuanced understanding of the other's views. I conclude with some thoughts on where it is particularly necessary to try to agree, or at least to compromise, in the broader interests of transatlantic security.

It is simply not possible to defend a large, open, advanced society from all possible types of terrorism. As Richard Betts argued in the Spring 2002 issue of the Political Science Quarterly, the United States, for example, contains more than half a million bridges, nearly 500 skyscrapers, nearly 200,000 miles of natural gas pipelines, and more than 2,800 power plants, to say nothing of subways, restaurants, movie theatres, schools and malls. The list of critical infrastructure is far too long to protect everything. Certain special measures, such as providing extremely tight security around the nation's 104 nuclear power plants, cannot be extended to all possible targets.

But by focusing on the worst possible attacks, NATO countries can establish priorities and make further progress in protecting their countries. Several guidelines should inform both future efforts and politicians' efforts to speak to a transatlantic audience about what broad principles should guide the next steps in enhancing homeland security.

Firstly, it was correct to focus initially on preventing al Qaida from carrying out attacks similar to those of 9/11. We need not stretch our imaginations to identify other key vulnerabilities vividly exposed both by the attacks in Istanbul, Madrid and London and by narrowly avoided attacks in the UK, Germany and elsewhere. We must also recognise that future attacks could encompass not only skyscrapers or train stations, but also chemical plants or the air circulation systems of stadiums.

Secondly, we should focus first and foremost on prevention. This involves obtaining good domestic intelligence on terrorist suspects while impeding their movements, their financial transactions and their communications, rather than focusing on point defence of key national assets or on mitigating the consequences of successful attacks. The latter tasks are important but are not as optimal as preventive efforts. And on this issue, the United States could benefit from the advice and expertise on offer from some of its European Allies. As Bénédicte Suzan and Jeremy Shapiro wrote in the Spring 2003 edition of Survival, one key aspect of France's largely successful campaign against domestic terrorism in the 1980s and 1990s was the development of extensive human intelligence (HUMINT) networks-an advanced capacity for HUMINT collection that is not strong enough within the US counter-terrorist arsenal.

Thirdly, since we cannot protect everything, we should worry most about possible terrorist strikes that would cause large numbers of casualties. Only slightly less critically, we should focus intensively on preventing attacks that would cause few casualties but create huge economic ripple effects, such as episodes of attempted smuggling that revealed gaping holes in shipping container security.

Here is another example of the latter scenario. If a shoulder-launched surface-to-air (SAM) missile took down a commercial airliner in route between Brussels and New York, casualties might be relatively limited compared to the 9/11 attacks-a tragedy for those involved to be sure, but in and of itself not debilitating to a NATO Ally or the Alliance as a whole. The effects on transatlantic air travel, however, would be devastating. Those economic effects also could endure much longer than those of September 11 since it would take time to form a strategy for avoiding future SAM attacks.

Another example could be the use of a radiological weapon that uses conventional explosives to disperse radioactive material in an urban area. Peter Zimmerman and Cheryl Loeb noted in the January 2004 edition of Defence Horizons that such an attack probably would not kill many people, but would likely cause mass panic. It would also probably require a costly and time-consuming cleanup-as well as the implementation of disruptive security measures throughout the attacked country.

There are also general areas of homeland security where important progress has occurred in some ways but where key shortcomings remain. Consider the transatlantic vulnerability to biological attack. Lawrence Wein and Edward Kaplan have pointed out that although antibiotic stocks for addressing any anthrax attack are now fairly robust, means of quickly delivering the antibiotics are not. Longer-term worries about biological attacks remain acute, since there could be many types of infectious agents for which antidotes and vaccines prove unavailable (or non-existent) when they are most needed.

As for air travel, most passengers are still not screened for explosives, cargo carried on commercial jets is usually not inspected, and private planes face minimal security scrutiny. For all the security improvements that have been made by United States and European carriers, moreover, fewer have been made to many other international carriers that transport large numbers of people to and from NATO countries. Given the porousness of borders and the globalisation of the world's airline industry, these problems must be solved.

More specifically, the US private sector has done very little to protect itself. Richard Falkenrath, Deputy Commissioner of Counterterrorism at the New York City Police Department, testified to a January 2005 hearing of the Senate Committee on Homeland Security and Government Affairs that from chemical plants to trucks shipping hazardous materials to skyscrapers, vulnerabilities are often acute and not far different from how they existed prior to 2001. Owners of private infrastructure know that the chances of any one facility they own being attacked are miniscule, so they are not apt to incur added costs-and concede to shareholders and neighbours that their facilities might be vulnerable-on their own volition. Yet viewed from a national perspective, this means that certain systemic vulnerabilities remain unaddressed.

The creation of the Department of Homeland Security (DHS) has not automatically led to better protection against such threats in the United States, as the hapless response to Hurricane Katrina suggests. DHS has many capable and dedicated individuals serving within it. Reorganisations, however, can distract attention from efforts to identify remaining key vulnerabilities and then mitigate them, as Deputy Commissioner Falkenrath also noted in his testimony. Carrying out a major governmental overhaul during what is essentially a time of war is a risky proposition. It is also not the way the United States has typically responded to national crises. The Department of Defense was not created during World War II, but afterwards. The Goldwater-Nichols Pentagon reorganisation in 1986 was carried out during a time of relative international peace. For their part, European countries would do well to avoid the belief that simply creating a new institution will begin to confront the problem.

I would suggest a future policy emphasis on several areas, some of which are detailed further below. More detail on these suggestions can be found in Protecting the Homeland 2006/2007, Michael d'Arcy, Michael O'Hanlon, Peter Orszag, Jeremy Shapiro, and James Steinberg (Brookings Institution, 2006). Most of these measures would be as helpful for European countries as they would be for the United States and Canada. The organising philosophy should be to protect against attacks with potentially catastrophic impact, in human, economic or political terms. In the interest of cost effectiveness, action should focus on the prevention of attacks rather than on the site defence of potential targets or consequence mitigation after attacks have occurred. A blend of all approaches will be needed. My suggestions include:

creating incentives for the private sector to protect itself more effectively, especially in sectors such as the chemical industry and high-rise buildings;

developing a better and more rigorous security system for container cargo;

greatly expanding screening of cargo on airplanes;

creating national standards for driver's licenses with biometric indicators (not photos) and, specifically, improving the biometric indicators used on US passports;

encouraging more large-city police departments to build dedicated counter-terrorism cells as New York has done;

moving to the next step in using information technology in the war on terror by creating of a "google-like" search capacity across different police and intelligence databases for correlations of suspicious behaviour;

examining how NATO countries can develop a quick-manufacture capacity for vaccines and antidotes to new pathogens that it does not now possess. As Kendall Hoyt reported in the March 3rd, 2006 edition of the New York Times, this could also be of great importance in addressing such scenarios as a possible mutation of the bird flu H5N1 virus to a form highly dangerous to humans.

To detail a couple of these ideas, it is always sound to begin discussion of a new homeland security agenda by focusing on intelligence-the most important homeland security effort since an ounce of prevention is worth a pound of cure (or consequence management). Since there is too much to protect, the only way to make homeland security successful is to stop most terrorists before they can attempt an attack.

As Jeremy Shapiro and Dean of the Lyndon B. Johnson School of Public Affairs James Steinberg have recently argued, the transatlantic homeland security agenda requires further work. For example, an assistance and extradition treaty was signed between the United States and European Union in June 2003. But as James Steinberg has written, there is still a need for measures on both sides of the Atlantic that allow the admission of intelligence information as evidence in court while protecting against its disclosure.

There are some areas where existing European efforts at homeland security exceed those of the United States. In particular, as Michael d'Arcy of King's College in London has argued, the American choice of using just a facial image as the biometric indicator in its passports is unwise. Photographs are inherently unreliable. The United States should follow the European Union in incorporating fingerprints data, and ideally both sides of the Atlantic will move to using iris data in time.

So on intelligence cooperation, and on matters such as international trade and air travel where the nature of the effort is inherently transnational, Europe, Canada and the United States must endeavour to agree or at least find a middle ground where they can each feel comfortable with a common policy. But on many other matters, such as several of those detailed above, it is acceptable for Allies to differ in their threat perceptions and their corresponding degree of preparation against the future terror threat. Nation-states are still sovereign, after all, and even close Allies can disagree when making decisions about their own relative prioritisations of risk, cost, civil liberties, law enforcement, and social cohesion.

Paradoxically, these different perceptions could even prove beneficial by allowing NATO Allies to specialise in the counter-terrorism efforts that they do best. Too often, we appear implicitly to suppose that homeland security and the broader struggle against terrorism should be conducted through military strikes abroad and a series of intelligence operations, or through multilateral law enforcement operations. In fact, homeland security efforts and anti-terrorism operations can encompass all of these options at different times and places, and often at the same time and place. In Afghanistan, NATO leaders have come to realise that measures to provide stability must be comprehensive in nature, encompassing political and economic as well as military means. In a similar way, defending national territories effectively requires the adoption of "full spectrum" capabilities ranging from law enforcement and intelligence collection to combat operations abroad.

NATO can do best what it knows: acting as a clearinghouse for information and even capability exchange where the specialisations of each NATO Ally are brought to the table and coordinated. This would add tremendous value to the struggle against terror-whether we agree to call that struggle a war, or not.