Mr. Richman has written an extremely timely and relevant piece, which proposes if not a novel way of countering cross-jurisdictional law-enforcement conflicts in the war on terrorism, one that is realistic and holds some promise for success. Importantly, he has also—perhaps inadvertently—highlighted a key debate in the current struggle against international terrorism: the extent to which the federal government should create entirely new bureaucracies and institutions, charged with collecting and analyzing domestic intelligence, including information on American citizens.

When the role of the CIA, America’s first peacetime intelligence service, was initially being discussed in the late 1940s, the above questions were very much on the minds of those involved. Given our experiences in World War II, including our knowledge of the frightening Gestapo apparatus in Nazi Germany and the new realities of the emerging conflict with the Eastern Bloc, what was the appropriate structure for this new service? Should its creators risk the use of the more unsavory tools in the espionage arsenal (e.g., blackmail, misinformation, and covert paramilitary action) against Americans by granting the CIA and the new Director of Central Intelligence a domestic intelligence-gathering function and law-enforcement powers? Or should those functions (split into intelligence collection, counterintelligence, and law enforcement) remain the purview the FBI and therefore remain subject to Fourth Amendment safeguards?

President Truman, with the support of, among others, William “Wild Bill” Donovan, the grandfather of modern American foreign intelligence and the head of the wartime Office of Strategic Services, took the latter view and insisted the National Security Act of 1947 include a provision barring any “internal security” function for the CIA. While the CIA has more than once in its history stepped over this line—most notably with the Operation CHAOS program, which investigated whether the radical student protests of the 1960s and 1970s were being directed by foreign agitators—this postwar structure has largely worked well. It was especially appropriate to the nature of Cold War operations, in which the primary threat was Soviet espionage in the American political, diplomatic, business, and national-security communities. Accordingly, the FBI has amassed significant experience in the investigation and prosecution of these spies, though it suffered some significant embarrassments with moles, such as Robert Hanssen, and institutional failures, such as Waco or, of course, September 11.

The question, then, is just how much does the current big-picture “threat matrix” recommend changes in this way of doing business? I would argue, and I think Mr. Richman would agree, that the 9/11 Commission’s recommendation against the formation of a massive separate domestic-intelligence agency must form the basis for the way our society answers this important question. Mr. Richman argues that the extensive local contacts, multi-jurisdictional task forces and frequent cross-pollination of personnel among and between local and state law-enforcement agencies and the FBI means that the bureau, from a counterterrorism perspective, is more likely to get the right information in the right place at the right time. Mr. Richman’s larger point is that the bureau should actively seek to make local, state, and regional police forces significant stakeholders, especially in the emerging domestic-intelligence net. In doing so, Mr. Richman says, the FBI will foster both accountability and effectiveness in domestic intelligence by linking it to the trust and local contacts with particular communities, which police departments are likely to have in greater abundance than the FBI, the Homeland Security Department, and other federal agencies, who are often seen as “folks from Washington come to help the local boys.”

I think his perspective is probably accurate, although the record of joint terrorism task forces since 9/11 (among other subjects) warrants some skepticism. The broader point, unaddressed in Mr. Richman’s piece, is that for domestic intelligence-gathering and counterintelligence on American soil, a law-enforcement model—federal or local—is likely to be more effective and less prone to abuse than an imported foreign-intelligence model.

Law enforcement at any stratum of government is going to be focused on cases rather than the broader policy initiatives or, unfortunately, political agendas of federal agencies including to some extent even the CIA. These cases will tend to be particular, with a certain set of facts and a limited scope. The FBI hierarchy will assign agents to solve these cases, and either apprehend the perpetrator or prevent him or her from succeeding. A neophyte domestic-intelligence agency, however, would have a structural incentive to justify its own existence by going beyond this case-based approach. Instead of gathering evidence for a courtroom, a domestic intelligence agency would likely begin to use the emerging tools of the technological dragnet, such as data-mining, to monitor vast pools of Americans, looking for general trends suggesting terrorist activity. What types of religious texts, for instance, are being imported into Detroit mosques? What kinds of protest groups are certain foreign students drawn to? What’s the chatter on left-wing Web sites? And so on.

The FBI does things—or at least in the past has done things—very differently. As pointed out by the 9/11 Commission, the FBI, as a law-enforcement agency, is used to performing its functions in compliance with the law and the Constitution. Its agents have been trained extensively, for instance, in procedures to maintain integrity in the chain of custody and how to comply with a wiretap ascertainment requirement (meaning that if it’s the wrong person on the line being tapped, the agent has to shut down the surveillance), details that the foreign intelligence operations are able to largely ignore. By housing the functions of domestic intelligence and criminal-justice investigation in one agency, subject to constitutional limits and checks on abuse, we encourage the sharing of resources, experiences, and information in a way that enriches each and is mindful of constitutional principles. (Of course, we must actively work to ensure the opposite does not occur—that law-enforcement agencies take on more characteristics of intelligence-gathering agencies, which operate outside the parameters of our Fourth Amendment’s restraints.)

At the time of writing, it appears that the general consensus in Washington on intelligence reform favors more bureaucratic centralization (a whole other can of worms) in addition to new-agency creation. Unfortunately, some in Congress and the current administration are looking to break down certain legal barriers that have kept the FBI from turning into a pure domestic-surveillance agency. In just one example, the Senate is considering the Tools to Fight Terrorism Act, which would expand many of the new surveillance and investigative powers granted in the 2001 USA Patriot Act. Among dozens of other provisions, the bill would, for instance, expand administrative subpoena power to allow the Department of Justice to demand Americans’ personal records in terrorism cases without going through a grand jury or judge, and would further limit the discretion of judges in dealing with secret evidence. Clearly, we have much work to do.

The argument for keeping separate the tactics and mission of a foreign intelligence service and those of a domestic counterintelligence agency seems to suggest that such legislation might not be the best use of Congress’s time. Instead of trying to make the modern FBI a secret, domestic “foreign” intelligence service, Congress should be trying to make sure it has adequate investigative resources, including the most up-to-date information technology, to become the best national law enforcer and constitutional protector it can be; and that it uses the extensive powers it already has (and had even before 9/11) more consistently, aggressively, and effectively than it did before 9/11. Such a focus, though perhaps less chest-poundingly sexy than proclaiming a new “Son of Patriot Act” or a brand new intelligence agency, would be an important adjunct to Mr. Richman’s debate-starting essay.