Geof, you raise three basic issues. The first is whether a domestic intelligence agency would have less regard for constitutional rights than the FBI, the second whether adequate attention has been paid to security measures that do not involve curtailing civil liberties, and the third whether the post 9/11 restrictions on civil liberties are more serious than I believe.
On the first, you are right that an agency that has no arrest or other law enforcement powers--and that is the essence of a domestic intelligence agency that is separate from any criminal-investigation or other police body--will not worry about the exclusionary rule except in cases in which it believes the best response to a terrorist threat is prosecution (and hence referral to the FBI). On the other hand, since such an agency would have no powers of arrest, and little interest in obtaining evidence of crime, the occasions for invoking an exclusionary rule against it would be fewer. Moreover, one reason for an exclusionary rule is recognition of the heavy costs that a criminal prosecution imposes on people. A domestic intelligence agency wouldn't be in the prosecution business. And finally, the exclusionary rule has considerably eroded, as you know, and the erosion isn't going to be reversed. It is due partly to the Supreme Court's having become more conservative but partly as well to the increased efficacy of civil-rights litigation as an alternative method of enforcing the Fourth Amendment and other constitutional provisions--though I acknowledge that such litigation is unlikely to be effective against surveillance where the government refuses to disclose whose communications have been intercepted.
You are right that since terrorism is the use of violence for political ends, and since there is serious concern that extremist rhetoric by imams may encourage violence even if the rhetoric does not quite reach the level of incitement, a domestic intelligence agency would be investigating political activity, with the inevitable "chilling" effect. I cannot get excited about "chilling" the rhetoric of extremists who preach that Islamic suicide bombers go to heaven and their victims go to Hell and that it is the religious duty of every Muslim to wage holy war against the United States, Israel, and the West. But you're right that there's a history of abuse whenever a security agency interests itself in political activities.
But here it seems to me is the clincher in favor of a domestic intelligence agency. The FBI's able director, Robert Mueller, does not forswear engaging in the tactics that would be used by such an agency, including surveillance of radical imams. On the contrary, he claims that the FBI is engaging in those very tactics; that the reason we don't need a counterpart to MI5 or the Canadian Security Intelligence Service is that we have it already--in the "National Security Branch" of the FBI, an amalgam of the Bureau's intelligence, counterterrorism, and counterintelligence divisions. You may think such tactics inappropriate, but that is separate from the question whether, if they are to be used because of the terrorist threat, they should be used by the FBI or by a new agency. If as you suggest the FBI would be more squeamish in their use because of its preoccupation with prosecution and hence the exclusionary rule than a separate domestic intelligence agency would be, then that is another way of saying that the FBI will fail in its endeavor to duplicate the methods used by domestic intelligence agencies in the countries (virtually all but the United States) that have them. Those of us who think such tactics essential to combat the threat of terrorism want them to be utilized more efficiently than the FBI is likely to be able to do. It is not a compelling argument against a domestic intelligence agency that it would be more efficient than the FBI. And note that it would be a less intimidating body because it would not have the "badge and gun" culture of the FBI.
On the second issue, that of alternative security measures, I deliberately lumped together cost (and efficacy) and politics. There are two grounds for doubting whether, for example, inspecting all incoming cargoes for possible lethal contents is an alternative to, say, a domestic intelligence agency. The first ground of doubt is cost, and also whether an inspection program would be effective, given methods for shielding lethal cargoes from inspectors. The second, however, is that it may be politically impossible to obtain such a program. The political opposition may be base, selfish, corruptly motivated, short-sighted. But if it is overwhelming, that is a fact that has to be taken into account in deciding on whether to adopt an alternative measures that curtails civil liberties. Maybe civil liberties simply don't have the political support that shippers have. That would be a brute fact that the designer of an effective set of counterterrorist measures could not ignore.
I have stressed in my writings on intelligence its limitation as a protection against terrorist attacks. I therefore strongly favor hardening and response measures as alternative means of protection. We cannot hope to detect all surprise attacks in advance.
On the last issue, that of the gravity of the inroads made by the post-9/11 security measures into civil liberties, my conception of suspending habeas corpus--which I agree that only Congress can do lawfully and therefore that the Supreme Court's decision in Hamdi was sound--is to give the executive branch carte blanche so far as detention of terrorist suspects or other enemy suspects is concerned; it would have no enforceable obligation to provide even the rudiments of due process in military courts--it wouldn't have to create military courts. But I do not agree that the NSA surveillance is a severe curtailment of civil liberties. There is a big difference between snooping and imprisoning. I don't think people are as ignorant as you suppose of the amount of privacy that they are willingly giving up in order to enjoy the benefits of digitization. I think most people if fully informed would actually prefer a government agency to be monitoring their emails and phone calls for national security purposes than their employer to be monitoring their emails and phone calls for the employer's purposes, because the government agency's interest is much more limited.
Now you may respond that I am utterly naive in supposing that the NSA will limit its snooping to communications that its data-mining programs flag as being suspect. But we have such a leak-prone government that I think it much less likely than in the 1960s and early 1970s, before the Church and Pike hearings, that abuses of electronic surveillance can be concealed. But I have no objection to, and indeed favor, creating strong safeguards against such abuses. Congress wants more oversight; civil libertarians want more warrants. Both are mistaken, Congress because its oversight is intermittent and highly politicized, civil libertarians because warrants are designed for situations in which the government has a focused suspicion (probable cause). Warrants are fine for the monitoring of known or suspected terrorists; they are unusable for discovering who the terrorists are. That is the essential need at the moment, and it requires a much wider net with a much finer mesh.
We should be imaginative about alternative safeguards, such as (1) requiring that all interceptions that involve an intelligence officer actually reading or listening to the contents be reported periodically to an independent monitoring body (some counterpart to the General Accountability Office) together with the reason for and results of the interception; and (2) forbidding under stiff criminal penalties any use of intercepted communications for any purpose other than national security, including enforcement of ordinary criminal laws. So snooping would lead to imprisoning only if the interception produced or led to evidence of criminal terrorist activity.
No doubt these safeguards would in operation be imperfect, like everything in U.S. government. But the need is to balance privacy interests against national security. The recently foiled Canadian plot should remind us that our national security is endangered, justifying electronic surveillance designed to identify terrorists, not merely monitor their activities once they have been identified by other means.
Let me finally take this opportunity to correct an embarrassing mistake in my first posting: the subtitle of my new book Not a Suicide Pact is The Constitution in a Time of National Emergency, not in a Time of National Security--would this were a time in which we could be justified in feeling that the nation was secure.