National security federalism in the age of terror.

• Author: Waxman, Matthew C.

INTRODUCTION I. NATIONAL SECURITY FEDERALISM BEFORE AND AFTER 9/11 A. National Security Federalism and Domestic Intelligence in Historical Context 1. Vertical arrangement of national defense powers and the centralization of national security 2. Twentieth-century domestic intelligence abuses and oversight reform B. Post-9/11 National Security and Domestic Intelligence C. National Security and Local Police After 9/11 1. The context 2. Federal efforts to promote local national security activities II. VERTICAL RELATIONS AND NATIONAL SECURITY POWERS: LEADING ACCOUNTS AND THEIR LIMITS A. Top-Down Accounts 1. Hierarchical or principal-agent accounts 2. Trickle-down accounts B. Pushback Accounts C. Limits of Leading Accounts III. NATIONAL SECURITY FEDERALISM IN THE AGE OF TERROR A. Accountability and National Security Intelligence 1. Accountability and political processes 2. National security accountability as reasoned deliberation 3. Accountability oversight B. The Economics of National Security as Regulation 1. Intelligence law and policy as regulation 2. Uniformity and variation of intelligence law and policy C. The Future of National Security Federalism and Counterterrorism Intelligence CONCLUSION INTRODUCTION

National security law scholarship tends to focus on the balancing of security and liberty, and the overwhelming bulk of that scholarship is about such balancing at the federal level. That is, scholarship about national security institutional architecture focuses almost entirely on horizontal allocations, or relations among coordinate branches of the federal government, with an eye ultimately toward how those horizontal arrangements strike and enforce substantive liberty-security balances. (1)

This Article challenges that common focus by supplementing it with an account of the vertical axis: "national security federalism." It shows that the near-exclusive concentration of contemporary national security law scholarship on horizontal relationships among the federal branches is too limited because significant, tangible effects of security policies on liberty and many other interests today take place at the local level as a result of actions by local government actors. (2) Consider, for example, intelligence gathering related to combating terrorism. There are more than 700,000 local police officers from about 17,000 state and local law enforcement agencies who may conduct relevant activities (3) such as surveillance, profiling-based investigation, and data collection and sharing. (4) If an individual is being watched as a potential terrorism threat because of his appearance, it may be a local officer watching. (5) If a government agent is looking around a mosque and asking questions of members, it may be a local cop. (6) If data are being mined for suspicious patterns, local officials may have collected and passed on some of that data. (7) Most of this activity is done in the service of broad local law enforcement and policing mandates. Critically, however, it also contributes to a national security policy principally led by the federal government. (8)

These activities, moreover, are governed by a complex web of law: not just federal law, but also state statutes and state constitutional doctrine, municipal legislation and regulations, judicial consent decrees, and state and local administrative guidelines. For example, some municipal police departments' investigatory powers are limited by state law and some by locally set administrative guidelines. (9) Those investigatory powers are overseen by a variety of institutional mechanisms, including state and local legislatures, agency oversight boards, external audits, and courts. (10) Local law enforcement agencies vary in how they codify and apply guidelines to regulate collection, use, dissemination, and retention of data related to counterterrorism. (11) Most, but not all, states have electronic surveillance statutes, covering different types of communications and regulated by different standards and processes. (12) And racial and ethnic profiling is regulated by states and many local governments through a combination of legislation, court decisions, and administrative guidelines. (13)

This is not by any means to deny the predominance of federal national security law and policy, on account of the massive resources and capabilities of the federal government and the primacy of federal law in this context. Rather, it is to point out that subfederal law and institutions in many cases affect daily lives of individuals more directly and in different ways across jurisdictions. The result is an uneven, textured legal and policy landscape with regard to national security.

The federalism frame emphasizes vertical relations and institutional arrangements between federal and state or local governments for promoting and mediating policy values over the long term, and produces the uneven, textured legal and policy landscape. I argue that this frame yields insights to guide refinement of the emergent national security intelligence architecture. I emphasize two important values that have been neglected or undertheorized in existing scholarship with respect to local national security functions: (1) accountability and the way vertical intergovernmental arrangements enhance or degrade it, and (2) efficiency and the way those arrangements promote public policy effectiveness. In those respects my account reveals important potential benefits of our shared federal-local national security system, and offers ways to enhance these values, especially if terrorism threats evolve to include a greater domestic component.

I assume throughout this Article that state and local governments will continue to play a significant role in national security intelligence. Not only is this a political, practical, and historically contingent reality, but harnessing state and local institutions for national security is needed to address parts of the national security challenge for which those institutions are much better suited than the federal government could ever be.

A major theme running through the Article, however, is that to the extent state and local governments continue playing a major role in the emergent national security architecture, such that some national security intelligence will be decentralized, there remains a large question as to what degree these functions will be localized. By "decentralized" I mean a managerial concept, whereby policy is set at the top but executed at a local level with some limited discretion about how best to do so. Even countries with powerful national police forces like France and Israel may delegate implementation of some counterterrorism intelligence or investigation functions to lower levels of command. (14) By "localized," I emphasize instead an agenda-setting role in law and policy, whereby state or local entities need not accept the federal policy entirely but are allowed some room to vary that policy as they choose. To some extent, distinctions between decentralization and localization (15) blur in the national security intelligence context, because so much of the policy is implementation discretion: we care a lot--from both a legal and policy perspective--about the details of how, to what extent, and with what sort of oversight government agencies and agents collect, analyze, and share information. Nevertheless, the analytic distinction is useful in thinking about those important legal and policy decisions that could be taken at several levels of government--such as how widely to cast the government's authority to collect and compile information on individuals or groups, or how tightly to constrain and monitor the government's authority to comb through and retain data on individuals.

In that light, a descriptive goal of this Article is to show that significant national security localism continues to exist in the United States, despite some countervailing pressures, and to help explain why and how it operates. A prescriptive goal is to better understand in what specific contexts such localism should be celebrated, and how vertical intergovernmental relations might better be structured to harness it in advancing simultaneously a range of policy priorities. This Article does not focus on the first-order question of whether national security federalism or localism would be optimal compared to sweepingly different constitutional structures if we were drafting from scratch, such as radically centralizing police powers related to national security. Instead, this Article accepts some vertical division of power as given in the United States--as historically determined to a large degree by foundational constitutional compromises, even if also driven by some contemporary policy imperatives--and asks how we can make the most of that basic structure. Much more empirical research in this area is needed, (16) so the specific policy conclusions are necessarily tentative, but this Article can help guide further work.

In terms of substantive focus, this Article homes in on intelligence issues, which are just one element of national security policy, and puts aside some other significant elements such as disaster response. Intelligence is an area in which federal and state or local governments have significant resources and legal powers that raise especially salient civil liberties concerns and that overlap with other core local government responsibilities. While those factors make intelligence a fruitful case study of national security federalism generally, they also give rise to some distinct intergovernmental dynamics. Moreover, whereas most federalism discussion focuses on federal-state relations, (17) this Article focuses largely on federal-local relations and mostly groups state and local governments together. This is not to deny important differences in the way state and local governments may view national security matters or the public functions that...