{\rtf1\ansi \deflang1033\deff0{\fonttbl {\f0\froman \fcharset0 \fprq2 Times New Roman;}}{\colortbl;\red0\green0\blue0;\red0\green0\blue0;} {\stylesheet{\fs20 \snext0 Normal;} }{\info{\author schd2}{\operator schd2} } \ftnrstcont\ftnrstcont\notabind\margl1440\margr1440\hyphhotz936\enddoc\fet2\ftnrstcont\aftnnar\viewkind1\lytprtmet\subfontbysize \sectd \sbknone\pgndec\headery1440\footery1440 {\footer { \posxc\nowrap \plain \fs24 {\field{\*\fldinst { PAGE }}}\par} \par} {\*\pnseclvl1\pndec\pnstart1{\pntxta .}} {\*\pnseclvl2\pnlcltr\pnstart1{\pntxta .}} {\*\pnseclvl3\pnlcrm\pnstart1{\pntxta .}} {\*\pnseclvl4\pndec\pnstart1{\pntxtb (}{\pntxta )}} {\*\pnseclvl5\pnlcltr\pnstart1{\pntxtb (}{\pntxta )}} {\*\pnseclvl6\pnlcrm\pnstart1{\pntxtb (}{\pntxta )}} {\*\pnseclvl7\pndec\pnstart1{\pntxta .}} {\*\pnseclvl8\pnlcltr\pnstart1{\pntxta .}} {\*\pnseclvl9\pnlcrm\pnstart1} {\field{\*\fldinst {\lang4105 SEQ CHAPTER \\h \\r 1}}{\fldrslt }}\pard \fs24\qc {\plain \fs24 \b Federalism in Comparative Institutional Perspective: \par }{\plain \fs24 \b the Case of Preemption Doctrine}{\plain \fs24 \par }{\plain \fs24 \par }{\plain \fs24 by David S. Schwartz\par }{\plain \fs24 Assistant Professor of Law\par }{\plain \fs24 University of Wisconsin Law School\par }{\plain \fs24 \par }{\plain \fs24 \par }\pard \fs24 {\plain \fs24 \par }\pard \fs24\qc\sl480\slmult1 {\plain \fs24 \b {\tc {Introduction}}}{\plain \fs24 \par }\pard \fs24\sl480\slmult1 {\plain \fs24 \tab Federalism is one of the fundamental institutional choice questions in our constitutional \softline scheme; it is the allocation of power between federal and state sovereignties. Preemption \softline doctrine is a key component of federalism, because it guides courts in deciding whether a federal \softline or state regulatory scheme will control a particular activity, and thus allocates power between \softline federal and state regulators.\par }{\plain \fs24 \tab Consider the following four cases. The voters of California pass a referendum that \softline legalizes cultivation and possession of marijuana for medical purposes, but this {\u8220\'93}medical \softline marijuana{\u8221\'94} experiment is for all practical purposes nullified by federal anti-drug laws.{}{\plain \fs24 \super 1{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 1}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Oakland Cannabis Buyers Cooperative v. United States, 532 U.S. 483 (2001).}}} }{\plain \fs24 Several \softline states seek to regulate regional air pollution problems within the state by controlling vehicle \softline emissions, but the laws are preempted by a provision in the federal Clean Air Act designed to \softline lessen the regulatory compliance burden on automobile manufacturers.{}{\plain \fs24 \super 2{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 2}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Engine Mfrs. Assn v. South Coast Air Quality Mgmt. Dist., 124 S. Ct. 1756 (2004).}}} }{\plain \fs24 A state consumer \softline protection law permits a class action on behalf of victims of widespread small-dollar-amount \softline legal violations, but a federal law may give the lender the right to contract out of class actions and thereby effectively insulate itself from liability.{}{\plain \fs24 \super 3{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 3}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Green Tree Financial Corp. v. Bazzle, 123 S. Ct. 2402 (2003).}}} }{\plain \fs24 State tort law deems a 1987 model car \softline manufactured without an airbag to be a defective product, but that law is nullified by a federal \softline safety regulation issued by the Secretary of Transportation.{}{\plain \fs24 \super 4{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 4}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Geier v. American Honda Motor Corp., 529 U.S. 861 (2000).}}} }{\plain \fs24 Each of these decisions was non-unanimous on key points, and, in my view, probably wrongly decided.\par }{\plain \fs24 \tab My starting point is an intuition that there is currently a {\u8220\'93}preemption problem{\u8221\'94} in law and \softline public policy. My stronger intuitive claim is that there is {\u8220\'93}too much{\u8221\'94} federal preemption of state \softline law. Numerous beneficial state regulations falling well within the states{\u8217\'92} constitutional authority \softline are nullified by federal law; and even if one disagrees about the beneficial character of the state \softline laws I value, there are values inherent in federalism itself {\u8211\'96} understood as a robust degree of state \softline legislative autonomy {\u8211\'96} that are undermined by preemption doctrine. A weaker intuitive claim is \softline that preemption doctrine is {\u8220\'93}a mess.{\u8221\'94}{}{\plain \fs24 \super 5{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 5}{}{\plain \fs24 }}} }{\plain \fs24 Courts{\u8217\'92} institutional choices in preemption cases are \softline guided by a thin veneer of rules covering broad judicial discretion to enact the judges{\u8217\'92} own \softline policy preferences.\par }{\plain \fs24 \tab This paper examines whether comparative institutional analysis can be employed to \softline improve judicial performance in preemption decisions. In Part I, I consider whether there is in \softline fact a {\u8220\'93}preemption problem{\u8221\'94} {\u8211\'96} deficient institutional performance by courts in preemption cases. \softline While in important respects, it is difficult to reach a firm conclusion, I argue that federalism \softline principles should receive weight that they often do not receive from courts considering this \softline question. In Part II, I recast the basic issue of preemption in institutional choice terms. Finally, in Part III, I examine particular categories of preemption cases and consider to what extent \softline comparative institutional analysis can be mapped onto existing legal doctrine. I argue that while \softline courts may be able to make meaningful comparisons between state and federal political processes \softline in some preemption cases, their ability to choose between political and market processes is too \softline flawed to take such determinations away from the political process. \par }{\plain \fs24 \par }\pard \fs24\qc\sl480\slmult1 {\plain \fs24 \b {\tc {I. Preemption and Federalism}}}{\plain \fs24 \par }\pard \fs24\li720\sl480\slmult1 {\plain \fs24 }{\plain \fs24 \b {\tc {A. The Values of Federalism} \tcl2}}{\plain \fs24 \par }\pard \fs24\sl480\slmult1 {\plain \fs24 \tab The doctrine of federalism itself embodies choices emerging from comparative \softline institutional analysis. The apportionment of authority by the Constitution reflects institutional \softline choices no doubt guided by comparative judgments of which institution, federal or state, would \softline more beneficially handle the issues. In general, the Constitution{\u8217\'92}s assignment of paramount, but \softline limited powers to the federal government reflects a judgment that issues of war and foreign \softline affairs, interstate coordination and commercial development, and national unification are better \softline handled at the national level.{}{\plain \fs24 \super 6{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 6}{}{\plain \fs24 Cites __. Thus, for example, locating in the federal government the power to declare war \softline and make treaties reflects that judgment that it would benefit the nation to speak and act as a \softline unified whole, rather than as a diverse and possibly conflicting set of sovereignties. }}} }{\plain \fs24 The residuum of sovereignty {\u8220\'93}not delegated by the Constitution to \softline the United States, nor prohibited by it to the states,{\u8221\'94} remains with the states. \par }{\plain \fs24 \tab It is generally agreed that the overarching historical trend in the structure of federalism \softline has been toward greater federal power at the expense of the states. Public policy arguments \softline about federalism are arguments over whether legal doctrine should be used as a constraint on the \softline expansion of federal power. Proponents of national power have advanced the so-called {\u8220\'93}political safeguards of federalism{\u8221\'94} theory, which boils down to an argument that whatever power is to be \softline retained by the states in the federal system should be nothing more nor less than what the federal \softline political process allows it.{}{\plain \fs24 \super 7{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 7}{}{\plain \fs24 This literature is summarized and critiqued in Saikrishna B. Prakash and John C. Yoo, \softline }{\plain \fs24 \i The Puzzling Persistence of Process Based Federalism Theories,}{\plain \fs24 79 Tex. L. Rev. 1459 (2001).}}} }{\plain \fs24 Under this theory, the institutional choice questions of federalism \softline presented to courts are of little interest: courts should simply defer to determinations made by \softline federal lawmakers.\par }{\plain \fs24 \tab To argue that federalism questions are significant institutional choice questions, assumes \softline that policy benefits flow from the certain comparative advantages state sovereignty brings to the \softline federal model.{}{\plain \fs24 \super 8{\footnote \pard \fs24 {\ul0 \tab }{\plain \fs24 \super 8}{}{\plain \fs24 While {\u8220\'93}originalists{\u8221\'94} may argue that the preservation of federalist structure in the text of \softline the constitution is enough without more to support vigorous protection of the states{\u8217\'92} role, }{\plain \fs24 \i see, \softline e.g., }{\plain \fs24 Michael W. McConnell, }{\plain \fs24 \i Federalism: Evaluating the Founders{\u8217\'92} Design}{\plain \fs24 , 54 U. Chi. L. Rev. \softline 1484, 1491 (1988) (criticizing Raoul Berger{\u8217\'92}s }{\plain \fs24 \i Fderalism: the Founder{\u8217\'92}s Design}{\plain \fs24 for {\u8220\'93}ignor[ing] \softline the intellectual case for federalism{\u8221\'94})}{\plain \fs24 , that is probably not the case. Constitutional doctrines \softline lacking in practical value have a tendency to wither, and the continued existence of a meaningful \softline federalism doctrine reflects underlying normative values. The Court has shown an awareness to \softline make such normative arguments. The most ambitious normative justification for federalism in a \softline Supreme Court decision in modern times is Justice O{\u8217\'92}Connor{\u8217\'92}s opinion in }{\plain \fs24 \i Gregory v. Ashcroft}{\plain \fs24 , \softline 501 U.S. __, which has been frequently echoed in whole or in part in most of the subsequent \softline federalism revival cases. See New York v. United States (citing Gregory); United States v. Lopez \softline (same); }{\plain \fs24 \i id}{\plain \fs24 . (Kennedy, J., concurring) (same). \par }\pard \fs24\li1440\ri720 {\plain \fs24 This federalist structure of joint sovereigns preserves to the people \softline numerous advantages. It assures a decentralized government that will be \softline more sensitive to the diverse needs of a heterogenous society; it increases \softline opportunity for citizen involvement in democratic processes; it allows for \softline more innovation and experimentation in government; and it makes \softline government more responsive by putting the States in competition for a \softline mobile citizenry}{\plain \fs24 . 501 U.S. at 458.\par }\pard \fs24\sa240 {\plain \fs24 At greater length, the }{\plain \fs24 \i Gregory}{\plain \fs24 opinion asserts that {\u8220\'93}perhaps the principal benefit of the federalist \softline system is a check on abuses of government power.{\u8221\'94}\ul0 \tab }}} }{\plain \fs24 To the extent that {\u8220\'93}individual liberty{\u8221\'94} is promoted by fewer layers of {\u8220\'93}police{\u8221\'94} \softline power regulation, and by having such {\u8220\'93}police{\u8221\'94} regulations enacted by smaller units of government closer to the people (both of which are controversial propositions, to be sure), then \softline federalism limits that restrict general police powers to the states could be said to promote \softline individual liberty. Preservation of state autonomy may also protect a second aspect of {\u8220\'93}liberty,{\u8221\'94} \softline as the framers used that term: the power of a citizen to participate in republican government.{}{\plain \fs24 \super 9{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 9}{}{\plain \fs24 Cites}}} }{\plain \fs24 \softline Representativeness is arguably stronger at smaller levels of government,{}{\plain \fs24 \super 10{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 10}{}{\plain \fs24 This, too, is an empirically contestable assertion. See, e.g., Erwin Chemerinsky, }{\plain \fs24 \i the \softline Values of Federalism}{\plain \fs24 , 47 Fla. L. Rev. 499, 528 (1995). }}} }{\plain \fs24 and state autonomy \softline fosters this form of liberty insofar as the states are players in national politics, as lobbyists or \softline litigants, or as the source of countervailing political power through opposition or third party \softline politics.{}{\plain \fs24 \super 11{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 11}{}{\plain \fs24 Deborah Jones Merritt, }{\plain \fs24 \i The Guarantee Clause and State Autonomy: Federalism for a \softline Third Century}{\plain \fs24 , 88 Colum. L. Rev. 1, 5-7 (1988).}}} }{\plain \fs24 \par }{\plain \fs24 \tab State autonomy also offers diversity and experimentation in social policy. As has often \softline been observed, the ability of residents of the United States to relocate from state to state makes it \softline easier to avoid oppressive measures enacted at the state level.{}{\plain \fs24 \super 12{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 12}{}{\plain \fs24 See, e.g., McConnell, supra note __, at 1503-04.}}} }{\plain \fs24 Even where the state policies in \softline question are not oppressive, {\u8220\'93}[i]n a heterogenous society, a federal system can better satisfy \softline political preferences and economic needs, especially over time, than can a simple unitary \softline government.{\u8221\'94}{}{\plain \fs24 \super 13{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 13}{}{\plain \fs24 Bednar and Eskridge, supra note __, at 1467-68.}}} }{\plain \fs24 Finally, as Justice Brandeis famously put it, states can serve as {\u8220\'93}laboratories for \softline experimentation{\u8221\'94} in social policy.{}{\plain \fs24 \super 14{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 14}{}{\plain \fs24 New State Ice Co. v. Liebmann, 285 U.S., 262, 310-11 (1932) (Brandeis, J., dissenting).}}} }{\plain \fs24 Even in areas of concurrent legislative authority, the laboratories idea implies, there may be issues where the federal government should refrain from \softline occupying the legislative field and allow the best ideas to be tried on the state level.\par }{\plain \fs24 \par }{\plain \fs24 \tab }{\plain \fs24 \b {\tc {B. Preemption Doctrine} \tcl2}}{\plain \fs24 \par }{\plain \fs24 \ul0 \tab The general law of preemption can be briefly summarized as follows. {\u8220\'93}[U]nder the \softline Supremacy Clause, from which our pre-emption doctrine is derived, {\u8216\'91}any state law, however \softline clearly within a State{\u8217\'92}s acknowledged power, which interferes with or is contrary to federal law, \softline must yield.{\u8217\'92}{\u8221\'94}{}{\plain \fs24 \super 15{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 15}{}{\plain \fs24 Gade v. National Solid Waste Management Assn., 505 U.S. 88, 108 (1992). }}} }{\plain \fs24 Assuming that legislation is within Congress{\u8217\'92} enumerated powers, its power to \softline displace state law as part of that legislation is {\u8220\'93}plenary.{\u8221\'94}{}{\plain \fs24 \super 16{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 16}{}{\plain \fs24 }{\plain \fs24 \i E.g.,}{\plain \fs24 United States v. Lopez, 514 U.S. 549, 670 (1995) (Souter, J., dissenting).}}} }{\plain \fs24 The Supremacy Clause means in part \softline that federal law is {\u8220\'93}in state law,{\u8221\'94} just as much a part of the corpus of law regulating the citizens \softline within a state as that state{\u8217\'92}s law.{}{\plain \fs24 \super 17{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 17}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Testa v. Katt, 330 U.S. 386 (1947); Evan H. Caminker, }{\plain \fs24 \i State Sovereignty and \softline Subordinacy: May Congress Commandeer State Officers To Implement Federal Law?}{\plain \fs24 , 95 \softline Colum. L. Rev. 1001, 1036-38 (1995)}{\plain \fs24 . }}} }{\plain \fs24 A judicial decision applying a federal law to preempt a state \softline law is a form of constitutional adjudication that nullifies a state law, in effect striking down the \softline state law as {\u8220\'93}violat[ing] the Supremacy Clause.{\u8221\'94}{}{\plain \fs24 \super 18{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 18}{}{\plain \fs24 Southland Corp. v. Keating, 465 U.S. 1, 12 (1984).}}} }{\plain \fs24 \par }{\plain \fs24 \tab The black letter preemption doctrine tells us that preemption, while constitutionally \softline based in the Supremacy Clause, is an issue of Congressional intent.{}{\plain \fs24 \super 19{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 19}{}{\plain \fs24 Gade, 505 U.S. at 108.}}} }{\plain \fs24 }{\plain \fs24 Courts have organized the \softline intent question into {\u8220\'93}express{\u8221\'94} and {\u8220\'93}implied{\u8221\'94} preemption. Express preemption cases involve the interpretation of federal statutes with an express provision dealing with preemption.{}{\plain \fs24 \super 20{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 20}{}{\plain \fs24 }{\plain \fs24 \i See, e.g., }{\plain \fs24 Employee Retirement Income Security Act, 29 U.S.C. {\u167\'a7} 29 U.S.C. {\u167\'a7} 1144(a) \softline (providing that ERISA {\u8220\'93}shall supersede any and all State laws insofar as they may now or \softline hereafter relate to any employee benefit plan{\u8221\'94}).}}} }{\plain \fs24 }{\plain \fs24 Implied \softline preemption cases involve statutes with no such preemptive language, and try to determine the \softline preemptive intent of Congress in other provisions of the statute or its legislative history.\par }{\plain \fs24 \tab Implied preemption is divided, by the doctrine, into further subcategories. {\u8220\'93}Field{\u8221\'94} \softline preemption is a finding by the Court that Congress intended to occupy a particular field of \softline legislation to the exclusion of the states; any state law in the field will be preempted irrespective \softline of whether the law is consistent with federal law.{}{\plain \fs24 \super 21{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 21}{}{\plain \fs24 }{\plain \fs24 \i See, e.g., }{\plain \fs24 English v. General Elec. Co., 492 U.S. 72, 79 (1990). }}} }{\plain \fs24 }{\plain \fs24 Under {\u8220\'93}conflict{\u8221\'94} preemption, in contrast, a \softline state law is void only if it conflicts with federal law.{}{\plain \fs24 \super 22{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 22}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Caleb Nelson, }{\plain \fs24 \i Preemption}{\plain \fs24 , 86 Va. L. Rev. 225 (2000). }}} }{\plain \fs24 }{\plain \fs24 The notion of a conflict has been defined \softline by two further categories. If the commands of federal and state law are so in conflict that it is \softline impossible to comply with both, then the state law is void.{}{\plain \fs24 \super 23{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 23}{}{\plain \fs24 }{\plain \fs24 \i See, e.g., }{\plain \fs24 Florida Lime Growers v. Paul, 373 U.S. 132, 142-43 (1963). }}} }{\plain \fs24 }{\plain \fs24 Finally, a conflict will also be found \softline if the state law {\u8220\'93}stands as an obstacle to the accomplishment and execution of the full purposes \softline of Congress.{\u8221\'94}{}{\plain \fs24 \super 24{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 24}{}{\plain \fs24 E.g., Hines v. Davidowitz, 312 U.S. 52, 67 (1941).}}} }{\plain \fs24 \par }{\plain \fs24 \tab Federal preemption of state law is constrained by three federalism-based doctrines of \softline statutory interpretation. {\u8220\'93}If Congress intends to alter the usual constitutional balance between the \softline States and the Federal Government, it must make its intention unmistakably clear in the language of the statute.{\u8221\'94}{}{\plain \fs24 \super 25{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 25}{}{\plain \fs24 }{\plain \fs24 \i Gregory v. Ashcroft}{\plain \fs24 , 501 U.S. at 460 (internal quotations omitted).}}} }{\plain \fs24 A subset of this federalism-based {\u8220\'93}clear statement{\u8221\'94} rule is the long-established \softline presumption against preemption: {\u8220\'93}{\u8216\'91}where ... the field which congress is said to have preempted \softline includes areas that have been traditionally occupied by the States,{\u8217\'92} congressional intent to \softline supersede state laws must be {\u8216\'91}clear and manifest.{\u8217\'92}{\u8221\'94}{}{\plain \fs24 \super 26{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 26}{}{\plain \fs24 }{\plain \fs24 \i E.g., }{\plain \fs24 \i English v. General Electric Co.}{\plain \fs24 , 496 U.S. 72, 78-79 (1990)}}} }{\plain \fs24 }{\plain \fs24 \i }{\plain \fs24 These rules are connected to the doctrine of \softline constitutional avoidance, which holds that {\u8220\'93}where an otherwise acceptable construction of a \softline statute would raise serious constitutional problems, the Court will construe the statute to avoid \softline such problems unless such construction is plainly contrary to the intent of Congress.{\u8221\'94}{}{\plain \fs24 \super 27{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 27}{}{\plain \fs24 Solid Waste Agency of Northern Cook County v. Army Corps of Engineers, 531 U.S. \softline 159, 173 (2001) (internal quotations omitted); }{\plain \fs24 \i accord}{\plain \fs24 Ashwander v. Tennessee Valley Authority, \softline 297 U.S. 288, 347 (1936) (Brandeis, J., concurring) ({\u8220\'93}if a case can be decided on either of two \softline grounds, one involving a constitutional question, the other a question of statutory construction or \softline general law, the Court will decide only the latter{\u8221\'94}).}}} }{\plain \fs24 \par }{\plain \fs24 \b \par }{\plain \fs24 \b {\tc {\tab C. Is There a {\u8220\'93}Preemption Problem{\u8221\'94}?} \tcl2}\par }{\plain \fs24 \ul0 \tab \ul0 \tab }{\plain \fs24 \b {\tc {1. Too Much Preemption?} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab The question whether or not there is too much preemption is difficult, if not impossible to \softline answer. On the one hand, a lot of state laws are preempted. Preemption {\u8220\'93}is almost certainly the \softline most frequently used doctrine of constitutional law in practice.{\u8221\'94}{}{\plain \fs24 \super 28{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 28}{}{\plain \fs24 Stephen A. Gardbaum, }{\plain \fs24 \i The Nature of Preemption}{\plain \fs24 , 79 Corn. L. Rev. 767, 768 (1994).}}} }{\plain \fs24 From the beginning of the \softline 1991-92 term to the 2001-02 term, the Supreme Court struck down 10 federal laws in the name of federalism, giving rise to talk of a {\u8220\'93}federalism revolution.{\u8221\'94}{}{\plain \fs24 \super 29{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 29}{}{\plain \fs24 Richard H. Fallon, Jr., }{\plain \fs24 \i The Conservative Paths of the Rehnquist Court{\u8217\'92}s Federalism \softline Decisions}{\plain \fs24 , 69 Chi. L. Rev. 429, __ (2002).}}} }{\plain \fs24 During that same period, the \softline court struck down 20 state laws in preemption cases; by comparison, the Court struck down 40 \softline state laws on all other constitutional grounds combined.{}{\plain \fs24 \super 30{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 30}{}{\plain \fs24 David S. Schwartz, }{\plain \fs24 \i Federalism and Preemption}{\plain \fs24 (manuscript).}}} }{\plain \fs24 \par }{\plain \fs24 \tab These numbers are far from the whole story, because state laws are held preempted in \softline numerous state and lower federal court cases. When a federal or state law is struck down on \softline federal constitutional grounds, the Supreme Court reviews the case in a high proportion of cases. \softline However, preemption cases are viewed as statutory interpretation cases {\u8211\'96} the {\u8220\'93}ordinary diet of the \softline law{\u8221\'94} in Justice Breyer{\u8217\'92}s words{}{\plain \fs24 \super 31{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 31}{}{\plain \fs24 Egelhoff v. Egelhoff, 532 U.S. 141, 160 (2001) (Breyer, J., dissenting).}}} }{\plain \fs24 {\u8211\'96} and are much less likely to be reviewed by the Supreme Court. \softline Thus the number of preemption decisions by the Supreme Court is the tip of a much larger \softline iceberg than other constitutional adjudication.\par }{\plain \fs24 \tab To get a picture of what may be going on with lower-court preemption decisions, let{\u8217\'92}s \softline look at just one federal statute, the Federal Arbitration Act. Between January 2002 and April \softline 2004, over 50 state laws were held preempted by this one federal statute in state and federal \softline decisions.{}{\plain \fs24 \super 32{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 32}{}{\plain \fs24 David S. Schwartz, }{\plain \fs24 \i State Judges as Guardians of Federalism: Resisting the Federal \softline Arbitration Act{\u8217\'92}s Encroachment on State Law}{\plain \fs24 , 16 Wash. U. J. Law & Pol. __, App. A \softline (forthcoming 2004) (hereinafter {\u8220\'93}Schwartz, }{\plain \fs24 \i State Judges}{\plain \fs24 {\u8221\'94}); }{\plain \fs24 \i see also}{\plain \fs24 Schwartz}{\plain \fs24 \i , Correcting \softline Federalism Mistakes, supra }{\plain \fs24 note __, at __.}}} }{\plain \fs24 Moreover, an individual preemption ruling does not tell the whole story, since the \softline preemption of one state law by a binding precedent will effectively preempt similar laws. The Supreme Court{\u8217\'92}s decision in }{\plain \fs24 \i Southland Corp. v. Keating}{\plain \fs24 ,{}{\plain \fs24 \super 33{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 33}{}{\plain \fs24 465 U.S. 1 (1984).}}} }{\plain \fs24 held that the FAA preempted certain \softline generic statutory language that appeared in at least 30 state laws other than the one before the \softline court.{}{\plain \fs24 \super 34{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 34}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Schwartz, }{\plain \fs24 \i State Judges}{\plain \fs24 , App. B.}}} }{\plain \fs24 This suggests that hundreds of state laws are held preempted each year.\par }{\plain \fs24 \tab While that sounds like an awful lot of preemption, there{\u8217\'92}s another way to look at the \softline matter. There are an awful lot of federal and state laws. Given the constitutional supremacy of \softline federal law, we would expect lots of state laws to be preempted, even if preemption doctrine \softline balanced state and federal interests perfectly. At the same time, a lot of state laws are not \softline preempted. A few nooks and crannies of law still remain outside the federal commerce power, \softline such as family law and ... well, family law.{}{\plain \fs24 \super 35{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 35}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 United States v. Morrison, 529 U.S. 598, 619 (2000); Schwartz, }{\plain \fs24 \i Federalism and \softline Preemption}{\plain \fs24 .}}} }{\plain \fs24 Even where state and federal regulatory authority is \softline concurrent, Congress leaves room for the operation of state law. Thus, in 17 out of the 37 \softline preemption cases decided by the Supreme Court in the 1991-2002 period, state laws were held \softline not preempted. \par }{\plain \fs24 \tab Counting the numbers of laws state laws preserved and preempted, even if it could be \softline done, may not really tell us whether there is {\u8220\'93}too much{\u8221\'94} preemption. Shifting the focus to the \softline values of federalism, can it be said that these values are being harmed or undermined by federal \softline preemption? Plainly, preemption of state law moves regulatory control from states to the federal \softline government, and therefore tends in the direction of undercutting the values of federalism, all of \softline which depend upon the continued existence of a fairly robust system of autonomous state legislatures. Yet, as a practical matter, it seems unlikely that these virtues of federalism {\u8211\'96} state \softline experimentation, for instance {\u8211\'96} could be entirely eliminated. Although under the current view of \softline preemption and federal supremacy, Congress could }{\plain \fs24 \i in theory}{\plain \fs24 displace the states from virtually \softline every field of legal regulation except... well, family law, it is hard to imagine Congress doing so. \softline Even if it wanted to, the federal government lacks the resources to take over all the fields of \softline regulation in which states engage heavily, such as the law of crimes, torts and contracts, to name \softline a few. Nevertheless, might there be an optimal level of state experimentation which is being \softline suppressed by excessive preemption? That question seems difficult to answer, even if there were \softline a consensus on the desirability of state law experimentation, because it seems impossible to \softline quantify.\par }{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {2. Is Preemption Doctrine a Mess?} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab So if the answer to the {\u8220\'93}too much preemption{\u8221\'94} question is indeterminate, perhaps there is \softline a preemption problem due to the {\u8220\'93}mess{\u8221\'94} of preemption doctrine.{}{\plain \fs24 \super 36{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 36}{}{\plain \fs24 }{\plain \fs24 \i See, e.g.,}{\plain \fs24 Calvin Massey, }{\plain \fs24 \i Federalism and the Rehnquist Court}{\plain \fs24 , 53 Hastings L. J. 431, \softline 508 (2002); Viet D. Dinh, }{\plain \fs24 \i Regulatory Compliance as a Defense to Products Liability: \softline Reassessing the Law of Preemption}{\plain \fs24 , 88 Geo. L. Rev. 2085, 2085 (2000).}}} }{\plain \fs24 Supreme Court justices have \softline been accused at times of being result-oriented, unprincipled or opportunistic in using federalism \softline principles, and this is nowhere more apparent than in preemption cases. The five justices who \softline have led the charge in the {\u8220\'93}federalism revival{\u8221\'94} (Rehnquist, O{\u8217\'92}Connor, Scalia, Kennedy and \softline Thomas) should be expected to incline toward the side of state autonomy and against federal \softline preemption, if guided by the federalism principles they espouse in the other federalism cases. \softline Yet these justices, during the {\u8220\'93}federalism revival{\u8221\'94} from 1991-2002 voted for preemption over 58% of the time, while the {\u8220\'93}liberal{\u8221\'94} justices who tended to vote for federal law in the other \softline federalism cases (Stevens, Souter, Ginsburg and Breyer), voted for preemption less than 37% of \softline the time.{}{\plain \fs24 \super 37{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 37}{}{\plain \fs24 Schwartz, }{\plain \fs24 \i Federalism and Preemption}{\plain \fs24 .}}} }{\plain \fs24 One might defend the conservative justices by saying that preemption cases are \softline statutory interpretation cases, and that the Supremacy Clause gives Congress, not the Court, the \softline choice whether or not to preempt state law. Justice Scalia has explicitly taken this position, \softline rejecting the argument for a federalism-based {\u8220\'93}presumption against preemption,{\u8221\'94} arguing instead \softline that {\u8220\'93}ordinary{\u8221\'94} rules of statutory interpretation must control.{}{\plain \fs24 \super 38{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 38}{}{\plain \fs24 Cipollone v. Liggett Group, 505 U.S. 504, 544-48 (Scalia, J., concurring in the \softline judgment in part and dissenting in part).}}} }{\plain \fs24 Yet this argument is \softline disingenuous, since preemptive intent {\u8211\'96} even in express preemption cases {\u8211\'96} is often ambiguous. \softline If one can roughly equate the clearcut cases of Congressional intent with unanimous Supreme \softline Court decisions, and equate non-unanimous decisions with instances where the statutory issue \softline could be decided either way, then arguably there is more room to give weight to federalism \softline concerns and indulge a {\u8220\'93}presumption against preemption.{\u8221\'94} Yet here, the federalism revival \softline justices are twice as likely to vote for preemption. Finally, in non-unanimous cases where a \softline preemption ruling would impose a deregulatory, market-oriented outcome, the {\u8220\'93}pro-federalism{\u8221\'94} \softline justices voted for preemption nearly 73% of the time, while the pro-nationalism justices voted to \softline preempt less than 30% of the time.{}{\plain \fs24 \super 39{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 39}{}{\plain \fs24 Schwartz, }{\plain \fs24 \i Federalism and Preemption}{\plain \fs24 .}}} }{\plain \fs24 \par }{\plain \fs24 \tab Nor are these cases typically 5-4, }{\plain \fs24 \i Bush v. Gore}{\plain \fs24 splits either. The voting breakdowns in preemption cases are often surprising.{}{\plain \fs24 \super 40{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 40}{}{\plain \fs24 The numbers imply this. So, for example, the 73% preemption rate for conservative \softline justices in deregulation preemption cases implies that they have voted for regulation about 27% \softline of the time, often joining liberal justices in doing so.}}} }{\plain \fs24 The picture that emerges is one of doctrinal confusion, \softline apparent inconsistency in voting patterns and principles, and a tendency toward {\u8220\'93}result-orientation{\u8221\'94} with conservatives more than twice as likely to vote for deregulation. It might be \softline added that the {\u8220\'93}ordinary experience-proved principles of ... pre-emption{\u8221\'94}{}{\plain \fs24 \super 41{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 41}{}{\plain \fs24 Geier v. American Honda Motor Co., 529 U.S. 861, 874 (2000). }}} }{\plain \fs24 are so malleable and \softline inconsistently applied that they do no operate to keep preemption decisions tightly lashed to the \softline intentions of Congress. The justices, some much more than others, tend to be all over the map in \softline preemption cases, articulating general principles in one case and abandoning them in the next {\u8211\'96} \softline perhaps more than in other sorts of cases.{}{\plain \fs24 \super 42{\footnote \pard \fs24 {\ul0 \tab }{\plain \fs24 \super 42}{}{\plain \fs24 Not to pick on anyone in particular, but }{\plain \fs24 \i compare}{\plain \fs24 Gade v. National Solid Wastes \softline Management Assn., 505 US. 88, 110 (1992) (Kennedy, J., concurring) (criticizing the preemption \softline of state laws based on {\u8220\'93}[a] freewheeling judicial inquiry into whether [state law] is in tension \softline with federal objectives{\u8221\'94}), }{\plain \fs24 \i with}{\plain \fs24 Circuit City Stores v. Adams, (per Kennedy, J.) (expanding \softline preemptive scope of Federal Arbitration Act because {\u8220\'93}there are real benefits to the enforcement \softline of arbitration provisions{\u8221\'94} while limiting preemptive scope {\u8220\'93}would call into doubt the efficacy of \softline alternative dispute resolution procedures adopted by many of the Nation{\u8217\'92}s employers, in the \softline process undermining the FAA{\u8217\'92}s proarbitration purposes{\u8221\'94}). \par }\pard \fs24 {\plain \fs24 \tab Sometimes the inconsistency even appears in the same case. }{\plain \fs24 \i See}{\plain \fs24 Cipollone v. Liggett \softline Group, 505 U.S. 504, 544 (1992) (Scalia, J., concurring in the judgment in part and dissenting in \softline part) (contending that majority creates new rule that express preemption provisions are to be \softline narrowly construed, but that {\u8220\'93}the lifespan of this new rule may have been blessedly brief, \softline inasmuch as the opinion that gives it birth in Part I proceeds to ignore it in Part V{\u8221\'94}).\par }\pard \fs24\sa240 {\plain \fs24 \tab }}} }{\plain \fs24 \par }{\plain \fs24 \par }{\plain \fs24 \tab Does this doctrinal mess matter? From the vantage point of institutional choice {\u8211\'96} the \softline selection of federal or state regulatory power {\u8211\'96} so long as a critical mass of state autonomy remains, perhaps the particular outcomes within the run of preemption decisions don{\u8217\'92}t matter. \softline The Supreme Court held state laws preempted in about 54% of the cases. If that leaves enough \softline room for state experimentation, arguably a coin toss (with a coin just slightly weighted toward \softline federal law) would serve just as well {\u8211\'96} if not from the vantage point of predictability, then at least \softline from the vantage point of federalism. \par }{\plain \fs24 \tab In these respects, there may not be a preemption problem at all, and you have to be asking \softline yourself why you shouldn{\u8217\'92}t stop reading this article right now. But I believe there is a reason to \softline consider comparative institutional analysis, because the courts have to decide. Preemption cases \softline arise as choice of law questions {\u8211\'96} whether federal or state law provides the rule of decision {\u8211\'96} that \softline must be answered to decide concrete disputes. Where state law prima facie controls a dispute, \softline and a federal law arguably applies preemptively, the court must answer the question whether \softline federal law applies and, if so, does it preempt state law. Since the preemption question must be \softline decided to resolve the case, and assuming a coin toss is ruled out, there must be a reasoned basis \softline to decide. Since the decision is an institutional choice, comparative institutional analysis is \softline worth considering.}{\plain \fs24 \b \par }{\plain \fs24 \par }\pard \fs24\qc\sl480\slmult1 {\plain \fs24 \b {\tc {II. Preemption as Institutional Choice}{\plain \fs24 :}{\plain \fs24 \b Some Conceptual Problems}}}{\plain \fs24 \par }\pard \fs24\sl480\slmult1 {\plain \fs24 {\tc {\tab }{\plain \fs24 \b A. Terminology Issues \par }{\plain \fs24 \b } \tcl2}{\tc {\tab \tab 1. {\u8220\'93}First Order{\u8221\'94} and {\u8220\'93}Second Order{\u8221\'94} Determinations} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab Preemption rulings raise some of the same problems as cases of constitutional judicial \softline review. In generic comparative institutional analysis terms, a preemption ruling, like a \softline constitutional challenge to a statute, involves two levels of institutional choice. The basic substantive policy question might usefully be called the {\u8220\'93}first-order{\u8221\'94} institutional choice. Will a \softline 50-unit condo complex be built on a stretch of beachfront or will the beachfront be maintained in \softline its natural state for environmental reasons? If the decision is allocated to the market, then market \softline forces will determine whether the condo is built; if to the political process, perhaps a zoning \softline board or an environmental agency will have a veto on the a market-driven development decision. \softline The {\u8220\'93}second-order{\u8221\'94} institutional choice is the question {\u8220\'93}who decides who decides?{\u8221\'94} If the \softline zoning or environmental restriction is challenged as an unconstitutional taking, the court will \softline decide this second-order question.\par }{\plain \fs24 \tab The fact that a court decides a case presenting the second order decision does not \softline necessarily mean the court will allocate that decision to itself. Komesar suggests, and I agree for \softline institutional choice purposes, that genuinely deferential court decisions allocate power to the \softline institution deferred to; whereas decisions based on non-deferential {\u8211\'96} and necessarily more \softline involved and in-depth {\u8211\'96} allocate the decisionmaking authority to the court itself.{}{\plain \fs24 \super 43{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 43}{}{\plain \fs24 Neil K. Komesar, }{\plain \fs24 \i Laws Limits: the Rule of Law and the Supply and Demand of Rights}{\plain \fs24 \softline (2001); Neil K. Komesar, }{\plain \fs24 \i Imperfect Alternatives: Choosing Institutions in Law, Economics and \softline Public Policy}{\plain \fs24 (1994).}}} }{\plain \fs24 \par }{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {2. State Common Law in the Preemption Context} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab To simplify the discussion somewhat, this paper cuts one significant analytical corner. In \softline many instances, the state law at stake in a preemption case is judge-made or {\u8220\'93}common{\u8221\'94} law. For \softline purposes of this paper, however, I will treat such state law rules as analytically indistinct from \softline tort rules created by state legislative or administrative bodies, i.e., the state{\u8217\'92}s {\u8220\'93}political process.{\u8221\'94} \softline There may be some risk in undercutting my analysis, insofar as the judicial process that leads to, for instance, a tort liability rule may be very different from the state{\u8217\'92}s political process. I believe \softline this shorthand equation is justified, however. Substantive common law rules are uniformly \softline subject to legislative control, and in many states, traditional common law fields such as tort, \softline contract and property are interlaced with statutes, if not densely codified. I believe that trying to \softline distinguish a state statute, for instance, permitting punitive damages from a long-standing \softline common-law doctrine that is substantively identical will not change the comparative institutional \softline analysis for preemption purposes.\par }{\plain \fs24 \par }{\plain \fs24 {\tc {\tab }{\plain \fs24 \b B. Assessing Institutional Performance} \tcl2}}{\plain \fs24 \par }{\plain \fs24 \tab The notion of {\u8220\'93}market failure{\u8221\'94} is probably the best analyzed and understood of the idea of \par }{\plain \fs24 failures in the three major institutions analyzed by Komesar. For comparative institutional \softline analysis to work as a positive policy-making strategy, as opposed to a critique, there must be \softline workable understandings of institutional failure in the other institutions as well. \par }{\plain \fs24 \par }{\plain \fs24 {\tc {\tab \tab }{\plain \fs24 \b 1. Political Process Failure} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab Komesar has persuasively critiqued discussions of the political process that either fail to \softline analyze that institution{\u8217\'92}s performance at all, or that do so by relying on a one-force rather than a \softline two force model of politics.{}{\plain \fs24 \super 44{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 44}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Komesar, }{\plain \fs24 \i Imperfect Alternatives}{\plain \fs24 , at 29, 35-44, 213-15, (critizing John Rawls, }{\plain \fs24 \i A \softline Theory of Justice }{\plain \fs24 and Pigouvian welfare economics for failing to analyze political institutions); \softline }{\plain \fs24 \i id}{\plain \fs24 . at 213-15, 221-30 (criticizing John Hart Ely and Bruce Ackerman for failing to consider both \softline majoritarian and minoritarian bias).}}} }{\plain \fs24 A complete analysis of the political process must, at least, \softline acknowledge that process failures can be caused either by majoritarian bias (domination of a minority by the majority) or minoritarian bias (the overrepresentation of small groups of high per \softline capita stakes holders dominating the dispersed, low per capita stakes interests of the majority). \par }{\plain \fs24 \tab In the preemption context, political malfunction is complicated by the interplay of two \softline levels of government. Consider a law prohibiting smoking in restaurants. While a one-force \softline {\u8220\'93}majoritarian{\u8221\'94} model would focus on California voters greater taste for anti-smoking laws than \softline voters in North Carolina, a more accurate, two-force model would recognize that tobacco \softline companies probably have far greater influence over the legislature of North Carolina than that of \softline California, making California more likely than North Carolina to enact anti-smoking legislation \softline even if voter preferences in the two states were roughly the same.\par }{\plain \fs24 \tab The alignment of political forces is likely to look different at the federal and state levels. \softline In North Carolina, tobacco companies are highly integrated in the economy, responsible for a \softline high proportion of jobs and revenue coming into the state. They are likely to exert minoritarian \softline pull in the legislative process simply by going about their business, as political leaders take the \softline view that {\u8220\'93}what{\u8217\'92}s good for R.J. Reynolds is good for the state.{\u8221\'94} But tobacco companies cost of \softline participation in California politics will be relatively higher: since they represent a much smaller \softline proportion of the California economy, their political influence in California could entail costly \softline campaign contributions and highly-funded lobbying or {\u8220\'93}public information{\u8221\'94} campaigns.\par }{\plain \fs24 \tab This raises an important dynamic of federalism and preemption: economy of scale of \softline political participation. For an industry substantial enough to exert influence in national politics, \softline it is likely to be more cost efficient to obtain a political result at the national level than to pursue \softline that result through several states. Rather than entering the political process to block regulation in \softline several states, it may be far more cost effective to pursue affirmative deregulation at the federal level and argue that state regulation is preempted: this has the effect of blocking regulation in all \softline 50 states in one blow. At the national level, an industry might exert influence either by spreading \softline its influence widely, or by exerting particularly strong influence over a concentrated group of \softline decisionmakers. Where the influence is focused on regulators, we get the minoritarian bias \softline associated with so-called {\u8220\'93}regulatory capture.{\u8221\'94} In the give-and-take that forms the legislative \softline process, a concentrated minority of legislators can exert disproportionate influence over policy on \softline their salient issue. Thus, where minoritarian influence is focused on a concentration of elected \softline officials {\u8211\'96} say, the tobacco companies{\u8217\'92} influence over Senators from tobacco states {\u8211\'96} the result \softline may be {\u8220\'93}log-rolling{\u8221\'94} politics.{}{\plain \fs24 \super 45{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 45}{}{\plain \fs24 "The members [of Congress] . . make a compact by which each aids the other. This is \softline logrolling."-Bryce: Commonwealth, vol. ii. part iii. chap. lxvii. page 125 (1889)}}} }{\plain \fs24 \par }{\plain \fs24 \tab While Komesar adds sophisitication to this two-force model by developing the concepts \softline of the costs (particularly information costs) of participation and the distribution of stakes, certain \softline key questions remain unanswered. In many instances, there will not be a consensus on what \softline constitutes {\u8220\'93}bias.{\u8221\'94} Is a national law against flag burning the product of majoritarian {\u8220\'93}bias,{\u8221\'94} or an \softline instance of the proper functioning of majority rule in a democracy? Does the impact of {\u8220\'93}trial \softline lawyers{\u8221\'94} on state legislation represent {\u8220\'93}minoritarian{\u8221\'94} bias, or do plaintiffs{\u8217\'92} lawyers represent a \softline broader constituency? Is a law providing for affirmative action in employment or public \softline contracting the result of majoritarian politics, or minoritarian {\u8220\'93}bias{\u8221\'94} (as Ackerman might \softline argued), or is it an example of well-functioning pluralist politics in which well-organized \softline minorities can achieve legislative successes? If the latter, what about special tax breaks for \softline SUVs? \par }{\plain \fs24 \ul0 \tab Nor will it always be clear whether the operation of the political process is minoritarian or \softline majoritarian. A referendum system of legislation {\u8211\'96} such as California{\u8217\'92}s ballot initiative process {\u8211\'96} \softline looks majoritarian on the surface; but if the vote is swayed by a highly organized concentration of \softline single-issue voters, or of high stakes holders who produce misleading advertizing and media \softline coverage, arguably the popular vote be recharacterized as suffering from minoritarian bias. \softline Courts often attribute a majoritarian character to the so-called {\u8220\'93}political branches{\u8221\'94} of \softline government, but that characterization is readily challenged. Superior access of well-organized \softline and well-monied interests, log-rolling politics, and regulatory capture can create severe \softline minoritarian bias within the political branches.\par }{\plain \fs24 \tab Finally, is bias a question of process only, or must the substantive outcome be assessed? \softline This raises the related question of whether the failure to engage in comparative institutional \softline analysis itself a failure of an institutional process. Komesar argues that legal and public policy \softline {\u8220\'93}analysts{\u8221\'94} fail when they do not attempt comparative institutional analysis, and implies that \softline judges have a similar obligation. But the market is a behavioral process that is functionally \softline unconscious, since it is driven by conscious decisions too numerous to make a unified \softline autonomous decision, and therefore cannot engage in analysis of any sort. The political process \softline might be seen as in between: it runs contrary to behavioral expectations to expect individual \softline voters to incur the cost of serious comparative institutional analysis when voting for or against \softline tort reform. Nor does it make much sense to expect interest groups to set aside their interests on \softline comparative institutional grounds {\u8211\'96} to expect the manufacturer{\u8217\'92}s association to support punitive \softline damages as the optimal vehicle to promote safety, or trial lawyers to decide that deregulatory tort \softline reform is better after all. Komesar offers a model of political behavior in which motivations are irrelevant.{}{\plain \fs24 \super 46{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 46}{}{\plain \fs24 Komesar, }{\plain \fs24 \i Imperfect Alternatives}{\plain \fs24 , 58-65.}}} }{\plain \fs24 This suggests that the political process should not be expected to be motivated by \softline analytical correctness. At the same time, legislative bodies have a capacity to engage in analysis, \softline conducting investigations, holding hearings and compiling evidence, and legal doctrine credits \softline them with doing so. Legal doctrine often credits the political process with greater {\u8220\'93}institutional \softline competence{\u8221\'94} than courts to determine whether regulation is to be preferred to the market. Why \softline should we expect less comparative institutional analysis from legislatures than from courts?\par }{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {2. Adjudicative Failure} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab Adjudicative failures can result from overwhelming numbers of cases that strain judicial \softline resources, from high participation costs that prevent litigants from access, or from complexity \softline that overtaxes a court{\u8217\'92}s analytical capacity.{}{\plain \fs24 \super 47{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 47}{}{\plain \fs24 Komesar, }{\plain \fs24 \i Imperfect Alternatives}{\plain \fs24 , __.}}} }{\plain \fs24 In the preemption context, I believe the salient \softline failure of courts is analytical, and in my view, adjudicative failures are particularly pronounced in \softline a cases where a court is faced with the institutional choice between the political process and the \softline market.\par }{\plain \fs24 \tab Courts do, of course, analyze the performance of the market. In some areas of the law, \softline first order decisions requiring a choice between the market or a regulatory response are delegated \softline to courts by statute. Certain antitrust or intellectually property laws, for instance, require courts \softline to assess market performance. In common law cases, whether or not judges self-consciously \softline analyze market performance, they do intervene in the market. In }{\plain \fs24 \i Boomer v. Atlantic Cement Co.}{\plain \fs24 , \softline for example, the court considered whether the plaintiff homeowners, who suffered ongoing harm to their property from pollution caused by the defendant cement company, were entitled to an \softline injunction stopping the cement plant{\u8217\'92}s operation, or were instead entitled to a one-time damage \softline remedy that would be much less costly to the defendant and allow it to keep operating. For \softline Komesar and Posner, who focus on the choice of remedy, the injunctive relief represents a \softline market response, whereas damages substitutes judicial for market decisionmaking. The {\u8220\'93}simple{\u8221\'94} \softline injunction in a nuisance (ongoing property damage) case allocates a property right which can the \softline be the subject of bargaining: the defendant can {\u8220\'93}buy{\u8221\'94} the injunctive right from the plaintiff, \softline paying a price set by {\u8220\'93}the market{\u8221\'94} {\u8211\'96} by privately transacting parties. In contrast, a damages \softline award {\u8211\'96} the amount of compensation for the property damage, which is tantamount to a forced \softline sale to the defendant {\u8211\'96} is set by the court (by the judge or perhaps a jury). While market values \softline are relied upon as relevant evidence of the amount of damages, it is fair to say that the court, and \softline not the market, determines the amount of damages.\par }{\plain \fs24 \tab Yet if one focuses on the substantive liability rule {\u8211\'96} air pollution causes legally \softline cognizable harm to the property rights of the homeowners {\u8211\'96} rather than on the mode of its \softline enforcement, it is plain that either remedial response is a judicial intervention into the market. \softline Without the liability rule, there is a plain market failure: there is no market for clean air, and \softline therefore no mechanism to allow consensual transactions between the plaintiffs and the \softline defendant, or to fairly establish prices for polluting activity. Without liability, the free market \softline gives the polluter the right to use the air for free. Thus, by imposing liability at all, the court \softline allocates authority to itself to intervene in an otherwise unregulated market. The remedial \softline responses differ only in how the judicial intervention feeds back into the market. The injunctive \softline response assigns a clean-air property right to the plaintiff, and thereby creates a market that can set its own price through bargaining. The damages response treats the pollution as a forced sale {\u8211\'96} \softline the plaintiff is not free to decline a transaction, as he would be in a free market {\u8211\'96} and sets the \softline price in the form of a damage award. That few commentators are troubled by the existence of \softline substantive nuisance liability in this air pollution case suggests a possible consensus or comfort \softline level with the institutional choice. While courts are imperfect institutions, in common law tort \softline situations they are often preferable to the market, which tends systematically to undervalue the \softline rights of tort victims. In any event, courts have done reasonably well in comparing their own \softline institutional performance to that of the market.\par }{\plain \fs24 \tab Yet cases involving rigorous judicial review of economic legislation, such as }{\plain \fs24 \i Lochner}{\plain \fs24 , \softline have been widely regarded as judicial failures. Comparative institutional analysis gives some \softline insight into the difference. Whereas in common law cases, the court compares its institutional \softline performance to that of the market, in economic due process cases the court must compare its own \softline institutional ability to assess market performance with that of the political process: is the court in \softline a stronger position than the political process to determine the need for regulation?\par }{\plain \fs24 \tab Most judges are generalists, not economists, but then so are most legislators. Both judges \softline and legislators are in a position to receive evidence and expert advice on complex questions of \softline economics. Some might argue that judges have an edge in deciding complex policy questions, \softline since their decisionmaking is based on a {\u8220\'93}contemplative{\u8221\'94} model freed from political pressures \softline and the {\u8220\'93}hurly burly{\u8221\'94} of politics. Yet that institutional advantage {\u8211\'96} which implies that a judge \softline can make a better decision than a legislature on any question whose complexity requires serious \softline contemplation {\u8211\'96} is more metaphorical than real. Judges have huge dockets, and it is not clear \softline that their work leaves their minds more free from clutter than legislators; even if that were so, judges tend to work alone, while legislators can more effectively divide labor by relying on the \softline work of colleagues. Moreover, judges{\u8217\'92} comparative freedom from political pressures may leave \softline room for more {\u8220\'93}unpopular{\u8221\'94} decisions, but that doesn{\u8217\'92}t necessarily mean }{\plain \fs24 \i better}{\plain \fs24 decisions; if \softline politicians{\u8217\'92} judgment is {\u8220\'93}clouded{\u8221\'94} by political pressures, judges{\u8217\'92} may be equally clouded by their \softline personal biases.\par }{\plain \fs24 \tab At the end of the day, it may be impossible to say whether a court or the political process \softline is in a better position to determine whether a market or deregulatory solution is to be preferred to \softline regulation. The comparative institutional analysis in a case like }{\plain \fs24 \i Lochner}{\plain \fs24 is further complicated \softline by the apples-to-oranges character of comparing market and political institutional failures: for \softline example, is minoritarian bias {\u8220\'93}worse{\u8221\'94} than monopoly? It may be that the only way to generate an \softline answer is to agree on a presumptive decisionmaker, and place a {\u8220\'93}burden of proof{\u8221\'94} on \softline comparative institutional analysis to overcome that presumption. Both the courts and the political \softline process perform very poorly in the of assessing market failure: as long as one institution is not \softline manifestly preferable to the other, our democratic norms prefer a wrong answer from the political \softline process to one from the judicial process. \par }{\plain \fs24 \tab The reaction to the }{\plain \fs24 \i Lochner}{\plain \fs24 era, characterized by a retreat from substantive due process \softline review in cases of economic legislation, can be understood in these terms. Arguably, both the \softline Lochner era decisions and the later decisions repudiating them engaged in institutional analysis \softline of sorts. The Lochner court pointedly described a supposed political process failure {\u8211\'96} a \softline minoritarian bias favoring special interests hoping to bid up wages for bakers by limiting supply \softline of bakers hours {\u8211\'96} as one of the grounds for invalidating the legislation. Lochner can be faulted \softline for an analysis that was not only single institutional, but one that did even that badly. The justices focused on the possible minoritarian bias underlying the law regulating bakers{\u8217\'92} hours, without \softline considering the minoritarian bias that would have opposed such legislation. While the political \softline process was probably not so flawed, the court failed to consider how the market was working. \softline Nor did it consider its own ability to make the second order decision compared to the political \softline process. The judicial system was arguably flawed by its own minoritarian bias coupled with an \softline absence of judicially manageable criteria to apply: an {\u8220\'93}arbitrariness test{\u8221\'94} for legislation was itself \softline arbitrary. Moreover, the courts ability to engage in economic analysis is not demonstrably \softline superior to that of the political process. The courts should not choose the market over the \softline political process in wage and hour cases, for example, not because those substantive decisions \softline are value-judgments reserved to the political process (as is typically argued), but because it is not \softline clear that political process malfunction that may lead to such decisions is worse than the market \softline malfunction. The failure of the economic due process decisions is not that courts allocated the \softline institutional choice to themselves, but rather that they allocated the decision to the market rather \softline than the political process without clear grounds to find that the political process failure was \softline worse than the market{\u8217\'92}s failure.\par }{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {3. Conclusions} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab Comparative institutional analysis, like the institutions themselves, is imperfect, and is \softline bound to suffer malfunction with issues presenting increasing complexity. There is simply too \softline much imprecision in the assessment of an institution{\u8217\'92}s performance, and in the apples-to-oranges \softline quality of comparing the performance of such differing institutions as markets and government, \softline to yield sharply distinct comparisons: the institutional comparisons will be inevitably gross under the best of circumstances, and often indeterminate, particularly in the political-process-versus \softline market comparison.. If the models of political participation and market performance are too \softline complex, judges{\u8217\'92} ability to understand them will be severely taxed, and there is likely to be \softline disagreement and indeterminacy even among experts. We can exchange the jargon of law for the \softline jargon of social science without adding to the quality of decisions. Thus, for courts to employ \softline comparative institutional analysis, they will likely have to rely on intellectual shortcuts and rules \softline of thumb to assess and compare insitutional failure. Moreover, in a range of cases, conclusions \softline to be drawn from comparative institutional analysis will be too uncertain to yield confidence-inspiring answers. Courts must continue to function, and thus need decision rules, in such cases. \softline I believe that in many, if not most, instances where a court has to compare institutional \softline performance of the market with that of the political process in a field of economic regulation, the \softline court{\u8217\'92}s ability to choose is not demonstrably superior to that of the political process, and the \softline court should defer. This is the intuition that has guided the judicial retreat from economic due \softline process doctrine. \par }{\plain \fs24 \par }\pard \fs24\qc\sl480\slmult1 {\plain \fs24 \b {\tc {III. The Comparative Institutional Dynamics of Preemption Cases}}}{\plain \fs24 \par }\pard \fs24\sl480\slmult1 {\plain \fs24 \tab }{\plain \fs24 \b {\tc {A. Institutional Choices in Preemption Cases} \tcl2}}{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {1. Courts, Congress, the States and the Market} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab Looked at one way, every preemption case entails a choice between the federal and state \softline political processes. The Supremacy Clause, which makes federal law supreme over state law, \softline gives Congress the second order {\u8211\'96} who decides who decides? {\u8211\'96} decision. Congress decides \softline whether the federal or state political process will control an issue. The controlling political process (federal or state) in turn decides whether the basic policy decision rests with itself or with \softline the market. Thus, in }{\plain \fs24 \i Geier v. American Honda Motor Corp.}{\plain \fs24 ,{}{\plain \fs24 \super 48{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 48}{}{\plain \fs24 529 U.S. 861 (2000).}}} }{\plain \fs24 }{\plain \fs24 the Court held that Congress gave \softline itself (and its delegee, the Department of Transportation), and denied to the states, the first order \softline decision of deciding how much safety regulation would be imposed on auto manufacturers. DOT \softline determined, beyond a certain threshold, that the decision whether to include airbags in cars for \softline the 1986-89 period would be determined by the market.\par }{\plain \fs24 \tab The role of the courts in this decision is imply to identify, in the context of resolving a \softline particular legal controversy, what Congress{\u8217\'92}s second order decision was. Thus, it is said that the \softline {\u8220\'93}intent of Congress{\u8221\'94} controls the preemption question.{}{\plain \fs24 \super 49{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 49}{}{\plain \fs24 }}} }{\plain \fs24 And the {\u8220\'93}judges{\u8217\'92} clause,{\u8221\'94} which \softline provides that {\u8220\'93}the judges in every state are bound{\u8221\'94} by federal law and by the Supremacy \softline principle, means that the state courts are to be interchangeable with federal courts in this \softline preemption role.{}{\plain \fs24 \super 50{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 50}{}{\plain \fs24 }}} }{\plain \fs24 \par }{\plain \fs24 \tab Yet this description is a simplification. As in virtually any statutory interpretation case, \softline the courts play a significant role with a substantial opportunity to exercise independent judgment. \softline Rarely is the statutory mandate so clear that the preemptive intent of Congress may be simply \softline identified. As with any case of statutory interpretation, statutory language may be ambiguous, or \softline the legislative history unclear.\par }{\plain \fs24 \tab The preemption question {\u8211\'96} the second-order institutional choice question about whether \softline state authority will be superseded by the law {\u8211\'96} is raised by virtually every act of Congress, and Congress may be more or less thoughtful and articulate in dealing with it. At one end of the \softline spectrum we might find an express preemption clause together with a {\u8220\'93}savings clause{\u8221\'94} that \softline indicates the limits of the preemption clause by identifying what state law is {\u8220\'93}saved{\u8221\'94} from \softline preemption. For instance, [ERISA __] At the other end of the spectrum, Congress may not have \softline considered preemption at all. And in between Congress may have considered preemption and \softline simply assumed federal law would nullify conflicting state law, without thinking about specifics \softline or hard cases.\par }{\plain \fs24 \tab Where Congress has been anything less then crystal clear and comprehensive about \softline preemption, there is always the possibility of a conscious effort to pass off the second-order \softline institutional choice question to an administrative agency or to the courts. How to strike the \softline balance between federal regulatory objectives and state autonomy is a difficult question that may \softline not have generated a consensus in the case of a particular enactment, or it may be that Congress \softline preferred to let the courts make the politically troublesome decision. The courts{\u8217\'92} traditional role \softline in interpreting statutes in cases where the statute{\u8217\'92}s application is subject to debate even implies \softline that Congress {\u8211\'96} and perhaps even the constitution{\u8217\'92}s definition of the judicial power {\u8211\'96} assumes \softline that a substantial role is delegated to courts in deciding the second-order preemption question.\par }{\plain \fs24 \tab So, rhetoric aside, it is the courts who decide who decides in most preemption cases. \softline This raises the question of the extent to which courts can and should be influenced by the first \softline order question. Should a court decide whether the market or a state political process is better \softline suited to set safety standards for passive restraint systems like airbags, the first-order question at \softline stake in }{\plain \fs24 \i Geier}{\plain \fs24 ? Traditionally, courts have allocated such decisions to themselves only in the \softline absence of regulation, in their common lawmaking capacity. But if they determine the second order question on a relatively standardless basis, and if the first order decision follows \softline automatically from the second order decision, as is often the case in litigation, arguably courts are \softline making this choice de facto. \par }{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {2. Regulation and Deregulation} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab Like population density on a map, activities may be densely regulated: the employment \softline relationship, for instance, is controlled by laws regulating wages and hours, race and sex equality, \softline safety and accident compensation, and benefits. Other activities are sparsely regulated, but like \softline areas of low population density, low density regulation may still be dotted by small clusters of \softline rules, or rimmed by rules at the edges. The content of books and newspapers is regulated by libel \softline and intellectual property laws, but otherwise is largely left to the market. These examples also \softline suggest that regulatory density may be a question of the scale of the regulatory {\u8220\'93}map.{\u8221\'94} Even if \softline one agrees that {\u8220\'93}employment{\u8221\'94} is densely regulated, by focusing on employee benefits one finds \softline an arguably low-density area of regulation.\par }{\plain \fs24 \tab ERISA, the primary law of employee benefits, provides a good example of considering \softline the interplay between state and federal regulation. ERISA has broad preemptive effect {\u8211\'96} \softline basically, with certain key exceptions, prohibiting states from regulating employee benefits. Yet \softline ERISA is frequently criticized for leaving wide regulatory gaps. For instance, the grounds for \softline challenging wrongful denial of benefits are very limited; and while ERISA establishes certain \softline requirements for benefit plans that qualify for the statute{\u8217\'92}s tax incentives, the statute does not \softline require any employer actually to provide any benefits whatever. States are not permitted to fill in \softline these regulatory gaps, for example, by providing causes of action or standards to challenge benefit denials, or by requiring safeguards for pension funds.{}{\plain \fs24 \super 51{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 51}{}{\plain \fs24 }}} }{\plain \fs24 \par }{\plain \fs24 \tab Since the commerce clause has been understood to embrace most activities, the federal \softline and state regulatory maps overlap considerably. But concurrent, or overlapping, regulation does \softline not create preemption by itself; indeed, parallel state and federal regulation may be more the rule \softline than the exception. An act of employment discrimination violates both federal law and the law \softline of most states; likewise an illegal sale of narcotics. The dual regulatory regime simply creates \softline the option of pursuing a federal or state remedy. A private plaintiff may initially sue under both \softline state and federal employment laws, though the rule against double recovery implies an election of \softline one remedy in the end. Where the remedy involves government intervention, some sort of \softline coordination (it is hoped) can be attained by federal and state authorities in prosecuting the case.\par }{\plain \fs24 \tab As seen above, preemption arises where there is conflict or inconsistency between the \softline state and federal regulatory {\u8220\'93}map.{\u8221\'94} A common pattern is federal deregulation. Where Congress \softline wishes to deregulate an area, it not only curtails federal regulation, but also preempts state law {\u8211\'96} \softline creating low regulatory density by clearing away state regulation. ERISA is arguably an example \softline of this; the Telecommunications Act of 1996 is widely agreed to be of this character.{}{\plain \fs24 \super 52{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 52}{}{\plain \fs24 }}} }{\plain \fs24 \par }{\plain \fs24 \tab Some federal deregulation may be more subtle. Congress can create low density \softline regulation, not necessarily by withdrawing regulation from an area, but by establishing a \softline regulatory standard that sets a ceiling rather than a floor. A regulatory floor creates a minimum \softline standard on which states may impose more stringent or protective regulation. A ceiling \softline effectively operates to displace state regulation; state regulations above the ceiling are preempted, while those at or below the ceiling are redundant. In }{\plain \fs24 \i Geier,}{\plain \fs24 }{\plain \fs24 the Supreme Court found the federal \softline standard to set a ceiling, preempting a state tort rule; the dissenters }{\plain \fs24 argued that the federal \softline standard should have been viewed as a floor, allowing the state regulation. \par }{\plain \fs24 \par }{\plain \fs24 \tab }{\plain \fs24 \b {\tc {B. The Second Order Decision: State versus Federal Political Process Failures} \tcl2}}{\plain \fs24 \par }{\plain \fs24 \tab The second order decision facing courts in preemption cases is whether the federal or \softline state political process will control the first order {\u8220\'93}who decides?{\u8221\'94} question. In comparing the \softline performance of the federal and state political processes, it is important to bear in mind that the \softline array of political forces on a single regulatory issue can look quite different at the two levels. It \softline would be useful here to attempt a rough categorization of the kinds of political process failures \softline that may underlie conflicting federal and state laws in a preemption case by looking at some \softline specific examples.\par }{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {1. Majoritarian Bias and State Experimentation: Controlled Substances } \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab The form of majoritarian bias most frequently discussed in legal literature (typically in the \softline literature of constitutional law) is the oppression of {\u8220\'93}discrete and insular minorites{\u8221\'94} by legislative \softline majorities.{}{\plain \fs24 \super 53{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 53}{}{\plain \fs24 }{\plain \fs24 \i See}{\plain \fs24 Ely, }{\plain \fs24 \i Democracy and Distrust}{\plain \fs24 (1980). The }{\plain \fs24 majoritarian bias that seemed most to \softline concern the Federalists among the framers was a kind of unchecked populism that might lead to \softline radical redistributions of property from the wealthy in the form of takings or repudiations of debt. \softline [cites __] While {\u8220\'93}the rich{\u8221\'94} may in some sense be {\u8220\'93}discrete{\u8221\'94} or {\u8220\'93}insular,{\u8221\'94} they have much greater \softline political power and have historically been much less likely to be targeted by legislative majorities \softline than politically weaker racial or religious minority groups.}}} }{\plain \fs24 Majoritarian bias may also take the form of social control legislation {\u8211\'96} typically \softline criminal laws {\u8211\'96} affecting groups that are not {\u8220\'93}discrete,{\u8221\'94} but shifting, because the regulation controls an activity. People who burn American flags, use recreational drugs, or seek to \softline terminate unwanted pregnancy through abortion are part of the target group when and to the \softline extent that they pursue the activity. Often the regulated group is politically weak relative to the \softline majority, in many instances precisely because membership in the group is ephemeral. In some \softline cases, the impulse to regulate may be something of a fad: This legislation often coalesces around \softline passing public outrage due to a high profile event, or perhaps around political leaders resorting to \softline demagogic tactics, and often produces some type of morals legislation usually in the form of a \softline criminal sanction. Deliberation and opposition can be impeded because the aroused electorate is \softline encouraged by proponents to view the issue simplistically, and potential opponents know that \softline they can become vulnerable targets in a re-election campaign. \par }{\plain \fs24 \tab Generally speaking, it probably cannot be shown that either federal or state political \softline processes are more susceptible than the other to this type of legislation. An argument can be \softline made in some instances that nationwide interest groups are in a better position to oppose such \softline legislation at the national level than in individual states where there influence is comparatively \softline weak. On the other hand, when such legislation succeeds at the federal level, it becomes the law \softline of the land.\par }{\plain \fs24 \tab The case of medical marijuana provides an example. In 1996, California voters enacted \softline Proposition 215, the Compassionate Use Act, through its direct referendum process. The Act \softline which exempts a patient or a patient{\u8217\'92}s primary caregiver from state laws criminalizing possession \softline or cultivation of marijuana {\u8220\'93}for the personal medical purposes of the patient upon the written or \softline oral recommendation or approval of a physician.{\u8221\'94}{}{\plain \fs24 \super 54{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 54}{}{\plain \fs24 Cal. Health & Saf. Code {\u167\'a7}{\u167\'a7} 11362.5 & sub. (d).}}} }{\plain \fs24 In }{\plain \fs24 \i Oakland Cannabis Buyers Cooperative v. United States}{\plain \fs24 ,{}{\plain \fs24 \super 55{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 55}{}{\plain \fs24 532 U.S. 483 (2001).}}} }{\plain \fs24 this statute was held preempted by the federal Controlled Substances Act, under \softline which marijuana is made a {\u8220\'93}schedule I{\u8221\'94} substance. These are substances which Congress has \softline determined to have {\u8220\'93}no accepted medical use,{\u8221\'94} and therefore no doctors in the United States can \softline prescribe them.{}{\plain \fs24 \super 56{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 56}{}{\plain \fs24 21 U.S.C. {\u167\'a7} 841(a)(1); }{\plain \fs24 \i see Oakland Cannabis Buyers}{\plain \fs24 , 532 U.S. at 489-90.}}} }{\plain \fs24 \par }{\plain \fs24 \tab The medical marijuana cases raise the {\u8220\'93}states as laboratories{\u8221\'94} in plain form. California \softline and other states seek not only to experiment in social policy, but even literally in biology and \softline medicine, to generate data that could well cause a revision of the scientific basis to the \softline Congressional conclusion that marijuana has no accepted medical use. Under a broad, but widely \softline accepted view of {\u8220\'93}obstacle{\u8221\'94} preemption, the state law defenses to the distribution and possession \softline crimes for marijuana would be preempted by federal law.\par }{\plain \fs24 \tab In comparative institutional terms, however, there is a strong argument to be made against \softline preemption. If it is acknowledged that the prohibition on marijuana is justified on practical \softline scientific grounds (as opposed to moral or irrational grounds), an institutional choice that allows \softline a limited state experiment has much to recommend it. The federal law may well have been the \softline result of majoritarian bias to begin with, and the political unfeasibility of experimenting with \softline medical marijuana at the federal level suggests a continuing majoritarian bias. The process at the \softline state level leading to California{\u8217\'92}s medical marijuana law, in contrast, appears to have functioned \softline comparatively well. The law was the result of a majority vote in a statewide referendum; no \softline discrete or insular minority appears to have been oppressed as the result of it. The widely \softline dispersed low stakeholders who made up this voter majority had to overcome a minoritarian bias in the form of the politically powerful prison guard lobby, which has successfully backed such \softline measures as the state{\u8217\'92}s three strikes law and generally opposes decriminalization policies. At \softline minimum the opposition position was well, if not disproportionately, represented.\par }{\plain \fs24 \tab At the same time, to the extent that policy experimentation in this field is desirable, states \softline have a comparative advantage. Few if any national politicians are calling for decriminalization \softline of drugs under federal law, for fear of majoritarian backlash. California avoided this problem \softline with its referendum system, which of course does not exist at the federal level. And conducting \softline the experiment on a small scale (though certainly the federal government could choose to limit \softline the scale as well) reduces the potential cost of failure.\par }{\plain \fs24 \par }{\plain \fs24 \tab }{\plain \fs24 \b \ul0 \tab {\tc {2. National Unification: Antidiscrimination Laws} \tcl3}}{\plain \fs24 \ul0 \tab \par }{\plain \fs24 \tab So why not allow states to {\u8220\'93}experiment{\u8221\'94} with reintroducing Jim Crow laws or allowing \softline employers to discriminate on the basis of race and sex?{}{\plain \fs24 \super 57{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 57}{}{\plain \fs24 }{\plain \fs24 \i Cf. }{\plain \fs24 Ann Althouse, Vanguard States, Laggard States: Federalism and Constitutional \softline Rights, 152 U. Penn. L. Rev. 1745 (2004) (suggesting that the concept of state experimentation \softline should extend to {\u8220\'93}laggard states{\u8221\'94} which experiment with retaining {\u8220\'93}traditions,{\u8221\'94} such as an all-male state military academy). }}} }{\plain \fs24 \par }{\plain \fs24 \tab There are significant and systematic differences between the drug laws and the \softline antidiscrimination laws. To begin with, states were allowed to experiment with race and sex \softline discrimination for nearly two centuries before a nationwide consensus developed that that social \softline policy was a failure. Nor does there seem to be any significant movement to experiment with a \softline return to discriminatory traditions: state antidiscrimination laws are for the most part \softline complementary and integrated with the federal scheme, and the small number of preemption issues that arise concern fine points, such as whether states may provide greater protection than \softline the federal standard.{}{\plain \fs24 \super 58{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 58}{}{\plain \fs24 Cal. Fed. S & L Assn. v. Guerra, 479 U.S. 272 (1987); }{\plain \fs24 \i but cf. }{\plain \fs24 Nevada Dept. of Human \softline Res. v. Hibbs, 538 U.S. 721 (2003).}}} }{\plain \fs24 \par }{\plain \fs24 \tab More generally, it might be said that the federal government has a comparative \softline institutional advantage over the states when it comes to policies aimed at national unification or a \softline nationwide implementation of an equality principle. This institutional choice is implicit in \softline several constitutional provisions or doctrines, such as the right to travel from state to state, the \softline privileges and immunities clause of Article IV, and the 14}{\plain \fs24 \super th}{\plain \fs24 amendment. When Congress \softline legislates under its 14}{\plain \fs24 \super th}{\plain \fs24 amendment power a national unification objective is likely to be present. \par }{\plain \fs24 \tab Federal antidiscrimination law is based on the 14}{\plain \fs24 \super th}{\plain \fs24 amendment insofar as it regulates the \softline states. That its regulation of private business (employers and places of public accommodation{}{\plain \fs24 \super 59{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 59}{}{\plain \fs24 {\u8220\'93}Public{\u8221\'94} in the sense of a privately-owned business open to the public.}}} }{\plain \fs24 ) \softline is based on the commerce clause arises from the controversial and debatable Supreme Court \softline doctrine requiring state action for 14}{\plain \fs24 \super th}{\plain \fs24 amendment legislation, and does not denigrate the national \softline unification objectives of laws like Title VII. Indeed, in building a case for the impact on \softline interstate commerce, the legislative history of Title VII is replete with arguments and examples \softline of how race discrimination in employment and places of public accommodation interfered with \softline interstate job mobility and travel for African Americans.{}{\plain \fs24 \super 60{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 60}{}{\plain \fs24 [Cites __]}}} }{\plain \fs24 \par }{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {3. National Coordination: Race to the Bottom and NIMBYism} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \ul0 \tab National legislation has sometimes been called for to prevent a so-called {\u8220\'93}race to the \softline bottom,{\u8221\'94} in which states compete for business relocation by providing a deregulatory climate. \softline The fear is that states will be discouraged from protective regulation {\u8211\'96} pollution controls, child \softline labor laws, safety standards {\u8211\'96} as they will be penalized by having business relocate, taking their \softline jobs and tax base with them to more permissive states. The race to the bottom tends to suppress \softline regulation throughout the system. This is a kind of political process failure to which a federal \softline system seems uniquely susceptible. \par }{\plain \fs24 \tab The political process failure shows up in at least two significant respects in this situation. \softline First, the deregulatory states may be adopting {\u8220\'93}free market{\u8221\'94} solutions due to minoritarian bias \softline from political powerful business interests which dominate the dispersed low stakes majority \softline interests in pollution control, product safety or the like. To be sure, it can always be argued that \softline deregulation serves majority interests, too, in the form of jobs and tax revenue. Perhaps the \softline deregulatory state is setting the balance between pollution and safety, on the one hand, and \softline economic development on the other at a sensible level. Yet, even if this is the case, the \softline deregulatory state creates externalities, in the form of pressures on other states to follow suit {\u8211\'96} \softline hence the race to the bottom. The majority voting for deregulation in one state, thus becomes a \softline minority in the larger national picture, one that wields disproportionate influence in creating the \softline downward pressure. A strong argument can be made that the national political process is the \softline preferable institution to regulate. Significantly, legislation to prevent a race to the bottom would \softline typically take the form of a federal floor, rather than a ceiling, and would not preempt more \softline stringent regulation by individual states.\par }{\plain \fs24 \ul0 \tab Majoritarian bias can also take the form of so-called {\u8220\'93}NIMBY{\u8221\'94} legislation,{}{\plain \fs24 \super 61{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 61}{}{\plain \fs24 NIMBY stands for {\u8220\'93}not in my back yard.{\u8221\'94}}}} }{\plain \fs24 in which \softline majorities refuse to take on the costs of socially necessary activities that generate externalities {\u8211\'96} \softline such as waste disposal sites or low-income housing. Typically, though not always, NIMBY \softline legislation occurs at the state or local level, and often takes the form of zoning regulation. Here, \softline too, the national government{\u8217\'92}s comparative advantage in interstate coordination would favor \softline preemption. \par }{\plain \fs24 \par }{\plain \fs24 \tab \tab }{\plain \fs24 \b {\tc {4. Federal Deregulation: Tort Reform} \tcl3}}{\plain \fs24 \par }{\plain \fs24 \tab Perhaps the example of Title VII illustrates the broader point that interstate commerce \softline regulation by Congress promotes national unification. Certainly, the dormant commerce clause \softline doctrines aimed at state protectionism or discrimination against out of state businesses are \softline justified on these grounds. It can always be, and often is, argued that uniform regulation, rather \softline than a patchwork of state laws, is more efficient and more conducive to growth of nationwide \softline industries. This argument raises the greatest difficulty for preemption doctrine generally, and for \softline an argument for state autonomy in economic regulation in particular. \par }{\plain \fs24 \tab From one point of view, national deregulatory legislation can look like the product of \softline minoritarian bias. The economy of scale of acheiving deregulation at the federal level makes the \softline federal political process an attractive target for politically powerful business interests. Whether \softline the mechanism is logrolling politics through the influence over a small, concentrated group of \softline political leaders, or broader access, the ability of such minoritarian interests to dominate the \softline political process is so commonplace that it forms the basis for most models of contemporary politics.\par }{\plain \fs24 \tab State regulations that come into conflict with federal deregulatory laws are more variable. \softline Most laws are given a public interest spin {\u8211\'96} a law to restrict competition from nondairy products \softline in a strong dairy state is likely to be fashioned a public health law, for instance {\u8211\'96} but in many \softline cases they are genuinely aimed at protecting the interests of a widely-dispersed, low-stakes \softline majority. The pollution control and tort laws in }{\plain \fs24 \i Geier}{\plain \fs24 and }{\plain \fs24 \i Manufacturer{\u8217\'92}s Association}{\plain \fs24 are \softline probably of the latter category.\par }{\plain \fs24 \tab Yet the problems identified earlier in assessing institutional performance come into play \softline here. In many instances, it will not be clear how the political processes actually functioned. \softline With the exception of the relatively unusual state referendum legislation, the state and federal \softline laws will be filtered through representative political processes where access depends on \softline organization. Courts may be unwilling or unable to assess the relative degrees of political \softline process failure in cases of interest groups pitted against one another. \par }{\plain \fs24 \tab In addition, what if a court genuinely believes that the more defective political process \softline produces a better policy result, for example striking a better balance between economic growth \softline and safety or pollution control? Should courts decide whether a state regulation is preempted by \softline a federal deregulatory regime on the basis of a direct and open analysis of whether the market \softline provides the preferred solution to a particular problem? For the reasons stated above, I believe \softline the answer is {\u8220\'93}no.{\u8221\'94} Courts{\u8217\'92} capacity for in-depth analysis of markets to compare the institutional \softline performance of market and political processes is not adequate to yield answers certain enough to \softline justify displacing the political process in favor of the market. \par }{\plain \fs24 \tab This principle alone weighs against preemption in a large number of cases. It is frequently the case that Congress cannot reasonably be said to have considered the preemptive \softline application framed by a lawsuit. Yet the courts approach is to engage in a {\u8220\'93}freewheeling inquiry{\u8221\'94} \softline into the objectives behind the federal law, and find preemption if the state law is inconsistent \softline with those objectives.{}{\plain \fs24 \super 62{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 62}{}{\plain \fs24 Gade v. National Solid Wastes Management Assn., 505 US. 88, 110 (1992) (Kennedy, \softline J., concurring).}}} }{\plain \fs24 In enacting the Federal Arbitration Act, for example, there is nothing in \softline the statute or its legislative history indicating that Congress sought to preempt state consumer \softline protection laws providing for class actions (laws which may not even have existed when the FAA \softline was enacted in 1925). Yet in the Court has seriously entertained the argument that such state \softline regulation is preempted, on the ground that the those laws interfere with the Congressional \softline objective of providing that arbitration agreements be enforced {\u8220\'93}as written.{\u8221\'94}{}{\plain \fs24 \super 63{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 63}{}{\plain \fs24 }{\plain \fs24 \i See }{\plain \fs24 Green Tree Financial Corp. v. Bazzle, 123 S. Ct. 2402 (2003); Schwartz, }{\plain \fs24 \i State \softline Judges}{\plain \fs24 __. }}} }{\plain \fs24 In cases such as \softline this, courts become an instrument of minoritarian bias, as they allow litigants to achieve, through \softline opportunistic legal arguments, deregulatory objectives that could only otherwise have been won \softline through exertion of minoritarian bias on the political process. \par }{\plain \fs24 \par }\pard \fs24\qc\sl480\slmult1 {\plain \fs24 \b {\tc {Conclusion: Doctrinal Space for Comparative Institutional Analysis}}}{\plain \fs24 \par }\pard \fs24\sl480\slmult1 {\plain \fs24 \tab I have argued that comparative institutional analysis suggests that, where morals \softline legislation is at issue, state political processes should be preferred to federal all other things being \softline equal. At the same time, if one political process (whether state or federal) protects a {\u8220\'93}discrete \softline and insular minority,{\u8221\'94} that political process is more likely to have functioned better than one \softline which dominates such a minority, and is to be preferred. On the other hand, a clear case that the federal law serves a national unification objective, comparative institutional analysis arguably \softline favors the federal political process. These observations dovetail with legal doctrine, insofar as \softline laws targeting political minorities tend to be morals legislation, which is the traditional domain \softline of the states. Some judges are beginning to suggest that doubts about the extent of federal power \softline should result in statutory rulings that the federal law does not apply, thereby avoiding the \softline preemption question.{}{\plain \fs24 \super 64{\footnote \pard \fs24\sa240 {\ul0 \tab }{\plain \fs24 \super 64}{}{\plain \fs24 }{\plain \fs24 \i [cites __]}}} }{\plain \fs24 \par }{\plain \fs24 \tab The legal-cultural norm that underlies the judicial retreat from economic due process \softline might be taken to express general disapproval of judicial decisions that make the enforceability \softline of economic legislation depend on perceived failures in the political process. Under this view, \softline courts should not engage in comparative institutuional analysis because they should not analyze \softline the political process, at least insofar as it produces economic legislation within a broad range of \softline {\u8220\'93}reasonableness{\u8221\'94} or {\u8220\'93}rationality.{\u8221\'94}\par }{\plain \fs24 \tab Yet there are two critical differences between judicial review and preemption cases that \softline make such objections less troubling. First, in preemption cases, the second order institutional \softline choice is not between the market and the political process, but between state and federal political \softline processes. To be sure, if the federal law is deregulatory, then federal law chooses the market over \softline state law as the institution to pursue the policy goal. But where Congress{\u8217\'92}s preemptive mandate \softline is clear, the court can defer to the decision by the federal political process to choose the market.\par }{\plain \fs24 \tab Second, in contrast to constitutional adjudication, in most cases the court cannot allocate \softline the second order decision to another institution. In preemption cases, a court is faced with a \softline special kind of conflict of laws decision: it must choose between two applicable laws, one state one federal, to supply the rule of decision in the case. It cannot decline to choose. To be sure, a \softline deference model does exist in preemption cases: the Supremacy Clause chooses federal over state \softline law in cases of conflict, and preemption doctrine purports to give Congress in all cases the \softline institutional choice of itself or the states. But in the majority of preemption cases, the deference \softline model is a thinly-veiled fiction. Most statutes do not have express preemption clauses, and rarely \softline is there evidence of an intent to displace state law.\par }{\plain \fs24 \tab The most difficult question involves deregulation. The national political process has to \softline have the ability to implement a deregulation agenda on the good faith belief that such a course \softline best promotes the national welfare, even at some risk of serious minoritarian bias. This is \softline supported by the interstate commerce coordination objective of the constitution and the \softline comparative institutional advantage of the national government in pursuing such a course. Some \softline deregulatory preemption is thus inevitable. In cases where state regulation would clearly nullify \softline the federal law, or in cases where Congress has been very clear about displacing state law, \softline comparative institutional analysis does not point away from the constitution{\u8217\'92}s Supremacy \softline principle. \par }{\plain \fs24 \tab But in a substantial proportion regulatory preemption cases, the search for Congressional \softline intent is so speculative that the court is on its own, and becomes de facto the primary \softline decisionmaker of the choice between deregulation and state regulation. In these cases, it is quite \softline likely that judges are in fact engaging in institutional analysis, albeit in an intuitive and possibly \softline single institutional way, consulting their own views on political process or market failure. I have \softline argued that comparative institutional analysis does not usually support a court allocating a \softline decision to the market when faced with a determination by the political process to regulate, because there is no strong reason to believe that courts are better able than the political process to \softline make the second order determination. Thus, courts should refrain from taking such decisions \softline away from the state political process in the absence of clear congressional intent. This doctrine is \softline consistent with the presumption against preemption and {\u8220\'93}clear statement{\u8221\'94} rules.\par }{\plain \fs24 \tab Preemption cases would be uncontroversial if the {\u8220\'93}better{\u8221\'94} law always prevailed. While \softline there is of course great room for disagreement about which policy is the wiser {\u8211\'96} nationwide \softline deregulation of telecommunications (the market) or locally-determined consumer protection (the \softline courts and the political process), for example {\u8211\'96} there is an argument to be made that there is \softline room for courts should at least engage in comparative institutional analysis of whose political \softline process has worked better in passing the laws in question.\par }\pard \fs24\notabind\sl480\slmult1 {\field{\*\fldinst TOC \\f}{\fldrslt {\plain \fs24 }}} \par }