Opinion

Justice O’Connor, with whom The Chief Justice and Justice Thomas join as to all but Part III, dissenting.

We enforce the “outer limits” of Congress’ Commerce Clause authority not for their own sake, but to protect historic spheres of state sovereignty from excessive federal encroachment and thereby to maintain the distribution of power fundamental to our federalist system of government. United States v. Lopez, 514 U. S. 549, 557 (1995); NLRB v. Jones & Laughlin Steel Corp., 301 U. S. 1, 37 (1937). One of federalism’s chief virtues, of course, is that it promotes innovation by allowing for the possibility that “a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.” New State Ice Co. v. Liebmann, 285 U. S. 262, 311 (1932) (Brandeis, J., dissenting).

This case exemplifies the role of States as laboratories. The States’ core police powers have always included authority to define criminal law and to protect the health, safety, and welfare of their citizens. Brecht v. Abrahamson, 507 U. S. 619, 635 (1993); Whalen v. Roe, 429 U. S. 589, 603, n. 30 (1977). Exercising those powers, California (by ballot initiative and then by legislative codification) has come to its own conclusion about the difficult and sensitive question of whether marijuana should be available to relieve severe pain and suffering. Today the Court sanctions an application of the federal Controlled Substances Act that extinguishes that experiment, without any proof that the personal cultivation, possession, and use of marijuana for medicinal purposes, if economic activity in the first place, has a substantial effect on interstate commerce and is therefore an appropriate subject of federal regulation. In so doing, the Court announces a rule that gives Congress a perverse incentive to legislate broadly pursuant to the Commerce Clause—nestling questionable assertions of its authority into comprehensive regulatory schemes—rather than with precision. That rule and the result it produces in this case are irreconcilable with our decisions in Lopez, supra, and United States v. Morrison, 529 U. S. 598 (2000). Accordingly I dissent.

I

In Lopez, we considered the constitutionality of the Gun-Free School Zones Act of 1990, which made it a federal offense “for any individual knowingly to possess a firearm … at a place the individual knows, or has reasonable cause to believe, is a school zone,” 18 U. S. C. §922(q)(2)(A). We explained that “Congress’ commerce authority includes the power to regulate those activities having a substantial relation to interstate commerce …, i.e., those activities that substantially affect interstate commerce.” 514 U. S., at 558–559 (citation omitted). This power derives from the conjunction of the Commerce Clause and the Necessary and Proper Clause. Garcia v. San Antonio Metropolitan Transit Authority, 469 U. S. 528, 585–586 (1985) (O’Connor, J., dissenting) (explaining that United States v. Darby, 312 U. S. 100 (1941), United States v. Wrightwood Dairy Co., 315 U. S. 110 (1942), and Wickard v. Filburn, 317 U. S. 111 (1942), based their expansion of the commerce power on the Necessary and Proper Clause, and that “the reasoning of these cases underlies every recent decision concerning the reach of Congress to activities affecting interstate commerce”); ante, at 2 (Scalia, J., concurring in judgment). We held in Lopez that the Gun-Free School Zones Act could not be sustained as an exercise of that power.

Our decision about whether gun possession in school zones substantially affected interstate commerce turned on four considerations. Lopez, supra, at 559–567; see also Morrison, supra, at 609–613. First, we observed that our “substantial effects” cases generally have upheld federal regulation of economic activity that affected interstate commerce, but that §922(q) was a criminal statute having “nothing to do with ‘commerce’ or any sort of economic enterprise.” Lopez, 514 U. S., at 561. In this regard, we also noted that “[s]ection 922(q) is not an essential part of a larger regulation of economic activity, in which the regulatory scheme could be undercut unless the intrastate activity were regulated. It cannot, therefore, be sustained under our cases upholding regulations of activities that arise out of or are connected with a commercial transaction, which viewed in the aggregate, substantially affects interstate commerce.” Ibid. Second, we noted that the statute contained no express jurisdictional requirement establishing its connection to interstate commerce. Ibid.

Third, we found telling the absence of legislative findings about the regulated conduct’s impact on interstate commerce. We explained that while express legislative findings are neither required nor, when provided, dispositive, findings “enable us to evaluate the legislative judgment that the activity in question substantially affect[s] interstate commerce, even though no such substantial effect [is] visible to the naked eye.” Id., at 563. Finally, we rejected as too attenuated the Government’s argument that firearm possession in school zones could result in violent crime which in turn could adversely affect the national economy. Id., at 563–567. The Constitution, we said, does not tolerate reasoning that would “convert congressional authority under the Commerce Clause to a general police power of the sort retained by the States.” Id., at 567. Later in Morrison, supra, we relied on the same four considerations to hold that §40302 of the Violence Against Women Act of 1994, 42 U. S. C. §13981, exceeded Congress’ authority under the Commerce Clause.

In my view, the case before us is materially indistinguishable from Lopez and Morrison when the same considerations are taken into account.

II

A

What is the relevant conduct subject to Commerce Clause analysis in this case? The Court takes its cues from Congress, applying the above considerations to the activity regulated by the Controlled Substances Act (CSA) in general. The Court’s decision rests on two facts about the CSA: (1) Congress chose to enact a single statute providing a comprehensive prohibition on the production, distribution, and possession of all controlled substances, and (2) Congress did not distinguish between various forms of intrastate noncommercial cultivation, possession, and use of marijuana. See 21 U. S. C. §§841(a)(1), 844(a). Today’s decision suggests that the federal regulation of local activity is immune to Commerce Clause challenge because Congress chose to act with an ambitious, all-encompassing statute, rather than piecemeal. In my view, allowing Congress to set the terms of the constitutional debate in this way, i.e., by packaging regulation of local activity in broader schemes, is tantamount to removing meaningful limits on the Commerce Clause.

The Court’s principal means of distinguishing Lopez from this case is to observe that the Gun-Free School Zones Act of 1990 was a “brief, single-subject statute,” ante, at 20, see also ante, at 19, whereas the CSA is “a lengthy and detailed statute creating a comprehensive framework for regulating the production, distribution, and possession of five classes of ‘controlled substances,’ ” ibid. Thus, according to the Court, it was possible in Lopez to evaluate in isolation the constitutionality of criminalizing local activity (there gun possession in school zones), whereas the local activity that the CSA targets (in this case cultivation and possession of marijuana for personal medicinal use) cannot be separated from the general drug control scheme of which it is a part.

Today’s decision allows Congress to regulate intrastate activity without check, so long as there is some implication by legislative design that regulating intrastate activity is essential (and the Court appears to equate “essential” with “necessary”) to the interstate regulatory scheme. Seizing upon our language in Lopez that the statute prohibiting gun possession in school zones was “not an essential part of a larger regulation of economic activity, in which the regulatory scheme could be undercut unless the intrastate activity were regulated,” 514 U. S., at 561, the Court appears to reason that the placement of local activity in a comprehensive scheme confirms that it is essential to that scheme. Ante, at 21–22. If the Court is right, then Lopez stands for nothing more than a drafting guide: Congress should have described the relevant crime as “transfer or possession of a firearm anywhere in the nation”—thus including commercial and noncommercial activity, and clearly encompassing some activity with assuredly substantial effect on interstate commerce. Had it done so, the majority hints, we would have sustained its authority to regulate possession of firearms in school zones. Furthermore, today’s decision suggests we would readily sustain a congressional decision to attach the regulation of intrastate activity to a pre-existing comprehensive (or even not-so-comprehensive) scheme. If so, the Court invites increased federal regulation of local activity even if, as it suggests, Congress would not enact a new interstate scheme exclusively for the sake of reaching intrastate activity, see ante, at 22, n. 33; ante, at 6 (Scalia, J., concurring in judgment).

I cannot agree that our decision in Lopez contemplated such evasive or overbroad legislative strategies with approval. Until today, such arguments have been made only in dissent. See Morrison, 529 U. S., at 657 (Breyer, J., dissenting) (given that Congress can regulate “ ‘an essential part of a larger regulation of economic activity,’ ” “can Congress save the present law by including it, or much of it, in a broader ‘Safe Transport’ or ‘Worker Safety’ act?”). Lopez and Morrison did not indicate that the constitutionality of federal regulation depends on superficial and formalistic distinctions. Likewise I did not understand our discussion of the role of courts in enforcing outer limits of the Commerce Clause for the sake of maintaining the federalist balance our Constitution requires, see Lopez, 514 U. S., at 557; id., at 578 (Kennedy, J., concurring), as a signal to Congress to enact legislation that is more extensive and more intrusive into the domain of state power. If the Court always defers to Congress as it does today, little may be left to the notion of enumerated powers.

The hard work for courts, then, is to identify objective markers for confining the analysis in Commerce Clause cases. Here, respondents challenge the constitutionality of the CSA as applied to them and those similarly situated. I agree with the Court that we must look beyond respondents’ own activities. Otherwise, individual litigants could always exempt themselves from Commerce Clause regulation merely by pointing to the obvious—that their personal activities do not have a substantial effect on interstate commerce. See Maryland v. Wirtz, 392 U. S. 183, 193 (1968); Wickard, 317 U. S., at 127–128. The task is to identify a mode of analysis that allows Congress to regulate more than nothing (by declining to reduce each case to its litigants) and less than everything (by declining to let Congress set the terms of analysis). The analysis may not be the same in every case, for it depends on the regulatory scheme at issue and the federalism concerns implicated. See generally Lopez, 514 U. S., at 567; id., at 579 (Kennedy, J., concurring).

A number of objective markers are available to confine the scope of constitutional review here. Both federal and state legislation—including the CSA itself, the California Compassionate Use Act, and other state medical marijuana legislation—recognize that medical and nonmedical ( i.e., recreational) uses of drugs are realistically distinct and can be segregated, and regulate them differently. See 21 U. S. C. §812; Cal. Health & Safety Code Ann. §11362.5 (West Supp. 2005); ante, at 1 (opinion of the Court). Respondents challenge only the application of the CSA to medicinal use of marijuana. Cf. United States v. Raines, 362 U. S. 17, 20–22 (1960) (describing our preference for as-applied rather than facial challenges). Moreover, because fundamental structural concerns about dual sovereignty animate our Commerce Clause cases, it is relevant that this case involves the interplay of federal and state regulation in areas of criminal law and social policy, where “States lay claim by right of history and expertise.” Lopez, supra, at 583 (Kennedy, J., concurring); see also Morrison, supra, at 617–619; Lopez, supra, at 580 (Kennedy, J., concurring) (“The statute before us upsets the federal balance to a degree that renders it an unconstitutional assertion of the commerce power, and our intervention is required”); cf. Garcia, 469 U. S., at 586 (O’Connor, J., dissenting) (“[S]tate autonomy is a relevant factor in assessing the means by which Congress exercises its powers” under the Commerce Clause). California, like other States, has drawn on its reserved powers to distinguish the regulation of medicinal marijuana. To ascertain whether Congress’ encroachment is constitutionally justified in this case, then, I would focus here on the personal cultivation, possession, and use of marijuana for medicinal purposes.

B

Having thus defined the relevant conduct, we must determine whether, under our precedents, the conduct is economic and, in the aggregate, substantially affects interstate commerce. Even if intrastate cultivation and possession of marijuana for one’s own medicinal use can properly be characterized as economic, and I question whether it can, it has not been shown that such activity substantially affects interstate commerce. Similarly, it is neither self-evident nor demonstrated that regulating such activity is necessary to the interstate drug control scheme.

The Court’s definition of economic activity is breathtaking. It defines as economic any activity involving the production, distribution, and consumption of commodities. And it appears to reason that when an interstate market for a commodity exists, regulating the intrastate manufacture or possession of that commodity is constitutional either because that intrastate activity is itself economic, or because regulating it is a rational part of regulating its market. Putting to one side the problem endemic to the Court’s opinion—the shift in focus from the activity at issue in this case to the entirety of what the CSA regulates, see Lopez, supra, at 565 (“depending on the level of generality, any activity can be looked upon as commercial”)—the Court’s definition of economic activity for purposes of Commerce Clause jurisprudence threatens to sweep all of productive human activity into federal regulatory reach.

The Court uses a dictionary definition of economics to skirt the real problem of drawing a meaningful line between “what is national and what is local,” Jones & Laughlin Steel, 301 U. S., at 37. It will not do to say that Congress may regulate noncommercial activity simply because it may have an effect on the demand for commercial goods, or because the noncommercial endeavor can, in some sense, substitute for commercial activity. Most commercial goods or services have some sort of privately producible analogue. Home care substitutes for daycare. Charades games substitute for movie tickets. Backyard or windowsill gardening substitutes for going to the supermarket. To draw the line wherever private activity affects the demand for market goods is to draw no line at all, and to declare everything economic. We have already rejected the result that would follow—a federal police power. Lopez, supra, at 564.

In Lopez and Morrison, we suggested that economic activity usually relates directly to commercial activity. See Morrison, 529 U. S., at 611, n. 4 (intrastate activities that have been within Congress’ power to regulate have been “of an apparent commercial character”); Lopez, 514 U. S., at 561 (distinguishing the Gun-Free School Zones Act of 1990 from “activities that arise out of or are connected with a commercial transaction”). The homegrown cultivation and personal possession and use of marijuana for medicinal purposes has no apparent commercial character. Everyone agrees that the marijuana at issue in this case was never in the stream of commerce, and neither were the supplies for growing it. (Marijuana is highly unusual among the substances subject to the CSA in that it can be cultivated without any materials that have traveled in interstate commerce.) Lopez makes clear that possession is not itself commercial activity. Ibid. And respondents have not come into possession by means of any commercial transaction; they have simply grown, in their own homes, marijuana for their own use, without acquiring, buying, selling, or bartering a thing of value. Cf. id., at 583 (Kennedy, J., concurring) (“The statute now before us forecloses the States from experimenting … and it does so by regulating an activity beyond the realm of commerce in the ordinary and usual sense of that term”).

The Court suggests that Wickard, which we have identified as “perhaps the most far reaching example of Commerce Clause authority over intrastate activity,” Lopez, supra, at 560, established federal regulatory power over any home consumption of a commodity for which a national market exists. I disagree. Wickard involved a challenge to the Agricultural Adjustment Act of 1938 (AAA), which directed the Secretary of Agriculture to set national quotas on wheat production, and penalties for excess production. 317 U. S., at 115–116. The AAA itself confirmed that Congress made an explicit choice not to reach—and thus the Court could not possibly have approved of federal control over—small-scale, noncommercial wheat farming. In contrast to the CSA’s limitless assertion of power, Congress provided an exemption within the AAA for small producers. When Filburn planted the wheat at issue in Wickard, the statute exempted plantings less than 200 bushels (about six tons), and when he harvested his wheat it exempted plantings less than six acres. Id., at 130, n. 30. Wickard, then, did not extend Commerce Clause authority to something as modest as the home cook’s herb garden. This is not to say that Congress may never regulate small quantities of commodities possessed or produced for personal use, or to deny that it sometimes needs to enact a zero tolerance regime for such commodities. It is merely to say that Wickard did not hold or imply that small-scale production of commodities is always economic, and automatically within Congress’ reach.

Even assuming that economic activity is at issue in this case, the Government has made no showing in fact that the possession and use of homegrown marijuana for medical purposes, in California or elsewhere, has a substantial effect on interstate commerce. Similarly, the Government has not shown that regulating such activity is necessary to an interstate regulatory scheme. Whatever the specific theory of “substantial effects” at issue ( i.e., whether the activity substantially affects interstate commerce, whether its regulation is necessary to an interstate regulatory scheme, or both), a concern for dual sovereignty requires that Congress’ excursion into the traditional domain of States be justified.

That is why characterizing this as a case about the Necessary and Proper Clause does not change the analysis significantly. Congress must exercise its authority under the Necessary and Proper Clause in a manner consistent with basic constitutional principles. Garcia, 469 U. S., at 585 (O’Connor, J., dissenting) (“It is not enough that the ‘end be legitimate’; the means to that end chosen by Congress must not contravene the spirit of the Constitution”). As Justice Scalia recognizes, see ante, at 7 (opinion concurring in judgment), Congress cannot use its authority under the Clause to contravene the principle of state sovereignty embodied in the Tenth Amendment. Ibid. Likewise, that authority must be used in a manner consistent with the notion of enumerated powers—a structural principle that is as much part of the Constitution as the Tenth Amendment’s explicit textual command. Accordingly, something more than mere assertion is required when Congress purports to have power over local activity whose connection to an intrastate market is not self-evident. Otherwise, the Necessary and Proper Clause will always be a back door for unconstitutional federal regulation. Cf. Printz v. United States, 521 U. S. 898, 923 (1997) (the Necessary and Proper Clause is “the last, best hope of those who defend ultra vires congressional action”). Indeed, if it were enough in “substantial effects” cases for the Court to supply conceivable justifications for intrastate regulation related to an interstate market, then we could have surmised in Lopez that guns in school zones are “never more than an instant from the interstate market” in guns already subject to extensive federal regulation, ante, at 8 (Scalia, J., concurring in judgment), recast Lopez as a Necessary and Proper Clause case, and thereby upheld the Gun-Free School Zones Act of 1990. (According to the Court’s and the concurrence’s logic, for example, the Lopez court should have reasoned that the prohibition on gun possession in school zones could be an appropriate means of effectuating a related prohibition on “sell[ing]” or “deliver[ing]” firearms or ammunition to “any individual who the licensee knows or has reasonable cause to believe is less than eighteen years of age.” 18 U. S. C. §922(b)(1) (1988 ed., Supp. II).)

There is simply no evidence that homegrown medicinal marijuana users constitute, in the aggregate, a sizable enough class to have a discernable, let alone substantial, impact on the national illicit drug market—or otherwise to threaten the CSA regime. Explicit evidence is helpful when substantial effect is not “visible to the naked eye.” See Lopez, 514 U. S., at 563. And here, in part because common sense suggests that medical marijuana users may be limited in number and that California’s Compassionate Use Act and similar state legislation may well isolate activities relating to medicinal marijuana from the illicit market, the effect of those activities on interstate drug traffic is not self-evidently substantial.

In this regard, again, this case is readily distinguishable from Wickard. To decide whether the Secretary could regulate local wheat farming, the Court looked to “the actual effects of the activity in question upon interstate commerce.” 317 U. S., at 120. Critically, the Court was able to consider “actual effects” because the parties had “stipulated a summary of the economics of the wheat industry.” Id., at 125. After reviewing in detail the picture of the industry provided in that summary, the Court explained that consumption of homegrown wheat was the most variable factor in the size of the national wheat crop, and that on-site consumption could have the effect of varying the amount of wheat sent to market by as much as 20 percent. Id., at 127. With real numbers at hand, the Wickard Court could easily conclude that “a factor of such volume and variability as home-consumed wheat would have a substantial influence on price and market conditions” nationwide. Id., at 128; see also id., at 128–129 (“This record leaves us in no doubt” about substantial effects).

The Court recognizes that “the record in the Wickard case itself established the causal connection between the production for local use and the national market” and argues that “we have before us findings by Congress to the same effect.” Ante, at 17 (emphasis added). The Court refers to a series of declarations in the introduction to the CSA saying that (1) local distribution and possession of controlled substances causes “swelling” in interstate traffic; (2) local production and distribution cannot be distinguished from interstate production and distribution; (3) federal control over intrastate incidents “is essential to effective control” over interstate drug trafficking. 21 U. S. C. §§801(1)–(6). These bare declarations cannot be compared to the record before the Court in Wickard.

They amount to nothing more than a legislative insistence that the regulation of controlled substances must be absolute. They are asserted without any supporting evidence—descriptive, statistical, or otherwise. “[S]imply because Congress may conclude a particular activity substantially affects interstate commerce does not necessarily make it so.” Hodel v. Virginia Surface Mining & Reclamation Assn., Inc., 452 U. S. 264, 311 (1981) (Rehnquist, J., concurring in judgment). Indeed, if declarations like these suffice to justify federal regulation, and if the Court today is right about what passes rationality review before us, then our decision in Morrison should have come out the other way. In that case, Congress had supplied numerous findings regarding the impact gender-motivated violence had on the national economy. 529 U. S., at 614; id., at 628–636 (Souter, J., dissenting) (chronicling findings). But, recognizing that “ ‘ “[w]hether particular operations affect interstate commerce sufficiently to come under the constitutional power of Congress to regulate them is ultimately a judicial rather than a legislative question,” ’ ” we found Congress’ detailed findings inadequate. Id., at 614 (quoting Lopez, supra, at 557, n. 2, in turn quoting Heart of Atlanta Motel, Inc. v. United States, 379 U. S. 241, 273 (1964) (Black, J., concurring)). If, as the Court claims, today’s decision does not break with precedent, how can it be that voluminous findings, documenting extensive hearings about the specific topic of violence against women, did not pass constitutional muster in Morrison, while the CSA’s abstract, unsubstantiated, generalized findings about controlled substances do?

In particular, the CSA’s introductory declarations are too vague and unspecific to demonstrate that the federal statutory scheme will be undermined if Congress cannot exert power over individuals like respondents. The declarations are not even specific to marijuana. (Facts about substantial effects may be developed in litigation to compensate for the inadequacy of Congress’ findings; in part because this case comes to us from the grant of a preliminary injunction, there has been no such development.) Because here California, like other States, has carved out a limited class of activity for distinct regulation, the inadequacy of the CSA’s findings is especially glaring. The California Compassionate Use Act exempts from other state drug laws patients and their caregivers “who posses[s] or cultivat[e] marijuana for the personal medical purposes of the patient upon the written or oral recommendation of a physician” to treat a list of serious medical conditions. Cal. Health & Safety Code Ann. §§11362.5(d), 11362.7(h) (West Supp. 2005) (emphasis added). Compare ibid. with, e.g., §11357(b) (West 1991) (criminalizing marijuana possession in excess of 28.5 grams); §11358 (criminalizing marijuana cultivation). The Act specifies that it should not be construed to supersede legislation prohibiting persons from engaging in acts dangerous to others, or to condone the diversion of marijuana for nonmedical purposes. §11362.5(b)(2) (West Supp. 2005). To promote the Act’s operation and to facilitate law enforcement, California recently enacted an identification card system for qualified patients. §§11362.7–11362.83. We generally assume States enforce their laws, see Riley v. National Federation of Blind of N. C., Inc., 487 U. S. 781, 795 (1988), and have no reason to think otherwise here.

The Government has not overcome empirical doubt that the number of Californians engaged in personal cultivation, possession, and use of medical marijuana, or the amount of marijuana they produce, is enough to threaten the federal regime. Nor has it shown that Compassionate Use Act marijuana users have been or are realistically likely to be responsible for the drug’s seeping into the market in a significant way. The Government does cite one estimate that there were over 100,000 Compassionate Use Act users in California in 2004, Reply Brief for Petitioners 16, but does not explain, in terms of proportions, what their presence means for the national illicit drug market. See generally Wirtz, 392 U. S., at 196, n. 27 (Congress cannot use “a relatively trivial impact on commerce as an excuse for broad general regulation of state or private activities”); cf. General Accounting Office, Marijuana: Early Experience with Four States’ Laws That Allow Use for Medical Purposes 21–23 (Rep. No. 03–189, Nov. 2002), http://www.gao.gov/new.items/d03189.pdf (as visited June 3, 2005 and available in Clerk of Court’s case file) (in four California counties before the identification card system was enacted, voluntarily registered medical marijuana patients were less than 0.5 percent of the population; in Alaska, Hawaii, and Oregon, statewide medical marijuana registrants represented less than 0.05 percent of the States’ populations). It also provides anecdotal evidence about the CSA’s enforcement. See Reply Brief for Petitioners 17–18. The Court also offers some arguments about the effect of the Compassionate Use Act on the national market. It says that the California statute might be vulnerable to exploitation by unscrupulous physicians, that Compassionate Use Act patients may overproduce, and that the history of the narcotics trade shows the difficulty of cordoning off any drug use from the rest of the market. These arguments are plausible; if borne out in fact they could justify prosecuting Compassionate Use Act patients under the federal CSA. But, without substantiation, they add little to the CSA’s conclusory statements about diversion, essentiality, and market effect. Piling assertion upon assertion does not, in my view, satisfy the substantiality test of Lopez and Morrison.

III

We would do well to recall how James Madison, the father of the Constitution, described our system of joint sovereignty to the people of New York: “The powers delegated by the proposed constitution to the federal government are few and defined. Those which are to remain in the State governments are numerous and indefinite…. The powers reserved to the several States will extend to all the objects which, in the ordinary course of affairs, concern the lives, liberties, and properties of the people, and the internal order, improvement, and prosperity of the State.” The Federalist No. 45, pp. 292–293 (C. Rossiter ed. 1961).

Relying on Congress’ abstract assertions, the Court has endorsed making it a federal crime to grow small amounts of marijuana in one’s own home for one’s own medicinal use. This overreaching stifles an express choice by some States, concerned for the lives and liberties of their people, to regulate medical marijuana differently. If I were a California citizen, I would not have voted for the medical marijuana ballot initiative; if I were a California legislator I would not have supported the Compassionate Use Act. But whatever the wisdom of California’s experiment with medical marijuana, the federalism principles that have driven our Commerce Clause cases require that room for experiment be protected in this case. For these reasons I dissent.