LIPEZ, Circuit Judge, concurring and dissenting.

Although I agree with the majority that the district court's decision cannot stand, I respectfully disagree with the majority's refusal to address the First Amendment issue at the core of this case. The majority focuses on the so-called upstream transactions — the acquisition, aggregation, and sale of prescriber-identifiable data by the plaintiffs — and concludes that such activity is not speech within the purview of the First Amendment. That conclusion is self-evident and beside the point. In enacting the Prescription Information Confidentiality Act ("the Prescription Act" or "the Act"), the New Hampshire Legislature chose to regulate the upstream transactions because it wanted to alter the message used by pharmaceutical detailers in pursuing a downstream transaction with health care professionals. In other words, the Act was designed to limit the speech of those detailers. The majority relies on the prudential doctrine of standing to avoid deciding whether that limitation violates the First Amendment. In my view, that avoidance is wasteful and unwise, unsupported by principles of standing, and analytically flawed.

Consequently, after examining the issue of standing, I address the issue that we should be addressing — whether the Act restricts protected commercial speech between detailers and prescribers and, if so, whether the State can justify that restriction under the commercial speech test of Central Hudson Gas Electric Corp. v. Public Service Commission, 447 U.S. 557, 566, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980). I conclude that the Act does restrict commercial speech, and that the State's interest in cost containment justifies that restriction. I also conclude, contrary to the majority, that we should remand the case for consideration of the plaintiffs' Commerce Clause challenge.

I.

The majority admits that speech is implicated by the Prescription Act and identifies that speech as "primarily [the] communications between detailers and doctors." It purports to refuse to address the Act's impact on that targeted speech, based on principles of standing, because "no detailer or doctor is a plaintiff here." However, not only do my colleagues misguidedly invoke standing to avoid explicitly resolving the constitutionality of the Act's restriction on communications between detailers and doctors, but they also accept the State's justification for the restriction without allowing the plaintiffs to establish the First Amendment values at stake. The majority's use of standing principles is thus doubly wrong.

A. The Prudential Policies of Third Party Standing

In Craig v. Boren, 429 U.S. 190, 97 S.Ct. 451, 50 L.Ed.2d 397 (1976), the Supreme Court considered whether a beer vendor could challenge on equal protection grounds an Oklahoma statute that prohibited the sale of "nonintoxicating" 3.2% beer to males under 21 and to females under 18. The question was whether the beer vendor had standing to raise the equal protection objections of 18- to 20-year-old males. The Court noted that the plaintiff had the requisite "injury in fact" to satisfy the constitutional standing requirement, id. at 194, 97 S.Ct. 451, leaving only a prudential concern about whether the plaintiffs should be allowed to raise third-party constitutional claims.

In concluding that the vendor's claims could go forward, the Court observed that it is "settled that limitations on a litigant's assertion of jus tertii are not constitutionally mandated, but rather stem from a salutary 'rule of self-restraint' designed to minimize unwarranted intervention into controversies where the applicable constitutional questions are ill-defined and speculative." Id. at 193, 97 S.Ct. 451. However, in the circumstances before the Court in Craig, such "prudential objectives" could not be furthered because "the lower court already ha[d] entertained the relevant constitutional challenge and the parties ha[d] sought or at least ha[d] never resisted an authoritative constitutional determination." The Court continued:

In such circumstances, a decision by us to forgo consideration of the constitutional merits in order to await the initiation of a new challenge to the statute by injured third parties would be impermissibly to foster repetitive and time-consuming litigation under the guise of caution and prudence. Moreover, insofar as the applicable constitutional questions have been and continue to be presented vigorously and "cogently," the denial of jus tertii standing in deference to a direct class suit can serve no functional purpose.

Id. at 193-94, 97 S.Ct. 451 (citation omitted).

There is no debate that the plaintiffs in this case also meet the requirements for Article III standing. Like the beer vendors in Craig, the plaintiffs here are direct targets of the challenged statute. By seeking to prevent pharmaceutical detailers from using prescriber data in their sales pitches to New Hampshire health care providers, the Act diminishes the market for the prescriber data collected, organized and sold by plaintiffs and thereby inflicts "a direct economic injury through the constriction of [the] buyers' market." 429 U.S. at 194, 97 S.Ct. 451. Thus, as in Craig, only the prudential standing doctrine is at issue, and here, too, pragmatic considerations are paramount. The district court heard evidence from about a dozen witnesses and considered voluminous other materials in preparing its thoughtful and comprehensive decision. Nothing in the extensive record even hints that the plaintiffs were unable or unwilling to aggressively litigate the First Amendment issues at stake in the "downstream" transactions between the detailers and physicians. Such an inability or unwillingness would counsel prudence in resolving the First Amendment issues raised by those transactions without the participation of the pharmaceutical companies or doctors. But here the First Amendment issues raised by the exchanges between detailers and physicians were explored exhaustively.

Moreover, the district court expressly confronted the question of third-party standing before proceeding with the case. The court told the parties that, if the State sought to invoke standing as a barrier to full resolution of the action, it would stay the case for thirty days to allow intervention by a pharmaceutical company. The court explained:

[I]t's very clear you are working closely with the pharmacy companies here. They don't want to be the ones to stand up and fight the doctors. They want you to do it. We all know what's going on here, and the reality is if they have to, they will come out from behind the scenes and get out into the forefront, because they want this information, and they want you to be fighting the battle for them. But if we have to, we'll get them in here. I just don't think it really matters.

So the state should think about that. If you want to fight on that issue, that's what I would do. I would first do an argument on third-party standing. If I think there's any issue with third-party standing, if the plaintiff asked for it, I will give them 30 days to amend to bring in a new plaintiff pharmacy company, in which case it seems to me the third-party standing argument disappears.

I didn't think we were going to be talking about third-party standing today, since it's not really raised in the briefs now. But if you want to press that, I think we'll have to deal with it that way.

(Emphasis added.) The Attorney General then said that "we don't intend to press that at this time." The issue was not addressed by either party on appeal.

In these circumstances, as in Craig, "a decision . . . to forgo consideration of the constitutional merits in order to await the initiation of a new challenge to the statute by injured third parties would be impermissibly to foster repetitive and time-consuming litigation under the guise of caution and prudence." 429 U.S. at 193-94, 97 S.Ct. 451. The prudence invoked by the majority serves no purpose and it ignores the judgment of the district court, based on its immersion in the details of the case, that the absence of the pharmaceutical companies as parties did not compromise the proper adjudication of the case.

I recognize that the Supreme Court's precedent on third-party standing since Craig, as well as our own precedent, set out a formal three-prong inquiry that could not be satisfied here because, as the majority observes, there is no indication in the record that pharmaceutical companies or health care providers who prescribe medication are unable to assert their own rights. See, e.g., Kowalski v. Tesmer, 543 U.S. 125, 129-30, 125 S.Ct. 564, 160 L.Ed.2d 519 (2004); Powers v. Ohio, 499 U.S. 400, 410-11, 111 S.Ct. 1364, 113 L.Ed.2d 411 (1991); Wine Spirits Retailers, Inc. v. Rhode Island ( Wine Spirits I), 418 F.3d 36, 49 (1st Cir. 2005). However, none of those cases suggests that the pragmatic factors emphasized by the Court in Craig no longer have force in comparable circumstances.

The prudential limitations on standing were designed to "add to the constitutional minima a healthy concern that if the claim is brought by someone other than one at whom the constitutional protection is aimed, the claim not be an abstract, generalized grievance that the courts are neither well equipped nor well advised to adjudicate." Sec'y of State of Md. v. Joseph H. Munson Co., 467 U.S. 947, 955 n. 5, 104 S.Ct. 2839, 81 L.Ed.2d 786 (1984); see also Miller v. Albright, 523 U.S. 420, 446, 118 S.Ct. 1428, 140 L.Ed.2d 575 (1998) (O'Connor, J., concurring) (stating that the requirement that a litigant assert his own legal rights "arises from the understanding that the third-party rightholder may not, in fact, wish to assert the claim in question, as well as from the belief that 'third parties themselves usually will be the best proponents of their rights'") (citation omitted). The Supreme Court has recognized that the "lessening" of these limitations may be justified where other concerns, such as the danger of chilling free speech, are more pressing, Munson, 467 U.S. at 956, 104 S.Ct. 2839, or where, as in Craig, such limitations do not serve the purpose for which they were designed.

Indeed, the Court in Tesmer conceded that it had been "quite forgiving with the[] criteria [for third-party standing] in certain circumstances," and identified the context of the First Amendment as one in which flexibility may be warranted. Tesmer, 543 U.S. at 130, 125 S.Ct. 564. In Munson, the Court described its conclusion to allow third-party standing in terms also applicable here: "The activity sought to be protected is at the heart of the business relationship between [the plaintiff] and its clients, and [the plaintiffs] interests in challenging the statute are completely consistent with the First Amendment interests of the [third parties] it represents. We see no prudential reason not to allow it to challenge the statute." 467 U.S. at 958, 104 S.Ct. 2839. Thus, notwithstanding the Court's more detailed articulation of the third-party standing inquiry since Craig, see Miller, 523 U.S. at 447, 118 S.Ct. 1428 (O'Connor, J., concurring), the pragmatic considerations highlighted in that decision remain relevant.

This case illustrates the importance of pragmatism. There is no reason to reject the district court's decision to proceed without a pharmaceutical company as a plaintiff unless that decision would result in a trial of the "generalized grievance that the courts are neither well equipped nor well advised to adjudicate," Munson, 467 U.S. at 955 n. 5, 104 S.Ct. 2839. The reality is that the court and the parties have expended substantial time, resources and energy to address comprehensively the First Amendment issue at the heart of this case. That issue has been vigorously tried and thoughtfully adjudicated. Given our authority to review the court's entire judgment, it is imprudent to avoid that issue.

B. The Unavoidable Issue

The majority's analysis reveals yet another reason why its reliance on standing is inappropriate. In the first part of its analysis, the majority finds no constitutional flaw in the Act's restriction on "certain information exchanges" because those transfers "are not . . . the sorts of exchanges valued by the Supreme Court's First Amendment jurisprudence." However, to reach that conclusion, the majority considers the societal benefits of a particular form of detailing — the very speech that it claims is beyond the scope of this appeal.

My colleagues insist that the limited scope of review "does not prevent consideration of New Hampshire's interest in combating detailing." I do not understand how the majority can have it both ways. If the constitutionality of the Act's impact on the detailers' speech is off limits in this case because a pharmaceutical company is not a party, how can the majority make a judgment about the low value of that speech in deciding that the Act regulates only conduct and not speech? Surely we must consider the plaintiffs' First Amendment contentions before concluding that the upstream information "exchanges" that make the speech possible are not worthy of First Amendment protection.

This inconsistency pervades the majority's decision. After making judgments about the nature of the detailing transaction and how it increases the likelihood that physicians will prescribe more expensive drugs, the majority asserts that "the legislature sought to level the playing field not by eliminating speech but, rather, by eliminating the detailers' ability to use a particular informational asset — prescribing histories — in a particular way." (Emphasis added.) Here the majority is characterizing the speech interest that is supposedly beyond the scope of its opinion, and characterizing it incorrectly. The very elimination of the detailers' ability to use "a particular informational asset" restricts the message they are allowed to disseminate and implicates the free speech concerns of the First Amendment.

Moreover, in discussing its alternative holding, which treats the plaintiffs' upstream transactions as speech subject to the First Amendment rather than conduct, the majority weighs the value of detailing, based on the regulated data, against the Legislature's policy objectives and the harms identified by the government. Again, the majority's conclusion that the Act does not violate the First Amendment rests on a judgment about the speech — i.e., the detailing — that the majority purports to place off limits for analysis. For example, the majority points to "substantial evidence" in the record

that, in several instances, detailers armed with prescribing histories encouraged the overzealous prescription of more costly brand-name drugs regardless of both the public health consequences and the probable outcome of a sensible cost/benefit analysis. By contrast, the record contains no evidence that in the absence of detailing, physicians have tended to prescribe generic drugs more often than either their patients' health or their patients' pocket-books warranted.

The majority ultimately concludes that "the state adequately demonstrated that the Prescription Information Law is reasonably calculated to advance its substantial interest in reducing overall health care costs within New Hampshire."

Thus, the majority does what it says standing doctrine forbids: it evaluates the Act based on the law's impact on the speech between detailers and prescribers. The majority's approach is hardly surprising given that this speech was the Act's target. What is surprising is the majority's failure to appreciate that reliance on standing principles is misplaced where, as here, the issue that the majority seeks to avoid is unavoidable. Although ostensibly limiting its First Amendment inquiry to the upstream transactions — the acquisition, aggregation, and sale of prescriber-identifiable data — and deciding in its primary holding that these transactions involve conduct only, the majority makes judgments about the nature, value, and consequences of the speech that occurs in the downstream transactions between detailers and doctors. As the majority discovered, it is impossible to assess the constitutionality of the Act without factoring in the Legislature's specific objective to limit the speech of the detailers.

Moreover, there is no reason to think that the majority's judgments about the statute would change in a case where a pharmaceutical company was a plaintiff. All of the relevant considerations were explored by the district court. They have similarly been explored in the majority's analysis because the majority could not characterize the upstream transactions as merely conduct without making judgments about the value of the "downstream" speech between the detailers and the doctors.

Thus, both the practicalities of this litigation and the nature of the First Amendment issue require that the case be analyzed as the parties tried it and the district court decided it. I therefore proceed with that analysis. Although my discussion will at times overlap with the majority's, I have chosen to present my complete view of the record and the governing law. The First Amendment question here is both important and close, and I wish to fully explain why, in the end, I conclude that the district court erred in declaring the Prescription Act unconstitutional.

II.

In recounting the background of this case, I draw heavily on the comprehensive and thoughtful recitation of the facts set out by the district court. See IMS Health Inc. v. Ayotte, 490 F.Supp.2d 163, 165-74 (D.N.H. 2007). Those facts are largely undisputed; the parties primarily contest their legal significance.

A. Pharmaceutical Sales and Marketing

More than three billion prescriptions are written each year by doctors and other licensed health care professionals, covering approximately 8,000 different pharmaceutical products. These prescriptions are filled by approximately 54,000 retail pharmacies; in 2004, such retail prescription sales totaled $168 billion. In an effort to increase and protect their share of this vast market, pharmaceutical companies engage in various promotional activities. The public is most familiar with direct-to-consumer advertising, in which the drug companies tout the virtues of their products in television commercials and other media, typically urging consumers to ask their doctors for the advertised drugs. However, the bulk of the drug companies' promotional efforts are aimed directly at physicians and other prescribers. The primary method for such promotion is detailing, which usually is accompanied by the provision of free drug samples that prescribers can distribute to patients. As inducements to increase their access to physicians who are sometimes reluctant to meet with them, detailers also frequently offer free meals and other gifts to the doctors and their staffs. As I shall explain, these practices are both widely used and widely criticized.

1. Detailing

Detailing is the face-to-face advocacy of a product by sales representatives who visit doctors' offices and hospitals to meet with the prescribing health care professionals. Although the objective of these visits is to make sales, detailers often provide valuable information about the drugs they are selling. Doctors may be alerted by a detailer to tests showing the risk of a drug interaction or a drug's side effects. One survey showed that most physicians meet with pharmaceutical representatives about four times a month. See Ashley Wazana, Physicians and the Pharmaceutical Industry: Is a Gift Ever Just a Gift?, 283 J. Am. Med. Ass'n 373, 375 (Jan. 19, 2000). Consumers Union has reported re-search showing many more encounters: "'[T]he average primary care physician interacts with 28 sales representatives each week; the average specialist interacts with 14.'" Consumers Union, Prescription for Change, http://www.consumersunion.org/pdf/drugreps.pdf (March 2006) (quoting research from Health Strategies Group). Whatever the frequency, it is undisputed that pharmaceutical detailing plays a substantial role in the dissemination of information about drugs to physicians.

Detailing focuses primarily on brand-name drugs that are entitled to patent protection. Once a patent expires, competitors may obtain approval to sell generic bioequivalent versions of the drug, which are equally effective for most patients but usually much less expensive than their brand-name counterparts. New Hampshire law provides that pharmacies may substitute a bioequivalent generic drug for a brand-name drug unless the prescriber specifies that the brand-name drug is "medically necessary." N.H.Rev.Stat. Ann. § 318:47-d (2003). Thus, once bioequivalent generic drugs become available, sales of the related brand-name drug tend to fall and detailing is no longer considered a cost-effective marketing technique. However, non-bioequivalent options also are available for some medical conditions, and the drug companies aggressively market to urge physicians to choose their patented brand-name medications over such alternatives. Thus, it is this choice — between a still-under-patent, branded drug and a similar, but biologically different generic medication — that is at the heart of this case.

As I will discuss below, studies indicate that detailing has "a significant effect on physician prescription behavior." Puneet Manchanda Elisabeth Honka, Symposium — Pharmaceutical Innovation and Cost: An American Dilemma: The Effects and Role of Direct-to-Physician Marketing in the Pharmaceutical Industry: An Integrative Review, 5 Yale J. Health Pol'y, L. Ethics 785, 809 (Summer 2005) ("While there seems to be little consensus about the size of the effect, it is clear that the effect is positive and significant in a statistical sense.").

2. Samples and Other Perks

Free samples and courtesy gifts are routinely given by detailers as part of their sales visits, and they are important tools in pharmaceutical marketing. Doctors rely on receiving drug samples that they can distribute to patients who are unable to afford the high cost of some medications. Keeping office doors open to detailers ensures that the doctors will have a continued supply of samples, and some physicians are therefore reluctant to restrict detailing. Even when drug cost is not an issue, the free samples are helpful to physicians who want to test new remedies before committing to them. A patient's positive results during a trial period may lead to a long-term prescription — the dettailer's desired outcome. En route to that objective, however, the free samples have provided access to helpful treatment that patients otherwise may not have received. The cost of the samples distributed annually by pharmaceutical representatives has been estimated at more than $11 billion.

It is not only the patients who benefit from the drug companies' largess, however. Physicians and other medical office staff members frequently receive "good will" gifts from detailers, including office supplies, free meals, and conference travel funding — perks that are designed to encourage long-term relationships with, and loyalty toward, the detailers. Studies have shown that these sorts of gifts can have a subtle effect on physicians, and, because they typically are unrelated to the provision of medical care, they have come under particular fire by both consumer advocates and medical professionals themselves. The Pharmaceutical Research and Manufacturers of America ("PhRMA") in 2002 adopted a voluntary code governing interactions with health care professionals that discourages such inducements

unless either the value of what is provided is insubstantial (less than $100) and the inducement is primarily for the benefit of patients, or the value of the inducement is minimal and the inducement is directly related to the provider's practice. For example, an occasional gift of a stethoscope is acceptable under the Code because it is not deemed to be of substantial value and the gift benefits patients. In contrast, an unrestricted gift certificate to a local bookstore may not be offered under the Code regardless of its value because it does not benefit patients and is unrelated to the health care professional's practice. The Code draws similar distinctions with respect to meals and entertainment.

490 F.Supp.2d at 168-69 (citations omitted).

3. Data Mining and Prescriber Profiles

When detailers enter medical offices to market their products, they are equipped not only with detailed information about the drugs they are attempting to sell but also with considerable knowledge about their audience. Much of that prescriber information is supplied by the plaintiffs and similar companies, who play a crucial behind-the-scenes role in the flirtation between pharmaceutical sales representatives and prescribers. These so-called "data mining" companies collect and organize information about doctors and their prescribing patterns, converting information gleaned from "thousands of sources" into a commodity for which the pharmaceutical industry pays substantial sums. From retail pharmacies and other entities, such as insurers, that acquire the data as part of the business they conduct, the data miners obtain information on every pharmaceutical sale, including the form, strength and dosage of the drug, the amount dispensed, and the name and address of the prescriber. The information includes an identifying code for each patient, although the patient is not personally identified. From other sources, including the American Medical Association, the plaintiffs obtain information about individual prescribers and their specialities.

The data mining companies weave the information together to produce, among other databases, "prescriber profiles" — individualized reports on the prescriptions being written by particular doctors. The information is then sold to third parties for various commercial uses, including pharmaceutical marketing, and also is provided at no charge for nonprofit purposes, such as academic and medical research. The data provide a historical view of a physician's prescribing practices, allowing the pharmaceutical companies to identify doctors who have displayed a willingness to try new products (the "early adopters") and to target doctors whose drug choices they seek to change. With knowledge of the physicians' prescribing history, the detailers are able to tailor their messages to those doctors' specific circumstances — for example, emphasizing the potential side effects of a competitor's brand-name product that the detailer knows the doctor has been using, or highlighting the advantages of the detailers' branded drug over the generic alternative the doctor routinely prescribes. The detailer's verbal message in favor of the brand-name drug may be furthered by the provision of free samples of the medication, encouraging what is initially a "no-cost" switch to the more expensive drug. The companies also use reports obtained shortly after detailing visits to assess whether the sales calls had an effect on the targeted prescribers' drug choices. The detailer's compensation is sometimes tied to the success of his or her efforts.

This use of prescriber-identified data has drawn sharp criticism on many fronts, including among physicians who object both to the disclosure of information they deem confidential and to the hard-sell messages delivered by detailers who may know more about their prescribing habits than do the doctors themselves. In 2006, the AMA responded to the concerns by initiating the Prescribing Data Restriction Program ("PDRP"), which allows physicians to restrict access to their prescribing data by pharmaceutical detailers. The AMA also developed guidelines for the use of prescribing data "to provide ethical guidance to the healthcare industry." The guidelines urge that companies, inter alia, "[c]ontinually reinforce that use of prescribing data to overtly pressure or coerce physicians to prescribe a particular drug is absolutely an inappropriate use." Neither the PDRP nor the guidelines have quelled the concerns. The PDRP has been criticized because prescriber information will be withheld only if doctors affirmatively opt out, and the opt-out choice must be renewed every three years. Voluntary guidelines are seen as insufficient to offset the commercial incentives to use the information. Some states, like New Hampshire, turned to legislation to address the concerns.

B. New Hampshire's Statutory Response

The Prescription Act prohibits the transmission or use of both patient-identifiable and prescriber-identifiable data for certain commercial purposes. Violators are subject to both criminal and civil penalties. N.H.Rev.Stat. Ann. § 318:55. In pertinent part, the statute provides:

Records relative to prescription information containing patient-identifiable and prescriber-identifiable data shall not be licensed, transferred, used, or sold by any pharmacy benefits manager, insurance company, electronic transmission intermediary, retail, mail order, or Internet pharmacy or other similar entity, for any commercial purpose, except for the limited purposes of pharmacy reimbursement; formulary compliance; care management; utilization review by a health care provider, the patient's insurance provider or the agent of either; health care research; or as otherwise provided by law. Commercial purpose includes, but is not limited to, advertising, marketing, promotion, or any activity that could be used to influence sales or market share of a pharmaceutical product, influence or evaluate the prescribing behavior of an individual health care professional, or evaluate the effectiveness of a professional pharmaceutical detailing sales force.

In effect, the statute prohibits the use of prescriber-identifiable data for all purposes related to detailing, but seeks to preserve access to the data for other uses — including other commercial purposes. I agree with the district court that the prohibited uses are narrowly defined and that the statute does not, for example, prohibit pharmaceutical companies from using prescriber-identifiable data for their own research. See 490 F.Supp.2d at 171.

1. Legislative History

In introducing the proposed legislation at a hearing before the Senate Committee on Executive Departments and Administration, Representative Cindy Rosenwald, one of the statute's co-sponsors, explained that it had two goals: "It will protect privacy and it will save money for the state, for consumers and businesses. It will accomplish these goals by prohibiting the sale or use of individual patient or prescriber identity for marketing brand name prescription drugs." A written attachment to her testimony, which included a section entitled "What H.B. 1346 will do," states that the law will, inter alia, "[h]opefully reduce the prescription drug costs for patients, employers the State Medicaid program."

About sixteen individuals testified at the hearing. A representative of the Department of Health and Human Services, Gregory Moore, emphasized both the privacy and cost reduction purposes of the legislation. He described the prescriber data as the physicians' "trade secrets" and further stated:

The Department also believes that these activities ultimately drive up the cost of prescription drugs and the cost of health care in the aggregate. Since no other state has passed legislation like this, it would be hard for us to quantify what that impact might be, but I find it unlikely the drug companies are sending details into doctors' offices for the purpose of selling doctors cheaper medication. In fact, I'm confident that, if you're a doctor, that one of the best ways to get a detailer into your office would be if you switched to prescribing a generic drug over a brand drug.

Also testifying in favor of the legislation was the president-elect of the New Hampshire Medical Society, Dr. Seddon Savage, who said the law "will deter marketing intended to manipulate the practice of individual physicians that is intended to increase market share for the individual companies, possibly at the expense of appropriate decision making for the patients." He further stated that "[n]umerous studies have shown that . . . [doctors'] decision making can be and sometimes is shaped by marketing efforts."

Savage's general testimony was reinforced by comments from Dr. Marc Sadowsky, a psychiatrist and the president of the New Hampshire Medical Society. He reported a phone conversation with a patient who said that her primary care doctor had thought a brand-name medicine might be better for her than the generic she was using. Sadowsky continued:

I said, "Well, you're doing fine on the generic and your co-pay is going to go up $40 a month, $500 a year. So, it is not entirely clear to me why we're doing this." . . . I think that that was an example of the primary care physician having been marketed to directly and didn't really have a clinical reason for doing it except that that was the last drug rep who came to see him and said this is a better medicine for anxiety, even though the person was asymptomatic at the time.

In Sadowsky's view, there was "no apparent reason" for the requested switch "except presumably that [the doctor] ha[d] been marketed to effectively."

Among those speaking against the statute was a representative of the New Hampshire Association of Chain Drug Stores, Stuart Trachy, who described the proposed legislation as "too broad" and observed that "the opt out program that the AMA is going to be instituting should take care of the concerns that we have heard in terms of specific doctors being concerned that their prescribing data is out there." A spokesman for plaintiff IMS, Robert Hunkler, stated that restricting prescriber-identifiable information would not lower health care costs because "pharmaceutical companies will[] in all likelihood continue to send sales reps to all doctors without the ability to more specifically hone in on the right people with the right message. It will likely incur more costs to the system." Hunkler also predicted that the acknowledged beneficial uses of the data, including medical research, would be compromised because the information would no longer be readily available. Responding to complaints from doctors that drug companies "know more about [their] prescribing behavior than [they] know," Hunkler stated that IMS was working toward greater access: "[W]e think that a preferable solution is to provide this information to doctors, to health researchers and others instead of turning out the light and taking it away from everyone." The American Medical Association also expressed opposition to the legislation, commenting in a prepared statement that the PDRP would "provide[] physicians with the tools they need to restrict information that they do not want shared while avoiding legislatively-mandated restrictions that could have unintended consequences."

2. Legislative Action and Legal Challenge

The Prescription Act was approved by the Legislature in May 2006, and it took effect on June 30 of that year. Four weeks later, on July 28, 2006, IMS and Verispan filed the complaint in this case, alleging that the Act violated the First Amendment and the Commerce Clause, and that it was void for vagueness and overbreadth. They sought declaratory and injunctive relief against the statute's enforcement. Meanwhile, in compliance with the Act, Verispan modified its databases so that it could identify and suppress all prescriber-identifiable data from New Hampshire prescriptions before the information was released to third parties. IMS also stopped selling prescriber-identifiable information obtained from New Hampshire sources to third parties.

During a four-day bench trial in January and February 2007, the court heard live testimony from ten witnesses, most of whom were physicians. A former detailer and a representative of each plaintiff also testified. The parties also submitted voluminous written materials, including a number of journal articles describing studies on detailing. The State highlighted the testimony of Dr. Jerry Avorn, a professor at Harvard Medical School whose research focuses on the use of prescription drugs and their outcomes, and who also works at Brigham and Women's Hospital in the Division of Pharmacoepidemiology and Pharmacoeconomics. Through Avorn's testimony on the medical literature and the testimony of practitioners who recounted specific experiences with detailing, the Attorney General sought to show that detailing in general, and use of prescriber-identifiable data in particular, influences physicians to prescribe brand-name drugs more frequently than would occur with "evidence-based" decision-making that was untainted by the detailers' marketing messages. The Attorney General asserted that the Act advanced the State's substantial interests in prescriber privacy, public health and cost-containment.

On their behalf, the plaintiffs elicited considerable testimony about the beneficial aspects of detailing and the use of prescriber-identifiable data to target physicians. For example, Dr. Thomas Wharton, Jr., director of cardiology at Exeter Hospital, testified that discussions initiated by drug company representatives provide "a very stimulating forum" for discussing the treatment of coronary disease. He also stated that the "level of discourse is elevated" when a drug representative knows his prescribing habits: "[I]f they know that I'm a user of the drug, they will direct what they have to say to me toward any brand-new information that might have come out rather than starting with the basics. If they know that I'm a user of a drug, I would think that they are more likely to come to me if a new adverse effect is announced regarding that drug." Plaintiffs also emphasized the lack of evidence showing that restriction of prescriber-identifiable data would lead to a decrease in drug costs and attempted to show that less efficient detailing would result, potentially increasing the pharmaceutical companies' marketing costs and, in turn, increasing the cost of their products.

C. The District Court's Decision

On April 30, 2007, the district court ruled that the Prescription Act impermissibly restricted commercial speech and therefore violated the First Amendment. It rejected the Attorney General's argument that the Act targeted only unprotected factual information rather than constitutionally protected speech and also rejected her contention that the statute regulated only non-speech "uses" of the prescriber-identifiable data. Having concluded that the Act restricted protected commercial speech, the court examined whether the Attorney General had sufficiently justified the regulation under the three-part inquiry set out in Central Hudson Gas Electric Corp. v. Public Service Commission, 447 U.S. 557, 566, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980).

Under Central Hudson, truthful commercial speech that does not promote unlawful activity may be limited only if it "(1) is in support of a substantial government interest, (2) 'directly advances the governmental interest asserted,' and (3) 'is not more extensive than is necessary to serve that interest.'" El Dia, Inc. v. P.R. Dep't of Consumer Affairs, 413 F.3d 110, 113 (1st Cir. 2005) (quoting Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343). The district court considered the State's asserted interests in protecting prescriber privacy, promoting public health, and containing health care costs. It concluded that the record did not reveal a distinct privacy interest that was supported by the Act and held that neither the public health interest nor the interest in containing health care costs was directly advanced by the statute.

In addition, the court found a "fundamental flaw" in the Attorney General's argument that the regulation was necessary because "pharmaceutical companies manipulate health care providers by using prescriber-identifiable data to enhance the effectiveness of highly persuasive but truthful commercial speech." 490 F.Supp.2d at 181. Instead of restricting such information, the court stated, "if the State is concerned that truthful detailing is causing health care providers to make inadvisable prescribing decisions, 'the remedy to be applied is more speech, not enforced silence.'" Id. (quoting Whitney v. California, 274 U.S. 357, 377, 47 S.Ct. 641, 71 L.Ed. 1095 (1927) (Brandeis, J., concurring)).

The court also addressed the third Central Hudson prong and found that the State could advance its health and cost-containment interests, and specifically the unnecessary prescription of brand-name drugs, without restricting protected speech. The court noted that the State could, inter alia, directly limit the samples and gifts given to prescribers and their staffs, educate health care providers about the health and cost implications of their prescribing decisions, require health care providers to participate in continuing education programs offering objective information about the advantages and disadvantages of different drug choices, or adopt a Medicaid pharmacy program that takes cost considerations into account.

Accordingly, the court held that the statute could not be enforced "to the extent that it purports to restrict the transfer or use of prescriber-identifiable data." Id. at 183. It therefore granted the plaintiffs' request for declaratory relief and a permanent injunction. It did not reach their vagueness or Commerce Clause arguments.

III.

The Attorney General continues to argue on appeal that the Prescription Act restricts only the use of information and that this regulation of non-expressive conduct does not implicate the First Amendment. From the Attorney General's perspective, the statute regulates a commercial transaction and not protected speech. See generally Neil M. Richards, Reconciling Data Privacy and the First Amendment, 52 UCLA L.Rev. 1149, 1194 (2005) (concluding that restrictions on use of consumer data to target advertisements were "not a regulation of speech at all, but rather a regulation of information use — the business activity of deciding to whom to market products"). At trial, the Attorney General contended that the Act did not restrict the content of the pharmaceutical manufacturers' advertising or marketing messages, which she acknowledges would trigger First Amendment scrutiny. Rather, the legislature made the "unusual" — and in the Attorney General's view — permissible choice "to strike at the source of the information," Day 1, AM Session, at 45, thereby regulating the distribution and use of a "commodity" rather than limiting a speaker's message.

Like the district court, I think this argument attempts to create a dividing line that does not exist in the factual context of this case. While the statute explicitly prohibits any "use" of prescriber-identifiable data, one of the Legislature's desired outcomes is the modification of the marketing messages communicated by pharmaceutical detailers. See, e.g., Defendant's Memorandum of Law in Support of its Objection to Plaintiffs Motion for Preliminary Injunction, at 30-31 ("By prohibiting the license, transfer, use, or sale of prescriber-identifiable prescription data for commercial purposes, the Act prevents pharmaceutical companies from using that information to pressure physicians into changing their prescriptions from less costly medications to name brand drugs for reasons unrelated to the clinical needs of patients."). The State has attempted to insulate this expression-based intention from First Amendment scrutiny by directing its legislation to an earlier step in the communicative process. However, it may not skirt the Constitution's requirements in such fashion. Indeed, the Attorney General seeks to minimize the impact of the Act by emphasizing that detailers may continue to use the same face-to-face marketing approach with physicians, notwithstanding the Prescription Act. But if the State acknowledges that the form of marketing conduct remains the same (i.e., face-to-face promotion by detailers), it is difficult to see how the statute may be viewed solely as a regulation of the commercial transaction itself, rather than as a limitation on the content of the expression that may be used to conduct that transaction. See U.S. West, Inc. v. FCC, 182 F.3d 1224, 1232 (10th Cir. 1999) (finding that prohibition of telecommunications companies' use of customer proprietary data for targeted marketing constitutes a restriction on protected commercial speech).

I recognize that there are three separate commercial activities involved here: first, the transfer of the data to data miners, including the plaintiffs, from the entities that acquire prescription information in the ordinary course of their businesses (such as pharmacies and insurance companies); second, the transfer of the data in aggregated form from the plaintiffs to the pharmaceutical companies; and, third, the marketing of drugs to prescribers by detailers whose sales pitches make use of the data. To serve its interests in protecting privacy, promoting public health and containing health care costs, the Legislature targeted the content of the message communicated in the third transaction. The statute restricts that message indirectly by imposing restrictions on the first two transactions. Because the statute's purposes are linked to the third transaction, I conclude — as did the district court — that the assessment of the statute's impact must be similarly focused. See IMS Health, 490 F.Supp.2d at 176 ("The law is . . . squarely aimed at speech that proposes a commercial transaction even though it does not explicitly bar such speech."); Boos v. Barry, 485 U.S. 312, 321, 108 S.Ct. 1157, 99 L.Ed.2d 333 (1988) (noting that "[r]egulations that focus on the direct impact of speech on its audience" must be viewed as speech-based for purposes of First Amendment analysis).

The Attorney General asserts that the Supreme Court drew "a sharp distinction" in Bartnicki v. Vopper, 532 U.S. 514, 121 S.Ct. 1753, 149 L.Ed.2d 787 (2001), between regulating the use of information — which she claims does not implicate the First Amendment — and regulating its disclosure. In Bartnicki, the Court held that the First Amendment protected a reporter's disclosure of the contents of an illegally intercepted communication about a matter of public interest. Id. at 518, 121 S.Ct. 1753. In its discussion, the Court described a prohibition against the "use" of the contents of an illegal wiretap as "a regulation of conduct," while holding that a prohibition against the "disclosure" of such material "is fairly characterized as a regulation of pure speech." Id. at 526-27, 121 S.Ct. 1753. The Attorney General seizes on this language to argue that the Prescription Act and its prohibition against "use" of prescriber-identifiable data is similarly immune from First Amendment attack. However, the examples of prohibited "uses" listed by the Court in Bartnicki are materially different from the prohibition at issue here. They involve conduct in which the impact on speech is non-existent or, at most, incidental — for example, using unlawfully intercepted information about a business rival to create a competing product or using illegally recorded information to trade in securities or for extortion. Id. at 527 n. 10, 121 S.Ct. 1753. Here, by contrast, the prohibited "use" at issue is the dissemination of a commercial message through marketing, advertising or promotion — expressions that unquestionably are entitled to First Amendment protection. See Thompson v. W. States Med. Ctr., 535 U.S. 357, 366-67, 122 S.Ct. 1497, 152 L.Ed.2d 563 (2002) (quoting Va. State Bd. of Pharmacy, 425 U.S. at 763, 96 S.Ct. 1817, for the proposition "that a 'particular consumer's interest in the free flow of commercial information . . . may be as keen, if not keener by far, than his interest in the day's most urgent political debate'").

The multi-step nature of the statutory prohibition — imposing the restraint on the providers of the underlying information rather than directly on the communicator of the message — does not remove that protection. Supreme Court precedent establishes that where the goal of a regulation relates to suppression of expression, even a restriction that indirectly achieves that objective may run afoul of the First Amendment. See Grosjean v. Am. Press Co., 297 U.S. 233, 249, 56 S.Ct. 444, 80 L.Ed. 660 (1936) (invalidating a license tax on publications with circulations of 20,000 or more that sold advertising "because, in light of its history and of its present setting, it is seen to be a deliberate and calculated device in the guise of a tax to limit the circulation of information to which the public is entitled"); see generally Minneapolis Star Tribune Co. v. Minnesota Comm'r of Revenue, 460 U.S. 575, 581, 103 S.Ct. 1365, 75 L.Ed.2d 295 (1983) (holding unconstitutional a tax on newsprint and ink used in the production of newspapers).

By contrast, legislation whose purpose is to regulate economic conduct, and which only incidentally affects speech, typically does not raise First Amendment concerns. See generally Rumsfeld v. Forum for Acad. Inst. Rights, Inc., 547 U.S. 47, 62, 126 S.Ct. 1297, 164 L.Ed.2d 156 (2006) (" FAIR") ("'[I]t has never been deemed an abridgement of freedom of speech or press to make a course of conduct illegal merely because the conduct was in part initiated, evidenced, or carried out by means of language, either spoken, written, or printed.'") (quoting Giboney v. Empire Storage Ice Co., 336 U.S. 490, 502, 69 S.Ct. 684, 93 L.Ed. 834 (1949)). Our circuit considered this principle at some length in two related decisions concerning a Rhode Island statute regulating the retail sale of alcohol. See Wine Spirits Retailers, Inc. v. Rhode Island, 481 F.3d 1, 6-7 (1st Cir. 2007) (" Wine Spirits II"); Wine Spirits Retailers, Inc. v. Rhode Island, 418 F.3d 36, 48-49 (1st Cir. 2005) (" Wine Spirits I"). Although the State relies on the Wine Spirits decisions in arguing that the Prescription Act falls outside the First Amendment's scope, those cases support a contrary conclusion.

The regulation at issue in Wine Spirits originally prohibited any "chain store organization" from holding a Class A retail liquor license, but gave the Department of Business Regulation the discretion to determine whether a business was a "chain store." Some businesses were evading the restriction by adopting chain-store-like features within a different business structure, described as "franchised package stores." The State responded by amending the statute to identify the specific conduct it sought to prohibit; i.e., it defined the term "chain store organization" to include businesses that participated in "a coordinated or common advertisement with one or more liquor licensed business in any advertising media" or that coordinated marketing strategies. At the same time, the State adopted a provision explicitly excluding franchisees from holding Class A liquor licenses. Wine Spirits had been operating as a franchisor of independently owned liquor retailers and, among other activities, provided marketing, advertising and business advice and services. In the first of the two cases, Wine Spirits claimed that the regulation improperly infringed on its right to communicate with its customers by, for example, designing advertisements and arranging for their placement in various media. Wine Spirits I, 418 F.3d at 49. In the second case, we also considered a claim by Wine Spirits' franchisees that the regulation imposed an improper limitation on the content of their advertising. Wine Spirits II, 481 F.3d at 6.

We found no First Amendment issue in either instance. In the first case, we stated that the regulation did not "prohibit the communication of advice between a franchisor and the holders of Class A liquor licenses," 418 F.3d at 47, but only forbade implementation of Wine Spirits' business model. We concluded that "[t]he provision of advertising and licensing services is not speech that proposes a commercial transaction and therefore does not constitute commercial speech." Id. at 49. In the later case, we observed that the prohibition on coordinated or common advertisements "does not target speech; each individual liquor licensee remains at liberty to disseminate information about its prices and products to other retail stores and to the public at large." 481 F.3d at 6. We observed: "The statute at issue here merely proscribes conduct — the launching of advertisements resulting from pre-agreed commercial strategies. Such a ban is not a ban on commercial speech." Id.

Thus, the Wine Spirits prohibition was against an acting-in-concert business approach — not against the message the liquor stores were seeking to disseminate. To be sure, the statute had an incidental impact on the speech of both the franchisor and franchisees. Wine Spirits was, in effect, prevented from marketing its services to particular businesses, and the franchisees could not distribute advertisements in coordination with other retail liquor stores. But the statute's objective was to regulate business methods, see supra n. 35, and, as we observed in Wine Spirits I, "the First Amendment does not safeguard against changes in commercial regulation that render previously profitable information valueless." 418 F.3d at 48.

Here, however, the Legislature did not simply prohibit a business model or strategy. Instead, it restricted the substance of the messages being communicated by pharmaceutical detailers in their sales pitches by curtailing information previously available to detailers. In other words, the State targeted, albeit indirectly, the speech of the detailers in order to achieve its multiple objectives. Such a regulation is a limitation on commercial speech, and the State consequently must bear the burden of demonstrating that it satisfies the Central Hudson test. See, e.g., 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484, 499, 116 S.Ct. 1495, 134 L.Ed.2d 711 (1996) (noting that "the State retains less regulatory authority when its commercial speech restrictions strike at 'the substance of the information communicated' rather than the 'commercial aspect of [it]'") (quoting Linmark Assocs., Inc. v. Willingboro, 431 U.S. 85, 96, 97 S.Ct. 1614, 52 L.Ed.2d 155 (1977)); cf. City of Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 429, 113 S.Ct. 1505, 123 L.Ed.2d 99 (1993) (noting the Court's prior "statements that the test for whether a regulation is content based turns on the 'justification' for the regulation") (citing Ward v. Rock Against Racism, 491 U.S. 781, 791, 109 S.Ct. 2746, 105 L.Ed.2d 661 (1989); Clark v. Cmty. for Creative Non-Violence, 468 U.S. 288, 293, 104 S.Ct. 3065, 82 L.Ed.2d 221 (1984)).

IV.

Before delving into the Central Hudson test and its application here, I pause briefly to clarify what this case is not about. We are not considering the State's authority to restrain untruthful, unlawful or otherwise misleading speech. Such communications — e.g., insider information about securities, fraudulent statements, or speech that would violate intellectual property laws — are routinely regulated without First Amendment inquiry. Although the State is concerned about the potentially misleading effect of the information provided by detailers to prescribers, it does not characterize the messages it seeks to restrict as categorically untruthful or deceptive. Thus, my analysis presumes that New Hampshire's prohibition on the use of prescriber-identifiable data affects communications that are truthful and otherwise lawful. As such, they may be limited only with adequate justification.

To justify a commercial speech restriction, the State bears the burden of proving the three elements of the Central Hudson test: (1) the restriction is in support of a substantial government interest; (2) it directly advances the asserted interest; and (3) it is "not more extensive than is necessary to serve that interest." Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343; El Dia, 413 F.3d at 113; see also Thompson, 535 U.S. at 367, 122 S.Ct. 1497. I consider each prong in turn.

A. Substantial Government Interest

The Attorney General maintains that the Prescription Act supports the State's substantial interests in protecting patient and prescriber privacy, promoting public health, and containing health care costs. Although the plaintiffs do not challenge the importance of the public health and cost-containment interests, they contend that the evidence in the record fails to prove that either interest is directly advanced by the statute as required by the second prong of Central Hudson. They wholly reject the Attorney General's contention that the Act serves a privacy interest.

I, too, accept as substantial the State's asserted interests in cost-containment and quality health care. However, I join the district court in rejecting on this record prescriber privacy as a sufficient interest to justify the Prescription Act. The State does not claim an interest in preventing public disclosure of the prescriber-identifiable data, and indeed it could not, as the statute allows the data to be disclosed and used for a myriad of purposes. See Defendant's Trial Memorandum, at 20 n. 10 (conceding that the law does not "attempt to keep prescriber-identifiable data secret or entirely private").

Rather, the Attorney General explains in her brief that the State's privacy interest is in the "patient-physician relationship," specifically in New Hampshire patients' "reasonable right to expect that their relationship with the physician is private, and [that] a pharmaceutical detailer is not manipulating the physician's prescribing behavior." The Attorney General contends that detailers have become "an invisible intruder in the physician's examination room."

However, the regulation does not in any cognizable way touch on the privacy of the examination room. Although the statute bars disclosure of patient-identifiable information as well as prescriber data, the plaintiffs do not challenge the prohibition on the use of specific patient data. Thus, no patient identifying information is at issue in this case. Any privacy justification must therefore reside in the prescriber-identifiable data. Rather than arguing that "the [prescriber-identifiable] data is being exploited to compromise patient privacy," the Attorney General argues that "pharmaceutical companies are using the data to help persuade doctors to make inadvisable prescribing decisions." 490 F.Supp.2d at 179. The district court properly recognized the flaw in this depiction of a privacy interest:

[W]hat the Attorney General claims as a distinct interest in protecting prescriber privacy is nothing more than a restatement of her contentions that the law can be justified because it prevents pharmaceutical companies from using prescriber-identifiable data in ways that undermine public health and increase health care costs.

Id. Accordingly, I join the district court in rejecting the Attorney General's argument that the Prescription Act is justified by a substantial privacy interest.

I thus turn to consider whether the Prescription Act is a narrowly tailored provision that directly advances the State's substantial interests in quality health care and cost-containment.

B. Advancing the Interest

The Attorney General asserts that the Prescription Act satisfies the second prong of the Central Hudson test — that it advances the State's interest — because it reduces the likelihood that prescribers will make unnecessarily expensive and unwise drug choices. I borrow the district court's well stated description of the Attorney General's logic:

The chain of reasoning . . . begins with the major premise that prescriber-identifiable data allows pharmaceutical companies to target health care providers for marketing and tailor marketing messages in ways that make detailing more persuasive. Next, it assumes that because prescriber-identifiable data makes detailing more persuasive, it inevitably leads to more prescriptions for brand-name drugs when compared with generic alternatives because only branded drugs are detailed. Finally, it assumes that any increase in the number of prescriptions written for brand-name drugs when compared to generic alternatives harms the public health and increases health care costs because branded drugs often turn out to be more harmful than generic alternatives and almost always are more expensive. Accordingly, a ban on the use of prescriber-identifiable data for marketing purposes promotes public health and contains health care costs by prohibiting pharmaceutical companies from using prescriber-identifiable data to promote the sale of brand-name drugs.

490 F.Supp.2d at 180.

The district court accepted the premise that detailing with prescriber-identifiable data is more persuasive, but found that the Attorney General had failed to establish a link between such detailing and any negative impact on public health or drug costs. On the health concern, the court found that it is "counterintuitive and unproven" that, on balance, "brand-name drugs are more injurious to the public health than generic alternatives." Id. In addition, the court was unpersuaded that the State's public health purpose was served by barring the use of prescriber data to target "early adopters" of new drugs because "the record does not establish either that early adopters are more likely to be influenced by detailing than other health care providers or that new drugs are generally more injurious to the public health than existing medications." Id.

The court found the Attorney General's position on cost-containment similarly deficient. It stated that "[n]on-bioequivalent generic drugs are not always as effective as brand-name alternatives," id., and found that the Attorney General had not proven that any reductions in health care costs stemming from reduced use of newer, more expensive medications "can be achieved without compromising patient care." Id. at 181. It thus found that none of the State's asserted interests was advanced by the Prescription Act. Moreover, to the extent that the Attorney General successfully drew a connection between truthful, non-misleading detailing based on prescriber-identifiable data and "inadvisable prescribing decisions," the district court opined that more speech, not less, was the remedy required by the First Amendment. Id.

I consider the State's showing on each of the two interests in turn.

1. Interest in the Quality of Health Care

To validate the Prescription Act on the basis of its impact on the quality of health care, the Attorney General needed to show that detailing with prescriber-identifiable data influences medical professionals to choose drugs that are less safe or less appropriate to meet patients' needs than the non-patented alternatives they would otherwise prescribe. I agree with the district court that no evidence in the record supports the proposition that newer, brand-name drugs are generally less safe or effective than older, generic ones.

The record does contain evidence that, at times, physicians are persuaded to prescribe new drugs that are less effective for patients. Dr. Avorn testified that, in the wake of extensive marketing for new hypertension medications, known as calcium-channel blockers, many doctors switched from "better, older, less-marketed products" to new products that gave patients "less benefits in terms of preventing strokes or heart disease." The record did not, however, support a conclusion that such occurrences were the norm; rather, the Attorney General's evidence primarily was directed toward showing that detailing routinely persuades health care professionals to prescribe patented medications when they offer no benefit over cheaper generic alternatives. In other words, the Attorney General's focus was on the unnecessarily high prices paid for functionally equivalent drugs. That circumstance is pertinent to the cost-containment interest I discuss in the next section, rather than to an interest in safe and appropriate health care.

Other evidence relevant to the interest in quality health care showed that detailers use prescriber-identifiable data to target early adopters, who then prescribe promoted new drugs that sometimes turn out to have harmful side effects. However, the Attorney General's argument is not that a greater number of physicians become early adopters because of targeted detailing; it claims the pharmaceutical companies use the data to identify physicians who already are inclined to adopt new drugs. In other words, the targeted doctors would likely have been among the first users of new drugs in any event. Thus, the possible adverse effect on health care stemming from reliance on the prohibited data would arise only from the possible difference in time between an early adopter's alert from a detailer and the physician's notice from another source. The record provides no basis for concluding that, in the ordinary case, that difference in time would have a significant health effect.

However, the evidence did indicate that access to early adopters was economically advantageous for the pharmaceutical companies. By soliciting the earliest possible use of new medications, the companies can maximize the financial advantage of their exclusive rights while their high-priced drugs are patent-protected. See, e.g., Day 3, PM Session, at 52 (Testimony of Dr. Avorn) ("[The drug companies] are very conscious that the patent life is ticking away, and there's a tremendous impetus on the part of the industry to be able to maximize their income as much as possible the minute the drug is released on the market."). While a few weeks or months delay in adoption of a new drug might make a substantial financial difference, the Attorney General has not shown that it would have material health consequences.

It is unsurprising that I find the Attorney General's showing on the State's health care interest to be inadequate — or at least undeveloped — given that justification's limited role in both the legislative process and the trial. Promoting quality health care was not one of the two purposes of the law identified by the Act's sponsor when she introduced the legislation, and the district court noted that the legislative history contained no "substantial support for the view that it was promoted as a public health measure, except to the extent that containing healthcare costs itself has a positive public health benefit." Tr. of Status Conference, April 11, 2006, at 44. In a colloquy with counsel toward the end of the trial, the court observed that it did not see "one shred of evidence in this record, either in the legislative history or in the trial" that prescription of higher-priced drugs instead of generics "produces unhealthy or less healthy outcomes for anybody in New Hampshire." Additionally, the plaintiffs effectively countered the Attorney General's limited showing on adverse health effects with evidence that targeted detailing is just as likely to offer health benefits; it allows drug companies to quickly alert prescribers when new drug side-effects are discovered and provides early notification to specialists of helpful new treatments for their patients. Thus, I agree with the district court that the record fails to show that the Prescription Act directly advances the State's interest in safer or better medical care. See 44 Liquormart, 517 U.S. at 505, 116 S.Ct. 1495 ("[A] commercial speech regulation 'may not be sustained if it provides only ineffective or remote support for the government's purpose.'") (quoting Central Hudson, 447 U.S. at 564, 100 S.Ct. 2343).

2. Interest in Containing Prescription Drug Costs

To justify the statute as a cost-control measure, the Attorney General has the burden of demonstrating that prescriber-identifiable data plays a significant role in the decisions of health care professionals to choose more expensive brand-name drugs over comparably effective, but less expensive, generic alternatives. See 44 Liquormart, 517 U.S. at 505, 116 S.Ct. 1495 ("[T]he State bears the burden of showing not merely that its regulation will advance its interest, but also that it will do so 'to a material degree.'") (quoting Edenfield v. Fane, 507 U.S. 761, 771, 113 S.Ct. 1792, 123 L.Ed.2d 543 (1993)). In other words, the Attorney General must show that (1) detailing generally has a persuasive effect on physicians and that (2) the use of prescriber-identifiable data magnifies that persuasive effect, increasing the physicians' tendency to prescribe unnecessary brand-name drugs.

a. The evidence

The impact of detailing on prescriber drug choice was amply documented by both empirical and anecdotal evidence. The following is a sampling of the evidence submitted to the legislature or at trial:

• Dr. Savage, president-elect of the New Hampshire Medical Society, testified at the Senate committee hearing that "[n]umerous studies" have shown that doctors' prescribing decisions "can be and sometimes [are] shaped by marketing efforts."

• During the trial, Savage's predecessor as president of the medical association, Dr. Sadowsky, related a particular instance when one of his patients, at the suggestion of her primary care doctor, asked for a brand-name drug that Sadowsky considered no better than a less expensive generic. See supra Section II.B.1. He attributed the request to detailing of the primary care physician. Sadowsky also testified:

I believe that detailing has had an [e]ffect on my prescribing. I think that just looking back I think that when medicines have gone off patent, I don't think that I thought about this consciously, but I think that my rate of prescriptions of those medicines declined in preference to the medicines I was being detailed about.

• The declaration submitted during the trial by Drs. Avorn and Kesselheim reported from their research and others' work that "[p]hysicians use of targeted prescriptions increases substantially after visits with sales representatives," Declaration, at 6, and the same result was reported in an article reviewing academic research on the effect and role of detailing. The article concluded that, "not only is detailing an important source of information, it affects physician prescription behavior in a positive and significant manner." Manchanda Honka, supra, at 787. The article cites multiple studies in which doctors acknowledged that detailing affected their prescribing behavior and reported one study showing that family physicians who relied least on sales representatives were most likely to prescribe generic drugs, "while only 12% of those who said they relied 'a great deal' on detailers prescribed generic drugs." Id. at 799.

• In her article reviewing 29 surveys exploring the relationship between physicians and pharmaceutical sales representatives, Ashley Wazana reported that "[t]here was an independent association between meetings with pharmaceutical representatives and formulary addition requests for the drug of the representative's company." See Wazana, supra, at 375. Most of the requested drugs, however, "presented little or no therapeutic advantage over existing formulary drugs." Id.

• A CALPIRG "white paper" contained in the Legislative History cited the finding of a Pennsylvania study that 40% of patients in a state assistance program received hypertension drugs different from those recommended by medical guidelines. According to the paper, the study reported that,

[i]f doctors had prescribed according to those guidelines, the state could have saved $11.6 million, or nearly 24% of the total money it spent on hypertension medicine. The study suggested that pharmaceutical promotion was partly at fault for the variance between the medicines that were recommended versus those that were prescribed.

Emily Clayton, CALPIRG, ' Tis Always the Season for Giving: A white paper on the practice and problems of pharmaceutical detailing (2004), at 4-5.

The Legislature was thus on solid ground in concluding that pharmaceutical detailing influences prescriber drug choices. The added benefit to pharmaceutical companies of marketing with access to prescriber-identifiable data, although less exhaustively covered, also was the subject of considerable testimony by the Attorney General's witnesses. Their testimony depicted targeted detailing as more aggressive and persuasive, and thus more potent than regular detailing in guiding prescriber behavior toward the detailer's desired outcome — the decision to use the sales representative's patented, brand-name drug. On the specific impact of detailing with prescriber-identifiable information, the evidence included the following:

• Dr. Gary Sobelson, a family practice physician, testified at trial that he was unaware of scientific evidence showing that the sale of prescriber-specific data increases drug costs, but observed that such knowledge "puts me at a disadvantage that I'm not comfortable being at." He told of being persuaded to prescribe a brand-name drug, Zithromax, instead of an equivalent generic Amoxicillin, based on an incorrect assumption that Zithromax, which had the advantage of requiring a shorter course of therapy, was minimally more expensive than the older Amoxicillin. After discovering that Zithromax was five times more expensive, he moved away from Zithromax because "I'm interested in prescribing rationally for my patients in a way that both maximizes their outcome but also helps maintain the lowest possible cost to both them individually and, frankly, to our society at large."

• Sobelson also described how detailers use prescriber-identifiable information when marketing to a physician who typically prescribes a competitor's equivalent product, citing two cholesterol-lowering medications, Lipitor and Zocor, in his example:

[W]hen a drug representative for Lipitor comes to see me, . . . they are going to know to present data that would focus me to why I should prefer Lipitor over Zocor. It's a very, very specific focus that particularly is fueled if they happen to know that 80 percent of my prescribing is Zocor. And so when the Lipitor rep comes around, they are going to have their targeted information provided by their marketing department. This is how we've learned from our study groups that you get doctors to move from Zocor to Lipitor.

• Sobelson's experience on the receiving end of the marketing dovetailed with the description provided by a former detailer of his strategy when he had prescriber information. Shahram Ahari testified that, when he knew a physician's patterns, "I have a fair idea why, and so it becomes almost a cat and mouse game when I get them to say their objections and for me to shift those objections or doubts and down-play or negate them altogether." By contrast, without prescriber-specific information,

it becomes less about the business and more about knowing the science of my drug. . . . [I]t puts the power of the detail more in the physician's hands because I don't truly know what his concerns are or what his perspectives or biases are. . . . [I]t shifts the power of the conversation to a more equal footing.

• A Boston Globe article included in the Legislative History reported similar information; a sales representative told of his understanding that, if he learned that a doctor was prescribing a competitor's product, his presentation should focus on undermining that product. Liz Kowalczyk, Drug Companies' Secret Reports Outrage Doctors, Boston Globe, May 25, 2003, at A1.

• Plaintiff IMS has explained the benefits of data-mining with a focus on prescriber-specific data: "By using a data-mining solution, IMS can pinpoint prescribers who are switching from one medication to another. A sales person can use this model to target doctors who have switched from the drug they are selling and to devise a specific message to counter that switching behavior." Paul Kallukaran Jerry Kagan, Data Mining at IMS HEALTH: How We Turned a Mountain of Data into a Few Information-rich Molehills, IMS Abstract.

• In both his testimony and declaration, Dr. Avorn stated that detailing becomes less information-focused and a more powerful tool of persuasion when the sales representative is armed with prescriber-specific information. In his joint declaration with Dr. Kesselheim, he related the "counter-detailing" experience of his research unit at Harvard Medical School, in which he and his colleagues used prescriber-specific data obtained from pharmacy records to choose physicians for educational visits by clinical pharmacists, accompanied by mailed "unadvertisements." He reported that these targeted interventions resulted in a 14 percent reduction in inappropriate prescriptions, Declaration at 9, and he saw significance in these results for commercial detailing:

Our educational programs (known as "academic detailing") focused on improving patient care through reducing excessive use of inappropriate medications. But when these techniques are used by companies whose main goal is simply to increase product sales, the impact on patients and on the health care system are quite different. The studies we have cited indicate that more physician-specific detailing will lead to more prescriptions of brand-name agents, often with no additional patient benefit but at much higher cost to patients and to state-based insurance programs, which will continue to drive up the cost of health care in New Hampshire.

Id. at 10.

Avorn echoed these observations at trial, explaining that prescriber-identified data was important to the success of his counter-detailing because "that's how we knew whom to visit, and we also knew what to say to them because we knew what drugs they were prescribing." In the declaration, he stated that restricting access to prescriber-specific information, "[m]aking it more difficult for manufacturers to tailor their marketing strategies to . . . individual physicians[,] would actually encourage detailers to present physicians with a more neutral description of the product that would emphasize presentation of information over promotion." Declaration, at 11; see also Day 3, PM Session, at 140 (Avorn Testimony) ("[I]f the sales rep knows my prescribing history, they will market to me or at me in a way that goes well beyond just providing me with the data. It's not really education at that point. It's not a level playing field.").

• An assumption that prescriber-identifiable detailing impacts drug choice is reflected in the professional guidelines cautioning against using the data aggressively. As noted above, the AMA has adopted suggestive guidelines against the use of "prescribing data to overtly pressure or coerce physicians to prescribe a particular drug." Such indirect evidence supports the State's view that eliminating access to the information will decrease the likelihood that physicians will be swayed by targeted marketing to prescribe unnecessary — and more expensive — brand-name drugs.

b. The district court's evaluation of the evidence

The district court concluded that, notwithstanding this evidence, the State's showing was insufficient to establish a link between the Prescription Act and cost-containment because other evidence showed that more expensive brand-name drugs will, at times, be the better therapeutic choice. The court acknowledged that "substantial deference" must be given to a legislature's predictive judgments "[w]hen a quality record establishes that the legislature conducted an extensive investigation, acquired considerable expertise in the regulated area, and incorporated express findings into the approved statute." 490 F.Supp.2d at 177 n. 12 (citing Turner Broad. Sys., Inc. v. FCC, 520 U.S. 180, 186, 117 S.Ct. 1174, 137 L.Ed.2d 369 (1997)). However, the court questioned the extent of the Legislature's investigation before adopting the initiative, noting, inter alia, that it acted quickly after the bill was introduced, made no express findings on the need for the legislation, and "cited no evidence as to how effective the restriction might prove to be." 490 F.Supp.2d at 177.

I am mindful that regulations that suppress commercial speech must be carefully evaluated. Nonetheless, the district court held the Attorney General to a higher standard of proof than is required by Supreme Court precedent. While a state legislature "does not have the broad discretion to suppress truthful, nonmisleading information for paternalistic purposes," 44 Liquormart, 517 U.S. at 510, 116 S.Ct. 1495, the Court's commercial speech cases "recognize some room for the exercise of legislative judgment." Id. at 508, 116 S.Ct. 1495. To earn that deference, the State must offer probative evidence that suppressing speech is essential to achieving its goal. However, a state legislature cannot reasonably be expected to undertake an investigation of the scope conducted by Congress in connection with the federal legislation at issue in Turner Broadcasting, the case cited by the district court, to justify a limited restriction on commercial speech. See Turner Broad. Sys., 520 U.S. at 187, 117 S.Ct. 1174 (noting that the record included "tens of thousands of pages" of materials acquired during three years of Congressional preenactment hearings, as well as additional expert submissions, sworn declarations, testimony, and industry documents).

In Turner Broadcasting, the Court observed that, given the exhaustive record, Congress's findings were entitled to "deference in part because the institution is far better equipped than the judiciary to amass and evaluate the vast amounts of data bearing upon legislative questions." 520 U.S. at 195, 117 S.Ct. 1174 (internal quotation marks and citations deleted). Although the contexts are different, the general principle of legislative deference also is compatible with the Court's commercial speech precedent. The question here, as there, is whether the government is able to support its restriction on speech by "'adduc[ing] either empirical support or at least sound reasoning on behalf of its measure[].'" Turner Broad. Sys., 512 U.S. at 666, 114 S.Ct. 2445 (quoting Century Commuc'ns Corp. v. FCC, 835 F.2d 292, 304 (D.C. Cir. 1987)); see Florida Bar v. Went For It, Inc., 515 U.S. 618, 628, 115 S.Ct. 2371, 132 L.Ed.2d 541 (1995) ("[W]e do not read our case law to require that empirical data come to us accompanied by a surfeit of background information. Indeed, in other First Amendment contexts, we have permitted litigants to justify speech restrictions by reference to studies and anecdotes pertaining to different locales altogether, or even, in a case applying strict scrutiny, to justify restrictions based solely on history, consensus, and 'simple common sense.'") (citations omitted). If the government makes the requisite showing, we defer to the legislative judgment to adopt the challenged measure.

The Attorney General has no empirical data showing the extent of the influence of prescriber-specific information on physicians' decision-making; nor can she document how much money the Prescription Act will save the State or consumers. The regulation was the first of its kind in the country, and it had been in effect for less than a year when the district court invalidated it. It is unreasonable in these circumstances to expect the Attorney General to provide extensive quantifiable data that might only become available after the statute has been in place for some time. I have described evidence here that establishes a plausible cause-and-effect relationship between targeted detailing and higher drug prices. What is missing is hard evidence of the global extent of this relationship. Clearly, it will be important going forward for the State to try to measure the cost-containment effect of its initiative, and it is possible that this ongoing assessment will indicate that the measure is not as effective as the State had hoped.

However, at this juncture, the Attorney General has established a factual basis justifying the initiative. She has adduced significant testimony based on relevant empirical research concerning the impact of detailing generally, supplemented by the personal experience of both prescribers and detailers, strongly indicating that sales pitches based on specific prescribing patterns have a particularly persuasive impact on drug choice. The extent of this empirical and anecdotal evidence, particularly in light of the Act's limited restriction on speech, distinguishes this case from those in which the Supreme Court has found more sweeping bans on commercial speech to be inadequately justified. For example, the Court in Edenfield noted the absence of any studies or anecdotal evidence to support a ban on in-person solicitation by accountants. 507 U.S. at 771, 113 S.Ct. 1792. In Shapero v. Kentucky Bar Ass'n, 486 U.S. 466, 108 S.Ct. 1916, 100 L.Ed.2d 475 (1988), which rejected a ban on direct-mail solicitations by lawyers, the State "assembled no evidence attempting to demonstrate any actual harm caused by targeted direct mail," Florida Bar, 515 U.S. at 629, 115 S.Ct. 2371. See also U.S. West, Inc., 182 F.3d at 1237 (noting that the government had presented "no evidence" showing that the harm to either of its two asserted interests "is real").

Moreover, as I have recounted, evidence from multiple sources indicated that the expense of unnecessary brand-name prescribing has in the past ranged into the billions of dollars nationally. This substantial evidence of needless spending, combined with evidence that detailing with prescriber-identifiable data contributes to that outcome, is enough to show that the Prescription Act "targets a concrete, non-speculative harm," Florida Bar, 515 U.S. at 629, 115 S.Ct. 2371, and that the Attorney General has sufficiently demonstrated that the State's interest in cost-containment would be furthered "to a material degree" by the limitation on speech it seeks to achieve through the Prescription Act. See, e.g., City of Los Angeles v. Alameda Books, Inc., 535 U.S. 425, 426, 122 S.Ct. 1728, 152 L.Ed.2d 670 (2002) ("[A] municipality may rely on any evidence that is 'reasonably believed to be relevant' for demonstrating a connection between speech and a substantial, independent government interest."); cf. Turner Broad. Sys., 512 U.S. at 666, 114 S.Ct. 2445 ("[T]he obligation to exercise independent judgment when First Amendment rights are implicated is not a license to reweigh the evidence de novo, or to replace Congress' factual predictions with our own. Rather, it is to assure that, in formulating its judgments, Congress has drawn reasonable inferences based on substantial evidence.").

Importantly, the district court made no finding that the Attorney General had failed to establish a relationship between detailers' use of prescriber-identifiable data and increased health costs. Instead, the court concluded that the Attorney General had failed to show that the Act advanced the State's interest because any cost savings might be offset by compromised health care for patients who would in fact benefit from the use of more expensive brand-name drugs.

It does not matter that detailing with prescriber-identifiable data sometimes has positive effects. The Attorney General's evidence indicated that the health care benefits of such marketing described by plaintiffs are largely achievable in other ways. News reports, for example, would highlight truly groundbreaking new therapies in a timely way and, indeed, pharmaceutical detailers with knowledge of physicians' medical specialties presumably would not need access to prescribing histories to effectively promote such innovations. Early adopters could be expected to respond quickly with an interest in trying the new medications — effectively identifying themselves to the sales representatives. In addition, as I already have observed, the statute does not bar drug companies from alerting prescribers to newly discovered problems with their medications. In other words, I see no message or interest of consequence that is foreclosed by the regulation. Cf. Thompson, 535 U.S. at 376, 122 S.Ct. 1497 (noting that "the amount of beneficial speech prohibited by the [statute]" would be "enough to convince us that the . . . advertising provisions were unconstitutional"); Greater New Orleans Broad. Ass'n, Inc. v. United States, 527 U.S. 173, 194, 119 S.Ct. 1923, 144 L.Ed.2d 161 (1999) (noting that the statute at issue "sacrifices an intolerable amount of truthful speech about lawful conduct when compared to all of the policies at stake"). Thus, the fact that detailing with prescriber-identifiable data may at times have a positive effect on health care does not negate the Act's role in advancing the State's interest in cost-containment.

C. Narrow Tailoring

In evaluating the narrow tailoring prong of the Central Hudson inquiry, the Court typically has asked "whether the extent of the restriction on protected speech is in reasonable proportion to the interest served." Edenfield, 507 U.S. at 767, 113 S.Ct. 1792; see also Greater New Orleans Broad. Ass'n, 527 U.S. at 188, 119 S.Ct. 1923 ("The Government is not required to employ the least restrictive means conceivable, but it must demonstrate narrow tailoring of the challenged regulation to the asserted interest — 'a fit that is not necessarily perfect, but reasonable'. . . .") (quoting Bd. of Trustees of State Univ. of N.Y. v. Fox, 492 U.S. 469, 480, 109 S.Ct. 3028, 106 L.Ed.2d 388 (1989)); Florida Bar, 515 U.S. at 632, 115 S.Ct. 2371 ("[T]he 'least restrictive means' test has no role in the commercial speech context.").

This "reasonable fit" standard of intermediate scrutiny has drawn criticism. See Thompson, 535 U.S. at 367-68, 122 S.Ct. 1497 (noting that "several Members of the Court have expressed doubts about the Central Hudson analysis and whether it should apply in particular cases"); Lorillard Tobacco Co. v. Reilly, 533 U.S. 525, 554-55, 121 S.Ct. 2404, 150 L.Ed.2d 532 (2001) (same); Greater New Orleans Broad. Ass'n, 527 U.S. at 184, 119 S.Ct. 1923 (recognizing the advocacy among judges, scholars and others for "a more straightforward and stringent test for assessing the validity of governmental restrictions on commercial speech"). However, the Court majority has adhered to the Central Hudson approach, observing repeatedly that, in the particular case at issue, "there is no need to break new ground" in assessing the validity of the challenged governmental restrictions on commercial speech. See Thompson, 535 U.S. at 368, 122 S.Ct. 1497; Lorillard Tobacco Co., 533 U.S. at 554-55, 121 S.Ct. 2404; Greater New Orleans Broad. Ass'n, 527 U.S. at 184, 119 S.Ct. 1923.

Nonetheless, the debate on Central Hudson's continuing viability seems to have influenced the Court's application of its framework. Multiple commentators have observed that intermediate scrutiny under Central Hudson has "come to resemble closely the 'narrowly tailored' requirement of strict scrutiny." Troy L. Booher, Scrutinizing Commercial Speech, 15 Geo. Mason U. Civ. Rts. L.J. 69, 77 (2004); see also R. Michael Hoefges, Regulating Professional Services Advertising: Current Constitutional Parameters and Issues Under the First Amendment Commercial Speech Doctrine, 24 Cardozo Arts Ent. L.J. 953, 989 (2007) (noting that recent precedent arguably "has pushed the fourth prong of the Central Hudson analysis closer than ever before to the least-restrictive-means requirement of strict constitutional scrutiny"); Emily Erickson, Disfavored Advertising: Telemarketing, Junk Faxes and the Commercial Speech Doctrine, 11 Comm. L. Pol'y 589, 602 (2006) ("[T]he broader trend has been one of higher scrutiny for commercial speech cases."); Elizabeth Spring, Sales Versus Safety: The Loss of Balance in the Commercial Speech Standard in Thompson v. Western States Medical Center, 37 U.C. Davis L.Rev. 1389, 1404 (2004) ("[T]he Court is now applying the Central Hudson test in a manner approaching strict scrutiny review.").

Indeed, in Thompson, a 5 to 4 decision, Justice Breyer in dissent chastises the majority for applying the commercial speech doctrine "too strictly" in striking down a statute prohibiting the advertising of compounded drugs. 535 U.S. at 388, 122 S.Ct. 1497. In finding that the regulation was not narrowly tailored, the majority proposed a variety of non-speech alternatives that the Government could have adopted to meet its objectives. The justices observed that "[i]f the Government could achieve its interests in a manner that does not restrict speech, or that restricts less speech, the Government must do so." Id. at 371, 122 S.Ct. 1497. From Justice Breyer's perspective, however, the majority "too readily assume[d] the existence of practical alternatives." Id. at 388, 122 S.Ct. 1497.

This case does not require us to decide if Thompson represents a departure in the Court's application of the narrow tailoring prong of Central Hudson. As I shall explain, even as applied by the majority in Thompson, Central Hudson's narrow tailoring requirement is satisfied here. As an initial matter, the restriction on speech imposed by the Prescription Act is significantly more limited than similar restrictions on commercial speech that have been considered by the Supreme Court. It is neither a complete ban on the marketing or advertising of a product or its price, see, e.g., Thompson, 535 U.S. at 360, 122 S.Ct. 1497 (compounded drugs); 44 Liquormart, 517 U.S. at 489, 116 S.Ct. 1495 (retail price of alcoholic beverages), nor a blanket prohibition on in-person solicitation, see, e.g., Edenfield, 507 U.S. at 763, 113 S.Ct. 1792 (accountants); Ohralik, 436 U.S. at 448-49, 98 S.Ct. 1912 (attorneys). Pharmaceutical sales representatives may continue to pitch their drugs directly to doctors and other health care providers, and the only message proscribed is one that incorporates an awareness of the doctor's prescribing practices. The detailers also may continue to use prescriber data provided by the plaintiffs for marketing, so long as the data aggregates prescribing patterns by speciality and zip code and not by individual provider. Thus, this case does not trigger the "special concerns [that] arise from 'regulations that entirely suppress commercial speech in order to pursue a non-speech-related policy,'" 44 Liquormart, 517 U.S. at 500, 116 S.Ct. 1495 (quoting Central Hudson, 447 U.S. at 566 n. 9, 100 S.Ct. 2343).

Despite the Act's limited scope, the plaintiffs maintain that it is broader than necessary to serve the State's objective and that it thus fails the narrow tailoring test. For multiple reasons, I reject the plaintiffs' contention and conclude that the State has met its burden of justifying the Prescription Act. The inadequacy of alternatives to satisfy the State's interests, the context of private communications, and the limited impact on the message sought to be disseminated lead me to conclude that New Hampshire has established "a 'reasonable fit' between its abridgment of speech and its . . . goal," 44 Liquormart, 517 U.S. at 507, 116 S.Ct. 1495.

1. Inadequacy of Alternative Measures

The plaintiffs argue that the State's cost-containment objective could have been achieved through measures that did not impact protected speech at all. The district court agreed and noted that, for example, the Legislature could have addressed the issue by "properly implementing" a Medicaid Pharmacy Program that takes into account the cost-effectiveness of brand-name drugs. 490 F.Supp.2d at 182. The court pointed out that New Hampshire's current program requires authorization for Medicaid patients to obtain certain drugs and that state regulations allow cost considerations to be taken into account when deciding which drugs should be subject to the authorization. 490 F.Supp.2d at 182. As a result, the court concluded that the State could prevent unnecessary expenditures on brand-name drugs by denying authorization requests for more expensive drugs that are no more effective than cheaper alternatives. Id.

This proposal and the other non-speech alternatives proposed by the parties and the district court lack equivalency with the Prescription Act in accomplishing the State's cost-containment goal. In response to the district court's suggestion that legislative changes be made in the Medicaid program, the Attorney General argues that such measures would not respond to the State's broader concern that physicians' drug choices for all patients are distorted by the detailers' access to prescriber-identifiable data. In addition, the Attorney General maintains that formularies also are affected by pharmaceutical detailing, citing evidence that physicians request additions to such lists even when the added drugs have "little or no therapeutic advantage over existing formulary drugs." Wazana, supra, at 375.

The court's other suggestions — requiring the State "to enter the intellectual marketplace" with its own information about proper drug choices; mandating participation in continuing medical education programs; or limiting the samples, meals and other ingratiating gifts provided by detailers to prescribers — are similarly imperfect. The Attorney General argues that the State lacks comparable resources to directly counter commercial detailing — for which the pharmaceutical companies spend billions of dollars — and the district court at trial noted Avorn's testimony that relying on medical education programs would be difficult because "it would be hard to find the right people and . . . [t]here would be disputes over what the content is."

I acknowledge that the suggestion that the State prohibit courtesy samples and other gifts to prescribers is not as easily dismissed. That prohibition could be implemented unilaterally and without expense to the State. Like the Prescription Act, such a ban would be directly aimed at diminishing the persuasive force of the detailers' message. As described above, the record contains evidence that the perks have a subtle influence on physicians' decision-making, increasing their affinity for particular sales representatives — and, presumably, for those representatives' drugs. In fact, a number of states have passed laws requiring that gifts to prescribers be publicly disclosed, and, as with the use of prescriber-identifiable data, professional guidelines have been adopted to reduce or eliminate such benefits.

While similar in intent, however, a ban on gifts and the ban on the use of prescriber-identifiable data are not interchangeable means of achieving the State's goal of cost-containment. The samples and gifts are merely a preparatory step in the marketing process; while they may increase the prescribers' susceptibility to the sales pitch, the State reasonably concluded that it is the sales pitch itself that has the most troubling effect on the prescribers' drug choice-and is most urgently in need of regulation. See Appellant's Brief at 42 (asserting that pharmaceutical companies use prescriber-identifiable data "to subtly manipulate physicians, in ways physicians are often unaware, to change their prescriptions for reasons other than the clinical needs of patients") (citing Avorn Declaration, at 9-11).

Moreover, Avorn testified that the remedies proposed by the district court "have been tried, not necessarily in New Hampshire, in particular, but nationally in terms of trying to restrict the freebies, trying to provide doctors with other means of learning, requiring that doctors take continuing ed courses." Avorn opined that the Prescription Act

was not just a flippant, oh, let's see what happens with this. It was more of a sense of people have tried everything they can try and we still have this massive distortion of what doctors are prescribing and what the State, and its citizens, are paying for drugs because of the very heavily and very effective promotional strategies that are going on out there; and this seemed like — given that those other avenues are probably not going to be viable, that this seemed to be a way of preserving the company's ability to give me their best shot in their sales argument, but not to do so with a kind of knowledge that really shouldn't have anything to do with teaching me something. . . .

I am thus satisfied that the State has eliminated the possibility that "alternative forms of regulation that would not involve any restriction on speech would be more likely to achieve the State's goal," 44 Liquormart, 517 U.S. at 507, 116 S.Ct. 1495 (emphasis added). To the contrary, Avorn's summary of other initiatives indicates that the State reasonably concluded that its legislation provided the only effective approach for achieving its objective.

In responding to the proposed alternatives through argument and evidence, the Attorney General in this case took steps that the majority in Thompson found lacking in the government's presentation there. The Court observed that "[n]o-where in the legislative history of the [Act] or petitioners' briefs is there any explanation of why the Government believed forbidding advertising was a necessary as opposed to merely convenient means of achieving its interests." 535 U.S. at 373, 122 S.Ct. 1497. The Court commented that "there is no hint" that the government had considered the alternatives proposed by the Court, or any other strategies. Id. In this case, the State offered expert evidence at trial and argued in its briefs on appeal in defense of its view that alternative strategies would not suffice. Thus, unlike in Thompson, the State has amply rebutted any impression that regulating speech was the first, or only, strategy it thought to try. Cf. id.

2. Focus on Private Communications

It is also significant that the Prescription Act restricts only private communications between the pharmaceutical detailer and prescribers, rather than a message disseminated to the public at large. In evaluating whether the Prescription Act advanced the State's cost-containment interest, the district court noted the Supreme Court's rejection in Thompson of a government interest "'in preventing the dissemination of truthful commercial information in order to prevent members of the public from making bad decisions with the information.'" 490 F.Supp.2d at 181 (quoting Thompson, 535 U.S. at 374, 122 S.Ct. 1497); see also 44 Liquormart, 517 U.S. at 503, 116 S.Ct. 1495 ("The First Amendment directs us to be especially skeptical of regulations that seek to keep people in the dark for what the government perceives to be their own good.").

This case differs from those in which the Court has rejected advertising bans that restrict the exchange of ideas in the "commercial marketplace." The Prescription Act neither "protects" the public from information about drugs nor prevents truthful advocacy by pharmaceutical representatives. Instead, it prevents sales representatives from crafting personal marketing messages on the basis of data that credible evidence indicates has been used to unduly influence prescribing choices. The Supreme Court on multiple occasions has reviewed regulation of such direct solicitations, upholding restrictions where the context raised concerns about the impact of the marketing on the recipient. See Edenfield, 507 U.S. at 765, 113 S.Ct. 1792 ("There are, no doubt, detrimental aspects to personal commercial solicitation in certain circumstances. . . .").

Two such cases provide a helpful contrast and offer guidance in this case. In Ohralik, the Court upheld a bar against in-person solicitation of prospective clients by lawyers in "'situation[s] that breed[] undue influence,'" 436 U.S. at 449, 98 S.Ct. 1912 (quoting Bates v. State Bar of Ariz., 433 U.S. 350, 366, 97 S.Ct. 2691, 53 L.Ed.2d 810 (1977)). Ohralik involved two young victims of an automobile accident, one who was approached while she was still hospitalized and the other on the day she was released from the hospital. Id. at 450-51, 98 S.Ct. 1912. The Court found that the State's compelling interest in "preventing those aspects of solicitation that involve fraud, undue influence, intimidation, overreaching, and other forms of 'vexatious conduct'" justified the limited restriction on speech. Id. at 462, 98 S.Ct. 1912. The Court further observed that "it hardly need be said that the potential for overreaching is significantly greater [than in the sale of ordinary consumer products] when a lawyer, a professional trained in the art of persuasion, personally solicits an unsophisticated, injured, or distressed lay person." Id. at 464-65, 98 S.Ct. 1912.

By contrast, the Court concluded in Edenfield that a ban on face-to-face solicitation by certified public accountants ("CPAs") did not survive First Amendment scrutiny. 507 U.S. at 765, 113 S.Ct. 1792. Although noting that face-to-face commercial solicitation may have "detrimental aspects," id., the Court also recognized that, "[i]n the commercial context, solicitation may have considerable value," id. at 766, 113 S.Ct. 1792. Among the advantages listed by the Court were "direct and spontaneous communication between buyer and seller," "enabl[ing] the seller to direct his proposals toward those consumers who he has reason to believe would be most interested in what he has to sell," and providing buyers "an opportunity to explore in detail the way in which a particular product or service compares to its alternatives in the market." Id. The Court ultimately found that the risks inherent in the Ohralik context did not exist in the accountant setting:

Unlike a lawyer, a CPA is not "a professional trained in the art of persuasion." A CPA's training emphasizes independence and objectivity, not advocacy. The typical client of a CPA is far less susceptible to manipulation than the young victim in Ohralik. Fane's prospective clients are sophisticated and experienced business executives who understand well the services that a CPA offers. In general, the prospective client has an existing professional relation with an accountant and so has an independent basis for evaluating the claims of a new CPA seeking professional work.

Id. at 775, 113 S.Ct. 1792 (citations omitted). The Court thus concluded that "the ends sought by the State are not advanced by the speech restriction," and that the rule against in-person solicitation "infringe[d] upon Fane's right to speak, as guaranteed by the Constitution." Id. at 777, 113 S.Ct. 1792.

In relevant respects, this case falls between Ohralik and Edenfield. Although the recipients of the marketing messages at issue here are, unlike in Ohralik, highly trained professionals, the solicitor in question — the pharmaceutical detailer — is schooled in the art of persuasion, like the lawyers in Ohralik. Unlike in Edenfield, there is substantial evidence that the detailer's persuasion has an impact and that confining the marketing interaction in the manner required by the Prescription Act would advance the State's interest. The detailer often has knowledge of drug details that are not readily available to the physician, and the evidence supports the State's view that adding prescriber-identifiable data into the mix lends weight to the detailer's message — and increases the likelihood that the targeted prescriber will choose the brand-name drug being promoted by the detailer.

This is not to suggest that the detailer's message is generally inaccurate or misleading. The advantage provided by prescriber-identifiable data may only be to re-focus the emphasis of the presentation. But where the record shows a real risk that "one-sided" presentations may give marketers "undue influence," the appropriateness of limiting speech veers much closer to Ohralik than Edenfield. See 44 Liquormart, 517 U.S. at 498, 116 S.Ct. 1495 (commenting that the State "may restrict some forms of aggressive sales practices that have the potential to exert 'undue influence' over consumers"); Ohralik, 436 U.S. at 462, 98 S.Ct. 1912 (noting state's legitimate interest in "preventing those aspects of solicitation that involve fraud, undue influence, intimidation, overreaching, and other forms of Vexatious conduct'") (emphasis added).

3. Calculation of Costs and Benefits

I already have described the alternative ways in which prescribers will have access to the helpful information that may no longer be available to them from pharmaceutical detailers as a result of the Prescription Act. See supra Section IV.B.2.b. The statute therefore suppresses only a small amount of beneficial speech. "On the whole, then, the challenged regulation . . . indicate[s] that [the State] '"carefully calculated" the costs and benefits associated with the burden on speech imposed by its prohibition.'" Greater New Orleans Broad. Ass'n, 527 U.S. at 188, 119 S.Ct. 1923 (quoting Discovery Network, 507 U.S. at 417, 113 S.Ct. 1505 (quoting Fox, 492 U.S. at 480, 109 S.Ct. 3028)); see also U.S. West, Inc., 182 F.3d at 1238.

In this context, I conclude that the State has met its burden to justify the limited restraint on commercial speech imposed by the Prescription Act.

V.

There remains the plaintiffs' Commerce Clause challenge to the Act. I part company with my colleagues on that challenge because the majority's discussion of the issue, and its ready acceptance of the Attorney General's statement about the scope of the Act, further undermine the value of the majority's decision. There is a puzzling disconnect between the Attorney General's contention that the Act governs only transactions that take place within New Hampshire and the plaintiffs' contention that all of the conduct that the Act purports to regulate occurs outside the State. On the record before us, we do not have an adequate foundation for evaluating that disconnect and its implications for the Commerce Clause analysis. I therefore would remand this case to the district court with instructions to address the Commerce Clause issue in the first instance.

Under the Attorney General's interpretation of the statute-that the Act reaches only transactions that occur within New Hampshire-no Commerce Clause problem would exist. See Alliance of Auto. Mfrs. v. Gwadowsky, 430 F.3d 30, 35 (1st Cir. 2005) (explaining that, in evaluating whether a statute has impermissible extraterritorial reach, courts are obliged to adopt any reasonable construction consistent with the Constitution). The majority summarily deems that narrowing construction "reasonable," commenting that "it would make no sense to read the statute to regulate out-of-state transactions when the upshot of doing so would be to annul the statute." Yet a literal application of that narrowing construction would appear to leave the Act with negligible impact — hardly a reasonable outcome.

It is undisputed that none of the plaintiffs' transactions take place within New Hampshire. The district court found that "IMS and Verispan obtain all of their prescription information, including information on prescriptions filled in New Hampshire, from computers that are located outside of New Hampshire." 490 F.Supp.2d at 166. At trial, the court described the factual record on the Commerce Clause question as follows:

It's undisputed that prescriptions are generated in the state. It's undisputed that the prescriptions are filled within the state. It's undisputed that the pharmacies where they're filled [are] based in the state. It's undisputed that the pharmacy, as a part of its routine practices, unassociated with the sale of this information to pharmaceutical companies or IMS, transfers the information in the ordinary course of its business from a data center in the state to data centers outside the state. That the IMS software and Verispan software is applied to it outside the state. That it is then transferred from the [pharmacy] to IMS or Verispan outside the state, and it is thereafter sold to pharmaceutical companies and other clients outside the state.

The parties agreed that this summary, with some variations, was accurate and also agreed with the court's understanding that "the factual record that bears on the Commerce Clause question is undisputed."

Given these undisputed facts, however, it is unclear how much, if any, of the activity that the statute explicitly proscribes occurs within New Hampshire. For example, the "routine" transfer of prescriber-identifiable information from a local New Hampshire pharmacy to the pharmacy's out-of-state headquarters does not appear to be prohibited by the Act. Arguably, that electronic transfer would not be for an impermissible "commercial purpose" — involving, inter alia, "advertising, marketing, promotion, or any activity that could be used to influence sales or market share of a pharmaceutical product, influence or evaluate the prescribing behavior of an individual health care professional, or evaluate the effectiveness of a professional pharmaceutical detailing sales force." Consequently, the data would be outside New Hampshire before any transaction described by the Act occurs. The district court's factual summary suggests that most prescriber-identifiable data leaves New Hampshire in this permissible manner.

That understanding of the facts underlies the plaintiffs' argument that the Act seeks to prohibit the licensing, transfer, use, or sale of data identifying New Hampshire prescribers wherever such activity occurs. Plaintiffs' counsel explained their position during a colloquy with the court at trial:

The State has said, this doesn't apply outside of the state. . . . [O]ur reply to that has been . . . if it doesn't prohibit these transactions outside of the state, then the statute really loses all of its force and effectiveness. Because if Rite Aid's pharmacy in New Hampshire can transfer to its parent in Pennsylvania and its parent can transfer to IMS or Verispan in Pennsylvania, that's not prohibited. And then they can transfer it to Pfizer, wherever Pfizer's headquarters are outside of New Hampshire; and if Pfizer can then use it outside of New Hampshire for all of these various purposes that are prohibited, then there's absolutely no force or effect to this statute. And I think what the State is really arguing is that . . . all these transfers outside of the state, they are prohibited.

This statement stops one step short of demonstrating the most critical flaw in the Attorney General's narrowing construction of the Act. If her view of the Act were correct, not only could Pfizer buy and use New Hampshire data outside of New Hampshire "for all these various purposes that are prohibited," but the Act also would pose no barrier to the use of such data by detailers inside New Hampshire. This would be so because the Act does not apply to detailers and, as noted above, the undisputed facts suggest that the detailers routinely obtain the data from entities whose acquisition of the information, according to the Attorney General, was not restricted by the Act. Hence, the detailers' use of prescriber-identifiable data in New Hampshire doctors' offices would appear to involve no violation of the Prescription Act. In taking an indirect route toward its goal of regulating detailers' communications, presumably to avoid the First Amendment concerns that would be triggered by a direct restriction on speech, the Legislature may not have accomplished what it intended.

Of course, the Attorney General may believe that her concession that the Act does not apply to out-of-state transactions is not problematic because of her view that the Act bars detailers from using prescriber-identifiable data in their communications with New Hampshire prescribers if that data originated in New Hampshire, regardless of whether the pharmaceutical company purchased the information inside or outside of the state. Indeed, that understanding of the Act's scope is suggested by the Attorney General's comments during the parties' colloquy with the district court:

The reality of the situation here is we have . . . national chain pharmacies moving into the State of New Hampshire, setting up their own places of business, hiring pharmacists, hiring managers, establishing a place of business in the State of New Hampshire and then obviously agreeing to abide by the laws of the State of New Hampshire when they establish a place of business in this state; and then in the course of their business, they're collecting . . . these data. They're moving these data out of the state, for whatever purpose, in full knowledge of . . . the laws of the State of New Hampshire. . . .

Under this view of the law, New Hampshire places an embargo on the use of the prescriber-identifiable data before it is first released by the pharmacies. The Attorney General apparently contemplates that New Hampshire pharmacies and similar entities would be permitted to license, transfer, use or sell the information they accumulate only on the condition that the data not be used downstream for the prohibited commercial purposes.

However, the disconnect that I described earlier remains. The explicit language of the Act does not appear to impose such a restriction on the original transfers of data by New Hampshire pharmacies to entities outside the state. The Act proscribes only the transfer of prescriber-identifiable data for the specified commercial purposes. The transfer of data by New Hampshire pharmacies beyond New Hampshire's borders typically may not implicate those prohibitions. Transactions involving those commercial purposes occur farther downstream, and, so far as the record shows, primarily outside the state. Frankly, I am not sure that the Attorney General understood the import of her statement that the Act regulates only in-state transactions. Nor, given the state of the record, do I understand the majority's statement that, when the Act is interpreted as the Attorney General proposes, it "may result in a loss of profit to out-of-state data miners due to the closing of one aspect of the New Hampshire market for their wares." To the contrary, the statute's impact in New Hampshire appears negligible if it truly governs only transactions that occur within the state.

Although the Attorney General's concession was an attempt to sidestep the plaintiffs' Commerce Clause challenge, there may be an argument that such a step was unnecessary. When a state statute regulates commerce "wholly beyond the boundaries of the enacting state," it usually is invalid per se. Alliance of Auto. Mfrs., 430 F.3d at 35. Yet not every impact on interstate commerce is prohibited. "[T]he dormant Commerce Clause[] is not absolute and in the absence of conflicting legislation by Congress, 'the States retain authority under their general police powers to regulate matters of legitimate local concern, even though interstate commerce may be affected.'" Pharm. Care Mgmt. Ass'n v. Rowe, 429 F.3d 294, 311 (1st Cir. 2005) (quoting Maine v. Taylor, 477 U.S. 131, 138, 106 S.Ct. 2440, 91 L.Ed.2d 110 (1986)). Moreover, whether extraterritoriality is impermissible in every instance, or whether it transgresses the dormant Commerce Clause only when the challenged statute is discriminatory or protectionist in nature, appears to be another relevant consideration. See Peter C. Felmly, Comment, Beyond the Reach of States: The Dormant Commerce Clause, Extraterritorial State Regulation, and the Concerns of Federalism, 55 Me. L.Rev. 467, 491 (2003) (noting that recent Supreme Court cases considering the dormant Commerce Clause suggest an increased "focus on the territorial reach of state legislation . . . in stark contrast to the long-established concentration on state regulations that are discriminatory or protectionist in nature").

I have said enough to demonstrate the complexity of the Commerce Clause issue and the inadequacy of the record. There are missing details about how the prescriber-identifiable data generated by New Hampshire pharmacies flows to corporate offices out of state and the purpose of that information flow. The parties appear to have different assumptions about those details and their legal significance. Moreover, the plaintiffs' argument on the Commerce Clause spans only two and one-half pages in their sixty-page brief. The Attorney General's response is equally terse. I think it unwise to address the Commerce Clause issue based on a cursory briefing that provides neither legal analysis nor developed application of the law to the limited facts of record. Although the parties agreed at trial that the facts on the Commerce Clause claim were undisputed and that no further evidence was needed to resolve it, the plaintiffs do not address that evidence in any meaningful way in their briefs and the Attorney General does not address the evidence at all. The district court did not reach the claim.

Our comment about a similarly bare Commerce Clause claim in Wine Spirits II also should guide us here: "This sophisticated area of law requires developed argumentation, with evidentiary support." 481 F.3d at 15 (noting that the Supreme Court had "label[ed] as a 'critical consideration' regarding extraterritorial reach claims the 'overall effect of the statute on both local and interstate commerce'" (quoting Healy v. Beer Inst, Inc., 491 U.S. 324, 337 n. 14, 109 S.Ct. 2491, 105 L.Ed.2d 275 (1989))). I therefore would remand this case to the district court on the Commerce Clause issue.

VI.

I summarize my conclusions:

1. The prudential standing doctrine is inapplicable in the circumstances of this case, where the core First Amendment issue was vigorously litigated and comprehensively considered by the district court, and where the Prescription Information Act's constitutionality cannot be assessed without addressing its impact on the communications between detailers and prescribers;

2. The Act restricts commercial speech that is protected by the First Amendment, and the Attorney General therefore bears the burden of demonstrating that the statute satisfies the Central Hudson test;

3. Although the State has failed to prove that the Act is justified by substantial interests in privacy and quality health care, it has met its burden to show that the Act directly advances its interest in containing the cost of prescription drugs and is not more extensive than necessary to accomplish that objective.

4. Like the majority, I find the Prescription Act sufficiently clear to withstand plaintiffs' vagueness challenge when construed narrowly, consistent with its legislative history and applicable precedent.

5. The plaintiffs' contention that the Act violates the dormant Commerce Clause should be considered by the district court in the first instance. We should remand the case for that purpose.