dissenting:
There are few more annoying occurrences than the disruption of one’s sleep due to a late-night or early-morning phone call by a telephone solicitor. Whether the end result is simply mild frustration (due to an ill-timed solicitation) or actual fear (due to the phenomenon of “abandoned calls”), the invasion of privacy that the majority identifies due to unwanted telephone solicitations is certainly a compelling interest that government can seek to remedy. I do not disagree with the majority’s assessment that the regulations at issue in this case are modest attempts by the Federal Trade Commission (“FTC”) to provide some respite to the public from the barrage of such intrusions. Were these regulations to apply equally to all charities that seek to engage in telephone solicitations, I would concur in the majority’s decision to affirm. However, because the TSR in my view impermissibly distinguishes between the type of speech permitted based upon the identity of the speaker without any legitimate neutral justification, I must unfortunately, respectfully, dissent.
I.
As the majority correctly points out, a law or regulation is underinclusive when it restricts the speech of one group of speakers but not that of another similarly situated group without a legitimate “neutral justification” for the distinction. See City of Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 429-30, 113 S.Ct. 1505, 123 L.Ed.2d 99 (1993). Even where, as here, the government has a compelling interest in regulating a particular type of speech, its distinctions between similarly situated actors must reflect a “reasonable fit” between the restriction and the goal to be achieved by the disparate treatment. Id. at 417, 113 S.Ct. 1505. If the distinction restricts speech in a way that does not achieve its stated goal, or is made without sufficient evidentiary justification, the regulation is constitutionally underinclusive.
In Riley v. Nat’l Fed. Of the Blind of North Carolina, Inc. 487 U.S. 781, 108 S.Ct. 2667, 101 L.Ed.2d 669 (1988), the Supreme Court made clear that the two groups at issue here, charities that use “telefunders” and those that keep solicitation in-house, are similarly situated actors. Riley involved a North Carolina statute requiring all professional fundraisers to disclose to potential donors the percentage of the charitable contributions collected in the previous twelve months that was actually turned over to the charity. Id. at 784 *352n. 2, 108 S.Ct. 2667. The Supreme Court struck down the requirement, holding that the state’s interest in informing donors about the charity’s fundraising agreement did not outweigh the fundraisers’ significant speech interests. Id. at 798-99, 108 S.Ct. 2667.
Most importantly for our purposes, the Court noted that the statute’s key deficiency was that it necessarily only applied to those charities that used professional fundraisers. Id. at 784, 108 S.Ct. 2667. The Court recognized that in addition to the heightened scrutiny given to all regulations affecting charitable solicitations, courts should be especially cognizant of the comparative effect regulations that distinguish between charities based upon the professional/volunteer status of the solicitor can have on the success of the charity itself. See id. at 799, 108 S.Ct. 2667. Because charities using professional solicitors are likely to be smaller and less popular than those relying on in-house efforts, the statute’s application of the financial disclosure requirements to these professional fundraisers would place them at a competitive disadvantage. The state in Riley argued that even under this heightened scrutiny, its regulation was justified due to the increased risk of fraud inherent in telefun-der solicitation. The Court specifically rejected this assertion, holding that there exists “no nexus” between the percentage of funds retained by a solicitor and the likelihood of fraud. Id. at 793, 108 S.Ct. 2667. Because regulations that distinguish between charities can affect the financial success of one vis a vis another, such distinctions must have a compelling justification. With no evidence to support the state of North Carolina’s bald assertion of a greater likelihood of fraud, the court held that the state could not treat these similarly situated actors differently, and the distinction could not be upheld. Id. at 798-99, 108 S.Ct. 2667. It is against this backdrop that we must review the regulation now before us.1
The FTC asserts that it intended the amended TSR to further the dual goals of fraud protection and consumer privacy. However, it is unclear that either of these goals is advanced in any material way by the TSR’s inclusion of only solicitations made by telefunders rather than the broader category of all charitable solicitors in its speech restrictions. The district court found sufficient the FTC’s declaration that a professional telefunder’s financial interest in obtaining greater contributions leads it to be more likely to engage in fraudulent behavior than a non-professional solicitor. Nat’l Fed. of the Blind v. F.T.C., 303 F.Supp.2d 707, 721 (D.Md. 2004). However, the FTC, when pressed at oral argument, conceded that there existed “no empirical evidence ” in the administrative record to support this assumption. In fact, the FTC conceded that the only evidentiary support the agency had for solely applying the TSR to telefun-ders was its intuitive, “common-sense” assumption that telefunders commit fraud more frequently. This is the exact assumption rejected by Riley. Further, as it is not grounded on any fact that the FTC or this court has before it, I cannot accept it as a “neutral justification” for the speech limitation distinction.2
*353The same evidentiary deficiencies undercut the FTC’s conclusion that its regulations advance the goal of protecting consumer privacy. While it cannot be doubted that the government’s rules will prevent some degree of consumer annoyance, the FTC again presents no evidence that tele-funders are more likely to be violators of consumer privacy or engage in abusive telemarketing practices than in-house solicitors. There is no suggestion in the record that consumers are more likely to feel that their privacy is invaded when receiving a call from a telefunder than a volunteer or in-house employee of a charitable organization. The FTC has also provided no evidence showing that telefunders generate a majority, or even a significant percentage of the charitable solicitation calls. It is not enough for us to accept the agency’s rationale that the regulation is justifiable because it will help consumers in some small way. See Discovery Network, 507 U.S. at 426, 113 S.Ct. 1505, 123 L.Ed.2d 99 (1993)(holding that a selective ban on commercial newsracks was not justifiable simply because it would “in some small way limit the total number of news-racks.”). In order for it to pass the under-inclusiveness test, the FTC must show that the evidence illustrates that it makes sense to treat these similarly situated charities differently. This it does not do.
It is certainly not the case that the First Amendment requires the government to regulate all aspects of a problem before regulating any part of it. United States v. Edge Broad. Co., 509 U.S. 418, 434, 113 S.Ct. 2696, 125 L.Ed.2d 345 (1993). We generally give the government the latitude it needs to find ways to address problems in society in the manner in which it sees fit, and those judgments are deserving of deference.3 However, where, as here, no evidentiary basis has been articulated for the disparate treatment of similarly situated groups, such deference is unwarranted. See Florida Star v. B.J.F., 491 U.S. 524, 537-38, 109 S.Ct. 2603, 105 L.Ed.2d 443 (1989). The FTC has made no showing that the professional/non-professional distinction it employs in the TSR furthers the goals of fraud prevention and consumer privacy. It is this showing that Discovery Network’s “neutral justification” and “reasonable fit” requirements compel. 507 U.S. at 428-30, 113 S.Ct. 1505. Because these requirements are not satisfied, the TSR cannot survive.
The majority characterizes my analysis as reflecting a belief that government must act in an “all-or-nothing” manner. It asserts that I believe that the government cannot undertake “messy” solutions, but rather must only make “sweeping pro*354nouncements.” Majority Op. at 349. It is unclear to me how such broadly painted assumptions about my beliefs concerning the role of government could be ascertained from such a narrow dissent; in any event, nothing could be further from the truth. Rather, I simply believe that the Supreme Court meant what it said in Discovery Network that without a “neutral justification,” neither Congress nor executive agencies can create speech restrictions that treat similarly-situated actors differently. 507 U.S. at 428-30, 113 S.Ct. 1505. The FTC had ample opportunity during the rule-making process to create an evidentiary record to support the distinction it has drawn. It failed to do so, and the majority’s miseharacterization of my views and expressions of concern for the “democratic process” and “our structure of government” are no substitute. The government is certainly able to “make distinctions or to strike compromises,” Majority Op. at 349, but in doing so, it must abide by the First Amendment. Here, the FTC did not.
II.
Seeking to save the FTC from these evidentiary deficiencies, the majority holds that the distinction the TSR draws between in-house fundraisers and external telefunders is justified due to the agency’s jurisdictional restraints. The FTC Act gives the agency jurisdiction over “persons, partnerships and corporations,” but no authority over nonprofit organizations. 15 U.S.C. § 45(a)(2). The majority correctly notes that without the jurisdiction to regulate all nonprofit organizations, the FTC has no ability to correct the underin-clusiveness inherent in the TSR. According to the majority, some degree of underin-clusiveness is inevitable in all agency regulation, and an agency’s jurisdictional boundaries can therefore serve as the “neutral justification” for an underinclusive distinction.
The implications of this holding are staggering. If a regulation that places different restrictions on speech based upon the identity of the speaker can be upheld simply by relying on the jurisdiction of the agency as the “neutral justification” for the distinction, this court will have created a perverse incentive for all legislative bodies. Congress can restrict speech, even unconstitutionally, so long as it does so by parsing jurisdiction between various agencies. If, for example, Congress were to give a particular agency jurisdiction over one political party — for instance if the Department of Homeland Security were given jurisdiction over the Independent party in America — the majority’s holding would allow a regulation that restricted that party’s ability to raise funds to be upheld simply by referring to the agency’s jurisdiction.4 Restrictions on speech that go to the heart of the First Amendment, those differentiating solely on the identity of the speaker, would be given little to no scrutiny, as the legislative decision to disperse jurisdiction amongst various executive agencies would justify the disparities. The *355majority would allow Congress to do indirectly that which it cannot do directly, and that is to regulate speech in an underinclu-sive manner.
The majority’s sole focus on the FTC’s jurisdiction and that agency’s inability to promulgate an even-handed regulation overlooks the fact that Congress is well able to correct the constitutional deficiency.5 In this case it could have done so by simply expanding the FTC’s jurisdiction to include all charitable solicitations, thus authorizing the agency to pass a regulation that advances the goals of fraud protection and consumer privacy materially and constitutionally. Congress could have also simply required the Federal Communications Commission, the agency with oversight authority over all non-telefunder charitable solicitation, to pass a similar regulation, thus insuring that all charities were on a level playing field in their quest to solicit funds. However what it cannot do is hide behind jurisdictional barriers which it erected as a shield for its unconstitutional restrictions on speech based upon the identity of the speaker.6
III.
The Supreme Court has recognized that the ability to raise funds is the lifeblood of a charity. Riley, 487 U.S. at 789, 108 S.Ct. 2667. For very large or very popular charities, fundraising can be conducted in-house through employees of the charities or volunteers. The use of volunteers can even be a selling point to potential donors, as an individual’s willingness to dedicate his or her free time to a cause suggests a strong belief in its goals. However, for small or unpopular charities, the ability to engage in mass volunteer fundraising may be impossible, and thus the use of professional fundraisers may hold the key to the entity’s existence.
As charities compete for the finite charitable donation dollar, those charities that either choose to or are forced to use tele-funders are placed at a competitive disadvantage by the regulations that affect them and not their charitable counterparts. It may very well be the case that following the regulations laid down in the TSR would be a good practice for all charities. Calling consumers early in the morning or in the middle of the night appears unlikely to provide the most successful fundraising strategy. However, whatever decisions the individual charities make, Riley and Discovery Network compel the conclusion that the government must apply an even hand. The FTC has no evidence and no reason beyond jurisdictional authority justifying its decision to regulate charitable solicitations differently based upon the identity of the speaker.
The majority holds that this suffices, and that Congress can justify unconstitutional restrictions on speech based upon the identity of the speaker simply by delineating jurisdiction over those speakers be*356tween various executive agencies. This argument is untenable. It provides a jurisdictional loophole through which government can achieve indirectly that which it cannot do directly. I do not believe that such jurisdictional parsing can be a “neutral justification” for a regulation that imposes different speech restrictions based upon the identity of the speaker. I therefore respectfully dissent.
. In its well-reasoned opinion, the majority considers Riley but does not address its implications concerning underinclusiveness. While the regulations at issue here may be more limited than in Riley, there is no indication that its holding that charities are similarly situated actors that must be treated equally without compelling justification is any less applicable.
. The majority is correct in asserting that our review is not one for "substantial evidence,” but rather is a question as to whether we have a substantial basis for finding the regulation *353unconstitutional. Majority Op. at 347. Here we most certainly do. The lack of evidence (as opposed to "common-sense” assumptions) providing a "neutral justification” for the distinction between telefunders and in-house solicitors requires the finding of a First Amendment violation.
Further, even if such so-called "common sense” assumptions were a sufficient eviden-tiary basis for speech distinctions (and Discovery Network counsels that they are not), I would not accept this one. It could just as plausibly be argued that a solicitation conducted by someone with a personal interest in the success of a charity (such as an employee or a volunteer) may lead to more overreaching and thus a greater potential for fraud. Unlike the FTC, I cannot accept that one’s strongest motivations are always monetary.
. The majority's deference borders on subservience. By suggesting that after the judiciary finds a constitutional right, as in Riley, it is then up to the legislative and executive branches to define the scope of that right, the majority seems to suggest a far too limited role for the courts. However the "elaborate dance” between the governmental branches is to proceed, the judiciary cannot relinquish its role as the ultimate protector of constitutional guarantees.
. The majority criticizes this analogy without providing any basis for distinguishing it from the situation before us. The government could disperse jurisdiction over political parties among various agencies without "exhibiting] any disapproval of the content” of the parties’ messages. See Majority Op. at 350. If one of such agencies were then to create a set of restrictions that apply to all parties within its jurisdiction, and "impose burdens on speech without reference to the ideas or views expressed,” the majority’s reasoning would require us to uphold the restriction. Significantly, the majority is unable to point to any language in its opinion which would preclude such a result. The majority's unexplained assertion that my analogy is "so far afield” from the regulation at hand as to be "apples and oranges” reflects a conclusion rather than an analysis.
. The majority suggests that a "correction” by Congress of an agency’s jurisdiction would in and of itself, be underinclusive. However it posits no basis for this proposition. If true, all Congressional grants of jurisdiction that differentiate between groups would be under-inclusive. The First Amendment does not require this radical result; it simply requires some legitimate basis for differential treatment.
. The majority’s concern about judicial overreaching is understandable but inapplicable here. If Congress and the executive agencies have evidentiary reasons for their respective decisions to pass regulations that have the effect of treating similar groups differently, they may well pass constitutional muster. However when, as here, no such evidentiary basis exists, the effect of this Congressional choice is to create an unconstitutional restriction on speech. It is not overreaching to prevent this result.