concurring in the judgment:
The trichotomy of strict, intermediate and rational-basis scrutiny with its judicial ranking of “compelling,” “important,” and “legitimate” governmental interests, a trichotomy devised in equal protection cases, seems to me out of place when it comes to analyzing issues concerning the freedom of speech. To be sure, in Simon & Schuster, Inc. v. New York State Crime Victims Board, — U.S. -, 112 S.Ct. 501, 116 L.Ed.2d 476 (1991), the Court used Fourteenth Amendment “compelling interest” language to decide a First Amendment case. See also Austin v. Michigan State Chamber of Commerce, 494 U.S. 652, 657, 110 S.Ct. 1391, 1396, 108 L.Ed.2d 652 (1990). This prompted Justice Kennedy to examine the extension’s, evolution, only to discover that the Court had “adopted this formulation in First Amendment eases by accident rather than as the result of a considered judgment.” Id. 112 S.Ct. at 513 (Kennedy, J., concurring); see also Burson v. Freeman, — U.S. -, 112 S.Ct. 1846, 1850 n. 3, 119 L.Ed.2d 5 (1992) (plurality opinion); id. at -, 112 S.Ct. at 1858 (Kennedy, J., concurring). The intermediate standard of review, created for gender discrimination cases in Craig v. Boren, 429 U.S. 190, 97 S.Ct. 451, 50 L.Ed.2d 397 (1976), is another matter. The Court has done little more than flirt with the idea of applying it to First Amendment claims. The controlling opinion in Barnes v. Glen Theatres, Inc., — U.S. -, 111 S.Ct. 2456, 2461, 115 L.Ed.2d 504 (1991), offered a variant of intermediate scrutiny to decide whether nude dancing is entitled to First Amendment protection, but that view commanded only three votes. See id. 111 S.Ct. at 2467-68 (Scalia, J., concurring in the judgment). A majority of the court embarked on an intermediate scrutiny course with respect to free speech rights of prisoners in Procunier v. Martinez, 416 U.S. 396, 413, 94 S.Ct. 1800, 1811, 40 L.Ed.2d 224 (1974), thereafter quoted portions of Pro-cunier with approval in Seattle Times Co. v. Rinehart, 467 U.S. 20, 32, 104 S.Ct. 2199, 2207, 81 L.Ed.2d 17 (1984), and then retreated in Thornburgh v. Abbott, 490 U.S. 401, 408-14, 109 S.Ct. 1874, 1879-82, 104 L.Ed.2d 459 (1989). Regan v. Taxation with Representation, 461 U.S. 540, 103 S.Ct. 1997, 76 L.Ed.2d 129 (1983), did not even flirt with the idea. Then-Justice Rehnquist’s opinion for the Court sustained, against a First Amendment challenge, Congress’ refusal to allow organizations engaging in substantial lobbying to remain eligible for tax deductible contributions on the basis that it was “not irrational” for Congress to have made the distinctions that it did. Id. at 550, 103 S.Ct. at 2003; see also International Soc’y for Krishna Consciousness, Inc. v. Rumbaugh, — U.S. -, -, 112 S.Ct. 2701, 2706, 120 L.Ed.2d 541 (1992).
At all events, I doubt that the majority in this case truly means to apply the sort of intermediate scrutiny employed in the context of the Fourteenth Amendment. In such cases, courts have required that statutory classifications have factual support. Cf. Burlington Northern R.R. v. Ford, — U.S. -, -, 112 S.Ct. 2184, 2187, 119 L.Ed.2d 432 (1992). The government must prove that the means chosen are “substantially related” to achieving “important” governmental objectives. See, e.g., Craig v. Boren, 429 U.S. at 197, 97 S.Ct. at 456; Metro Broadcasting, Inc. v. FCC, 110 S.Ct. 2997, 3008-09 (1990). Without empirical support, the provision will not survive. That is the point of now-Justice Thomas’ opinion for this court in Lamprecht v. *1121FCC, 958 F.2d 382 (D.C.Cir.1992). I doubt whether such stringent requirements of proof, although an integral part of intermediate scrutiny, can or should be transplanted into the First Amendment area. The accuracy of legislative judgments giving rise to free speech claims is not, as a general matter, readily susceptible to empirical demonstration. Cf. Burson v. Freeman, 112 S.Ct. at 1856 (plurality opinion).
The majority’s opinion does not convince me that the proviso limiting 2 U.S.C. § 438(a)(4) could survive such stringent analysis. A last-minute floor amendment passed with limited discussion (see 117 Cong.Reo. 30,057-58 (1971)), the proviso was not supported by any legislative factual findings. The Federal Election Commission has not filled the void in its presentation to us. The record contains no factual support for my colleagues’ assumption that the proviso will increase the overall level of speech. Maj. op. at 1117. Allowing others to use contributor lists may reduce the return on a political committee’s investment in developing them. But we cannot be certain of this and still less can we say how much the proviso will reduce that return. If others were free to exploit the lists, the developers might increase their level of speech in order to remain the contributor’s exclusive object. From the standpoint of the contributors, they might be subjected to more solicitations from others competing for their dollars, which is to say that there would be more speech without the proviso than with it. We simply do not know the true effect of the proviso and, from all that appears, neither did Congress. Under the usual intermediate scrutiny review, such uncertainty would lead to unconstitutionality.
Nonetheless, I agree with the result my colleagues reach. The case seems to me rather easy. As both the majority opinion and Judge Ruth Bader Ginsburg’s concurring opinion discuss, the proviso neither directly limits the ability of citizens to speak nor conditions the receipt of a government benefit on the waiver of the right to speak. Nor does it place restrictions on the use of information the government itself has generated and released to the public. The proviso simply restricts a citizen’s right to use another’s property (here, the contributor lists), property to which no one other than the owner has any separate, independent claim. Compare Pruneyard Shopping Center v. Robins, 447 U.S. 74, 81, 100 S.Ct. 2035, 2040, 64 L.Ed.2d 741 (1980); Lloyd Corp. v. Tanner, 407 U.S. 551, 92 S.Ct. 2219, 33 L.Ed.2d 131 (1972). In this respect, it is akin to the protection the Freedom of Information Act gives to trade secrets and other privileged or confidential information (5 U.S.C. § 552(b)(4)) and the trade secret protection some states accord customer lists. R. Mil-grim, Milgrim on Trade Secrets § 2.09[7][a][l], at 2-253 & n. 281 (1992). It is certainly “not irrational” for Congress to extend similar protection to a political committee’s contributor lists. If the government forced political committees to disclose their lists to the public without imposing any restrictions on how the lists might be used, this might constitute a taking under the Fifth Amendment. Cf. Ruckelshaus v. Monsanto Co., 467 U.S. 986, 1011-12 & n. 15, 104 S.Ct. 2862, 2877, & n. 15, 81 L.Ed.2d 815 (1984). The proviso preserves the property, avoids a taking, encourages compliance with the disclosure laws by removing an incentive to disobey them, and arguably enhances speech by allowing political committees to reap the full benefit of their labors. For its part, the defendant International Funding Institute did not have “a First Amendment right to ‘appropriat[e] to itself the harvest of those who have sown.’ ” San Francisco Arts & Athletics v. United States Olympic Comm., 483 U.S. 522, 541, 107 S.Ct. 2971, 2983, 97 L.Ed.2d 427 (1987) (quoting International News Serv. v. The Associated Press, 248 U.S. 215, 239-40, 39 S.Ct. 68, 72-73, 63 L.Ed. 211 (1918)). The proviso draws a line “to assure the due protection” of the rights of private property owners and the interests of citizens in learning the identity of contributors. Lloyd Corp., 407 U.S. at 570, 92 S.Ct. at 2229. The accommodation thus made, while perhaps not constitutionally compelled, is constitutionally permissible.