with whom Justice Rehnquist joins, dissenting.
Historically our society has gone to great lengths to protect minors charged with crime, particularly by prohibiting the release of the names of offenders, barring the press and public from juvenile proceedings, and sealing the records of those proceedings. Yet today the Court holds unconstitutional a state statute designed to protect not the accused, but the minor victims of sex crimes. In doing so, it advances a disturbing paradox. Although states are permitted, for example, to mandate the closure of all proceedings in order to protect a 17-year-old charged with rape, they are not permitted to require the closing of part of criminal proceedings in order to protect an innocent child who has been raped or otherwise sexually abused.
The Court has tried to make its holding a narrow one by not disturbing the authority of state legislatures to enact more narrowly drawn statutes giving trial judges the discretion to exclude the public and the press from the courtroom during the minor victim's testimony. Ante, at 611, n. 27. I also do not read the Court’s opinion as foreclosing a state statute which mandates closure except in cases where the victim agrees to testify in open court.1 But the Court’s deci*613sion is nevertheless a gross invasion of state authority and a state’s duty to protect its citizens — in this case minor victims of crime. I cannot agree with the Court’s expansive interpretation of our decision in Richmond Newspapers, Inc. v. Virginia, 448 U. S. 555 (1980), or its cavalier rejection of the serious interests supporting Massachusetts’ mandatory closure rule. Accordingly, I dissent.
I
The Court seems to read our decision in Richmond Newspapers, supra, as spelling out a First Amendment right of access to all aspects of all criminal trials under all circumstances. Ante, at 605, n. 13. That is plainly incorrect. In Richmond Newspapers, we examined “the right of access to places traditionally open to the public” and concluded that criminal trials were generally open to the public throughout this country’s history and even before that in England. The opinions of a majority of the Justices emphasized the historical tradition of open criminal trials. 448 U. S., at 564-573; id., at 580-591 (Brennan, J., concurring in judgment); id., at 599 (Stewart, J., concurring in judgment); id., at 601 (Blackmun, J., concurring in judgment). The proper mode of analysis to be followed in determining whether there is a right of access was emphasized by Justice Brennan:
*614“As previously noted, resolution of First Amendment public access claims in individual cases must be strongly influenced by the weight of historical practice and by an assessment of the specific structural value of public access in the circumstances.” Id., at 597-598.
Today Justice Brennan ignores the weight of historical practice. There is clearly a long history of exclusion of the public from trials involving sexual assaults, particularly those against minors. See, e. g., Harris v. Stephens, 361 F. 2d 888 (CA8 1966), cert. denied, 386 U. S. 964 (1967); Reagan v. United States, 202 F. 488 (CA9 1913); United States v. Geise, 158 F. Supp. 821 (Alaska), aff’d, 262 F. 2d 151 (CA9 1958), cert. denied, 361 U. S. 842 (1959); Hogan v. State, 191 Ark. 437, 86 S. W. 2d 931 (1935); State v. Purvis, 157 Conn. 198, 251 A. 2d 178 (1968), cert. denied, 395 U. S. 928 (1969); Moore v. State, 151 Ga. 648, 108 S. E. 47 (1921), appeal dism’d, 260 U. S. 702 (1922).2 Several States have longstanding provisions allowing closure of cases involving sexual assaults against minors.3
It would misrepresent the historical record to state that there is an “unbroken, uncontradicted history” of open proceedings in cases involving the sexual abuse of minors. Richmond Newspapers, supra, at 573. Absent such a history of openness, the positions of the Justices joining reversal in Richmond Newspapers give no support to the proposition that closure of the proceedings during the testimony of the minor victim violates the First Amendment.4
*6151 — 4 I — I
The Court does not assert that the First Amendment right it discerns from Richmond, Newspapers is absolute; instead, it holds that when a “State attempts to deny the right of access in order to inhibit the disclosure of sensitive information, it must be shown that the denial is necessitated by a compelling governmental interest, and is narrowly tailored to serve that interest.” Ante, at 606-607. The Court’s wooden application of the rigid standard it asserts for this case is inappropriate. The Commonwealth has not denied the public or the media access to information as to what takes place at trial. As the Court acknowledges, Massachusetts does not deny the press and the public access to the trial transcript or to other sources of information about the victim’s testimony. Even the victim’s identity is part of the public record, although the name of a 16-year-old accused rapist generally would not be a matter of public record. Mass. Gen. Laws Ann., ch. 119, § 60A (West Supp. 1982-1983). The Commonwealth does not deny access to information, and does nothing whatever to inhibit its disclosure. This case is quite unlike others in which we have held unconstitutional state laws which prevent the dissemination of information or the public discussion of ideas. See, e. g., Brown v. Hartlage, 456 U. S. 45 (1982); Smith v. Daily Mail Publishing Co., 443 U. S. 97 (1979); Landmark Communications, Inc. v. Virginia, 435 U. S. 829 (1978); Nebraska Press Assn. v. Stuart, 427 U. S. 539 (1976); Cox Broadcasting Corp. v. Cohen, 420 U. S. 469 (1975); NAACP v. Button, 371 U. S. 415 (1963).
The purpose of the Commonwealth in enacting § 16A was to give assurance to parents and minors that they would have this moderate and limited protection from the trauma, embarrassment, and humiliation of having to reveal the intimate details of a sexual assault in front of a large group of unfamiliar spectators — and perhaps a television audience— and to lower the barriers to the reporting of such crimes which might come from the victim’s dread of public testimony. Globe Newspaper Co. v. Superior Court, 379 Mass. *616846, 865, 401 N. E. 2d 360, 372 (1980); 383 Mass. 838, 847-848, 423 N. E. 2d 773, 779 (1981).
Neither the purpose of the law nor its effect is primarily to deny the press or public access to information; the verbatim transcript is made available to the public and the media and may be used without limit. We therefore need only examine whether the restrictions imposed are reasonable and whether the interests of the Commonwealth override the very limited incidental effects of the law on First Amendment rights. See Richmond Newspapers, 448 U. S., at 580-581 (plurality opinion); id., at 600 (Stewart, J., concurring in judgment); Pell v. Procunier, 417 U. S. 817 (1974); Saxbe v. Washington Post Co., 417 U. S. 843 (1974); Cox v. New Hampshire, 312 U. S. 569 (1941). Our obligation in this case is to balance the competing interests: the interests of the media for instant access, against the interest of the State in protecting child rape victims from the trauma of public testimony. In more than half the states, public testimony will include television coverage.
Ill
For me, it seems beyond doubt, considering the minimal impact of the law on First Amendment rights and the overriding weight of the Commonwealth's interest in protecting child rape victims, that the Massachusetts law is not unconstitutional. The Court acknowledges that the press and the public have prompt and full access to all of the victim’s testimony. Their additional interest in actually being present during the testimony is minimal. While denying it the power to protect children, the Court admits that the Commonwealth’s interest in protecting the victimized child is a compelling interest. Ante, at 607. This meets the test of Richmond Newspapers, supra.
The law need not be precisely tailored so long as the state’s interest overrides the law’s impact on First Amendment rights and the restrictions imposed further that interest. Certainly this law, which excludes the press and public only *617during the actual testimony of the child victim of a sex crime, rationally serves the Commonwealth’s overriding interest in protecting the child from the severe — possibly permanent— psychological damage. It is not disputed that such injury is a reality.5
The law also seems a rational response to the undisputed problem of the underreporting of rapes and other sexual offenses. The Court rejects the Commonwealth’s argument that § 16A is justified by its interest in encouraging minors to report sex crimes, finding the claim “speculative in empirical terms [and] open to serious question as a matter of logic and common sense.” Ante, at 609-610. There is no basis whatever for this cavalier disregard of the reality of human experience. It makes no sense to criticize the Commonwealth for its failure to offer empirical data in support of its rule; only by allowing state experimentation may such empirical evidence be produced. “It is one of the happy incidents of the federal system that a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.” New State Ice Co. v. Liebmann, 285 U. S. 262, 311 (1932) (Brandeis, J., dissenting). See also Chandler v. Florida, 449 U. S. 560, 579-580 (1981); Reeves, Inc. v. Stake, 447 U. S. 429, 441 (1980); Whalen v. Roe, 429 U. S. 589, 597, and n. 20 (1977).
The Court also concludes that the Commonwealth’s assertion that the law might reduce underreporting of sexual offenses fails “as a matter of logic and common sense.” This conclusion is based on a misperception of the Commonwealth’s argument and an overly narrow view of the protection the statute seeks to afford young victims. The Court apparently believes that the statute does not prevent any sig*618nificant trauma, embarrassment, or humiliation on the part of the victim simply because the press is not prevented from discovering and publicizing both the identity of the victim and the substance of the victim’s testimony. Ante, at 609-610. Section 16A is intended not to preserve confidentiality, but to prevent the risk of severe psychological damage caused by having to relate the details of the crime in front of a crowd which inevitably will include voyeuristic strangers.6 In most states, that crowd may be expanded to include a live television audience, with reruns on the evening news. That ordeal could be difficult for an adult; to a child, the experience can be devastating and leave permanent scars.7
The Commonwealth’s interests are clearly furthered by the mandatory nature of the closure statute. Certainly if the law were discretionary, most judges would exercise that discretion soundly and would avoid unnecessary harm to the child, but victims and their families are entitled to assurance of such protection. The legislature did not act irrationally in deciding not to leave the closure determination to the idio-syncracies of individual judges subject to the pressures avail*619able to the media. The victim might very well experience considerable distress prior to the court appearance, wondering, in the absence of such statutory protection, whether public testimony will be required. The mere possibility of public testimony may cause parents and children to decide not to report these heinous crimes. If, as psychologists report, the courtroom experience in such cases is almost as traumatic as the crime itself,8 a state certainly should be able to take whatever reasonable steps it believes are necessary to reduce that trauma. Furthermore, we cannot expect victims and their parents to be aware of all of the nuances of state law; a person who sees newspaper, or perhaps even television, reports of a minor victim’s testimony may very well be deterred from reporting a crime on the belief that public testimony will be required. It is within the power of the state to provide for mandatory closure to alleviate such understandable fears and encourage the reporting of such crimes.
IV
There is, of course, “a presumption of openness [that] inheres in the very nature of a criminal trial under our system of justice.” But we have consistently emphasized that this presumption is not absolute or irrebuttable. A majority of the Justices in Richmond Newspapers acknowledged that closure might be permitted under certain circumstances. Justice Stewart’s separate opinion pointedly recognized that exclusion of the public might be justified to protect “the sensibilities of a youthful prosecution witness ... in a criminal trial for rape.” 448 U. S., at 600, n. 5.9 The Massachusetts statute has a relatively minor incidental impact on First *620Amendment rights and gives effect to the overriding state interest in protecting child rape victims. Paradoxically, the Court today denies the victims the kind of protection routinely given to juveniles who commit crimes. Many will find it difficult to reconcile the concern so often expressed for the rights of the accused with the callous indifference exhibited today for children who, having suffered the trauma of rape or other sexual abuse, are denied the modest protection the Massachusetts Legislature provided.
Justice Stevens,dissenting.
The duration of a criminal trial generally is shorter than the time it takes for this Court’s jurisdiction to be invoked and our judgment on the merits to be announced. As a result, our power to review pretrial or midtrial orders implicating the freedom of the press has rested on the exception to the mootness doctrine for orders “capable of repetition, yet evading review.” See Richmond Newspapers, Inc. v. Virginia, 448 U. S. 555, 563; Gannett Co. v. DePasquale, 443 U. S. 368, 377-378; Nebraska Press Assn. v. Stuart, 427 U. S. 539, 546-547.
Today the Court expands that exception in order to pass on the constitutionality of a statute that, as presently construed, has never been applied in a live controversy. In this case, unlike the three cases cited above, the governing state law was materially changed after the trial court’s order had expired by its own terms. There consequently is no possibility “ ‘that the same complaining party will be subject to the same action again.’” Gannett Co. v. DePasquale, supra, at 377 (quoting Weinstein v. Bradford, 423 U. S. 147, 149).
The fact that the Massachusetts Supreme Judicial Court narrowly construed — and then upheld in the abstract — the state statute that the trial court had read to mandate the closure of the entire trial bears on our review function in other respects. We have only recently recognized the First *621Amendment right of access to newsworthy matter. See ante, at 603; Richmond Newspapers, Inc. v. Virginia, supra, at 582 (Stevens, J., concurring). In developing constitutional jurisprudence, there is a special importance in deciding cases on concrete facts. Cf. Minnick v. California Dept. of Corrections, 452 U. S. 105, 120-127; United States v. Raines, 362 U. S. 17, 21. Only in specific controversies can the Court decide how this right of access to criminal trials can be accommodated with other societal interests, such as the protection of victims or defendants. The advisory opinion the Court announces today sheds virtually no light on how such rights should be accommodated.
The question whether the Court should entertain a facial attack on a statute that bears on the right of access cannot be answered simply by noting that the right has its source in the First Amendment. See, e. g., Bates v. State Bar of Arizona, 433 U. S. 350, 380-381; Young v. American Mini Theatres, Inc., 427 U. S. 50, 61. For the right of access is plainly not coextensive with the right of expression that was vindicated in Nebraska Press Assn., supra.1 Because statutes that bear on this right of access do not deter protected activity in the way that other laws sometimes interfere with the right of expression, we should follow the norm of reviewing these statutes as applied rather than on their face.
It is not clear when, if ever, the Court will need to confront the question whether a mandatory partial-closure statute is unconstitutional. If the order hypothesized by the Supreme Judicial Court, instead of the trial court’s order, had actually been entered in this case, and if the press had been given prompt access to a transcript of the testimony of the minor victims, appellant might not even have appealed. At the *622very least the press, the prosecutor, and defense counsel would have argued the constitutionality of the partial-closure order in the context of the facts relevant to such an order, and a different controversy would have been framed for appellate review. In future cases the trial courts may voluntarily follow the direction of Justice Wilkins and make specific findings demonstrating a compelling state interest supporting the mandated partial-closure order. See 383 Mass. 838, 852-853, 423 N. E. 2d 773, 782 (concurring opinion). Or the record in future cases may plainly disclose a justification for a partiál closure that the Court would consider acceptable. Thus, aside from the illumination provided by live controversies, a decision to review only orders actually entered pursuant to the Massachusetts statute would advance the policy of avoiding the premature and unnecessary adjudication of constitutional questions;2 it is at least conceivable that no such order may ever have to be justified by the conclusion of the legislature that the mandatory closure of the trial during the testimony of a minor victim of a sex crime is necessary to serve important state interests.
The Court does not hold that on this record a closure order limited to the testimony of the minor victims would have been unconstitutional. Rather, the Court holds only that if ever such an order is entered, it must be supported by adequate findings. Normally, if the constitutional deficiency is the absence of findings to support a trial order, the Court would either remand for factfinding, or examine the record itself, before deciding whether the order measured up to constitutional standards. The infeasibility of this course of action — since no such order was entered in this case and since the order that was entered has expired — further demon*623strates that the Court’s comment on the First Amendment issues implicated by the Massachusetts statute is advisory, hypothetical, and, at best, premature.3
I would dismiss the appeal.
It certainly cannot be said that the victims in this case consented to testifying in open court. During a lobby conference prior to trial, the prosecutor informed the trial judge that she had interviewed the victims, that they were concerned about publicity, and would agree to press attendance only if certain guarantees could be given:
“Each of [the three victims] indicated that they had the same concerns and basically they are privacy concerns.
“The difficulty of obtaining any kind of guarantee that the press would not print their names or where they go to school or any personal data or take pictures of them or attempt to interview them, those concerns come from their own privacy interests, as welljis the fact that there are grandparents involved with a couple of these victims who do not know what hap*613pened and if they were to find out by reading the paper, everyone was concerned about what would happen then. And they stated that if it were at all possible to obtain a guarantee that this information would not be used, then they wouldn't object to the press being included. I explained that that is [a] very difficult guarantee to obtain because the Court cannot issue a conditional order, or anything like that, but I just wanted to put on the record what their concerns were and what they are afraid of.” App. 48a.
It is clear that the victims would “waive” the exclusion of the press only if the trial court gave them guarantees of strict privacy, guarantees that were probably beyond the authority of the court and which themselves would raise grave constitutional problems. See Oklahoma Publishing Co. v. District Court of Oklahoma County, 430 U. S. 308 (1977); Cox Broadcasting Corp. v. Cohn, 420 U. S. 469 (1975).
Cf. Stamicarbon, N.V. v. American Cyanamid Co., 506 F. 2d 532, 539-540 (CA2 1974), and cases cited therein.
See, e. g., Ala. Const., Art. VI, § 169 (1901) (repealed 1973); Fla. Stat. §918.16 (1979); Ga. Code § 81-1006 (1978); Miss. Const., Art. 3, § 26; N. H. Rev. Stat. Ann. § 632-A:8 (Supp. 1981); N. Y. Jud. Law § 4 (McKinney 1968); N. C. Gen. Stat. § 15-166 (Supp. 1981); Utah Code Ann. § 78-7-4 (1953).
It is hard to find a limiting principle in the Court’s analysis. The same reasoning might require a hearing before a trial judge could hold a bench conference or any in camera proceedings.
For a discussion of the traumatic effect of court proceedings on minor rape victims, see E. Hilberman, The Rape Victim 53-54 (1976); S. Katz & M. Mazur, Understanding the Rape Victim: A Synthesis of Research Findings 198-200 (1979), and studies cited therein.
As one commentator put it: “Especially in cases involving minors, the courts stress the serious embarrassment and shame of the victim who is forced to testify to sexual acts or whose intimate life is revealed in detail before a crowd of the idly curious.” Berger, Man’s Trial, Woman’s Tribulation: Rape Cases in the Courtroom, 77 Colum. L. Rev. 1, 88 (1977). The victim’s interest in avoiding the humiliation of testifying in open court is thus quite separate from any interest in preventing the public from learning of the crime. It is ironic that the Court emphasizes the failure of the Commonwealth to seal the trial transcript and bar disclosure of the victim’s identity. The Court implies that a state law more severely encroaching upon the interests of the press and public would be upheld.
See Hilberman, supra; L. Holmstrom & A. Burgess, The Victim of Rape: Institutional Reactions 222, 227 (1978); Berger, supra, at 88, 92-93; Libai, The Protection of the Child Victim of a Sexual Offense in the Criminal Justice System, 15 Wayne L. Rev. 977, 1021 (1969). Holmstrom and Burgess report that nearly half of all adult rape victims were disturbed by the public setting of their trials. Certainly the impact on children must be greater.
See Bohmer & Blumberg, Twice Traumatized: The Rape Victim and the Court, 58 Judicature 390 (1975); Katz & Mazur, supra; Holmstrom & Burgess, supra; Hilberman, supra; Berger, supra.
See also 448 U. S., at 580-581; id., at 582 (White, J., concurring); id., at 584 (Stevens, J., concurring); id., at 598 (Brennan, J., concurring in judgment).
For example, even though a reporter may have no right of access to a judge’s side-bar conference, it surely does not follow that the judge could enjoin publication of what a reporter might have learned about such a conference.
“But the most fundamental principle of constitutional adjudication is not to face constitutional questions but to avoid them, if at all possible.” United States v. Lovett, 328 U. S. 303, 320 (Frankfurter, J., concurring).
The “capable of repetition, yet evading review” exception to the mootness doctrine generally is compatible with our settled policy of avoiding the premature adjudication of constitutional questions, see Franks v. Bowman Transportation Co., 424 U. S. 747, 756, n. 8, for an order that is capable of repetition yet evading review generally is no less ripe for review the first time it is presented than it would be on subsequent occasions. But when the “order” that is presented for review the first time is formulated in the abstract, as was the ruling of the Supreme Judicial Court in this case, the policy requires the Court to defer review of such an order until it is entered in a live controversy.