concurring in part and dissenting in part.
I concur in the court’s judgment except insofar as it holds that 18 Pa.Cons.Stat. Ann. § 3209 (Supp.1991) (spousal notice) is unconstitutional. I also join all of the court’s opinion except for the portions concerning Section 3209 and those interpreting Justice O’Connor’s opinion in Hodgson v. Minnesota, — U.S. —, 110 S.Ct. 2926, 2949-51, 111 L.Ed.2d 344 (1990), to mean that the two-parent notification requirement without judicial bypass imposed an “undue burden” and was thus required to satisfy strict scrutiny.
*720I.
As the court suggests, the crux of this case concerns the identification of the constitutional standard that the lower courts must now apply in cases involving laws regulating abortion. For the reasons carefully explained in the court’s opinion, I agree that Webster v. Reproductive Health Services, 492 U.S. 490, 109 S.Ct. 3040, 106 L.Ed.2d 410 (1989), and Hodgson changed the law that we are bound to apply and that the test set out in Justice O’Connor’s opinions now represents the governing legal standard.
My disagreement with the majority regarding a single provision of the Pennsylvania Abortion Control Act, 18 Pa.Cons. StatAnn. § 3201 et seq. (1983 & Supp. 1991), results from disagreement about the portion of Justice O’Connor’s two-part test that must be applied to this provision. Under that test, as the majority explains, a law that imposes an “undue burden” must serve a “compelling” state interest. By contrast, a law that does not impose an “undue burden” must simply be “rationally” or “reasonably” related to a “legitimate” state interest. The majority holds that Section 3209 constitutes an undue burden. The majority therefore applies the first prong of the two-part test and strikes down Section 3209 on the ground that it does not serve a “compelling” interest. I do not believe that Section 3209 has been shown to impose an undue burden as that term is used in the relevant Supreme Court opinions; I therefore apply the second prong of the two-part test; and I conclude that Section 3209 is constitutional because it is “rationally related” to a “legitimate” state interest.
Although the majority and I apply different prongs of this two-part test, I see no indication that we disagree concerning the conclusion produced when either prong is applied to Section 3209. If the majority is correct that Section 3209 must satisfy heightened scrutiny, I agree that its constitutionality is doubtful. Similarly, I do not interpret the majority opinion to mean that Section 3209 cannot satisfy the rational relationship test. Indeed, the majority acknowledges that Section 3209 serves a “legitimate” interest. See majority opin. at 715, 716. Thus, my major disagreement with the majority concerns the question whether Section 3209 imposes an “undue burden,” and I will therefore turn to that question.
II.
A. Justice O’Connor has explained the meaning of the term “undue burden” in several abortion opinions. In Akron v. Akron Center for Reproductive Health, 462 U.S. at 464, 103 S.Ct. at 2510 (O’Connor, J., dissenting), she wrote that “an ‘undue burden’ has been found for the most part in situations involving absolute obstacles or severe limitations on the abortion decision.” She noted that laws held unconstitutional in prior cases involved statutes that “criminalized all abortions except those necessary to save the life of the mother,” inhibited “ ‘the vast majority of abortions after the first 12 weeks,’ ” or gave the parents of a pregnant minor an absolute veto power over the abortion decision. Id. (emphasis in original; citations omitted). She suggested that an “undue burden” would not be created by “a state regulation [that] may ‘inhibit’ abortions to some degree.” Id. She also suggested that there is no undue burden unless a measure has the effect of “substantially limiting access.” Id. at 463, 103 S.Ct. at 2509, quoting Carey v. Population Services International, 431 U.S. 678, 688, 97 S.Ct. 2010, 2017, 52 L.Ed.2d 675 (1977) (emphasis added in Justice O’Connor’s opinion).
Justice O’Connor reiterated the same analysis in Thornburgh v. American College of Obstetricians and Gynecologists, 476 U.S. 747, 106 S.Ct. 2169 (1986). She wrote (id. at 828, 106 S.Ct. at 2214 (O’Connor, J., dissenting), quoting Akron, 462 U.S. at 464, 103 S.Ct. at 2510 (O’Connor, J., dissenting)):
An undue burden would generally be found “in situations involving absolute obstacles or severe limitations on the abortion decision,” not wherever a state regulation “may ‘inhibit’ abortions to some degree.”
*721She also criticized the majority for taking an approach under which “the mere possibility that some women will be less likely to choose to have an abortion by virtue of the presence of a particular state regulation suffices to invalidate it.” Id. 476 U.S. at 829, 106 S.Ct. at 2214 (emphasis added).
Justice O’Connor’s application of the undue burden test in several cases further illustrates the meaning of this test. In Hodgson, 110 S.Ct. at 2950-51, Justice O’Connor found that no undue burden was imposed by a law requiring notice to both parents or judicial authorization before a minor could obtain an abortion. Justice O’Connor reached this conclusion despite statistics adduced by Justice Marshall to show that mandatory parental notice may inhibit a significant percentage of minors from obtaining abortions (id. at 2953-54) (Marshall, J., dissenting) and despite the district court’s finding, noted in Justice Marshall’s dissent, that the judicial bypass option “so daunted” some minors that they felt compelled to carry to term (id. at 2959, quoting 648 F.Supp. at 763).
Justice O’Connor has also suggested on more than one occasion that no undue burden was created by the statute upheld in H.L. v. Matheson, 450 U.S. 398, 101 S.Ct. 1164, 67 L.Ed.2d 388 (1981), which required parental notice prior to any abortion on an unemancipated minor. Instead, she has stated that this statute merely inhibited abortions to “some degree.” Thornburgh, 476 U.S. at 828, 106 S.Ct. at 2214 (O’Connor, J., dissenting); Akron, 462 U.S. at 464, 103 S.Ct. at 2510 (O’Connor, J., dissenting). In dissent in Matheson, Justice Marshall argued that the statute would result in substantial interference with abortions sought by minors. He wrote (450 U.S. at 398, 101 S.Ct. at 1164) (Marshall, J., dissenting) that “the minor may confront physical or emotional abuse, withdrawal of financial support or actual obstruction of the abortion decision.” These harms are almost identical to those that the majority in this case attributes to Section 3209. See majority opin. at 711-12. See also Planned Parenthood Association v. Ashcroft, 462 U.S. 476, 505, 103 S.Ct. 2517, 2532, 76 L.Ed.2d 733 (1983) (O’Connor concurring and dissenting) (statute requiring parental consent or judicial authorization “imposes no undue burden”).
Finally, Justice O’Connor has concluded that regulations that simply increase the cost of abortions, including regulations that may double the cost, do not create an “undue burden.” See Akron, 462 U.S. at 434-35, 103 S.Ct. at 2494-95 (maj. op.); at 466-67, 103 S.Ct. at 2511-12 (O’Connor, J., dissenting). Justice O’Connor reached this conclusion even though it seems clear that such increased costs may well deter some women.
Taken together, Justice O’Connor’s opinions reveal that an undue burden does not exist unless a law (a) prohibits abortion or gives another person the authority to veto an abortion or (b) has the practical effect of imposing “severe limitations,” rather than simply inhibiting abortions “ ‘to some degree’ ” or inhibiting “some women.” Thornburgh, 476 U.S. at 828, 829, 106 S.Ct. at 2213, 2214 (O’Connor, J., dissenting), quoting Akron, 462 U.S. at 464, 103 S.Ct. at 2510 (O’Connor, J., dissenting). Furthermore, Justice O’Connor’s opinions disclose that the practical effect of a law will not amount to an undue burden unless the effect is greater than the burden imposed on minors seeking abortions in Hodgson or Matheson or the burden created by the regulations in Akron that appreciably increased costs. Since the laws at issue in those cases had inhibiting effects that almost certainly were substantial enough to dissuade some women from obtaining abortions, it appears clear that an undue burden may not be established simply by showing that a law will have a heavy impact on a few women but that instead a broader inhibiting effect must be shown.
In this case, the plaintiffs, who made a facial attack1 on Section 3209, did not *722prove that this provision would impose an undue burden. Section 3209 does not create an “absolute obstacle” or give a husband “veto power.” Rather, this provision merely requires a married woman desiring an abortion to certify that she has notified her husband or to claim one of the statutory exceptions.
The plaintiffs also failed to carry their burden2 of proving that Section 3209 if enforced would have the kind of broad practical impact needed to establish an “undue burden” under the opinions discussed above. Clearly the plaintiffs did not substantiate the impact of Section 3209 with the degree of analytical rigor that should be demanded before striking down a state statute. Cf. Akron, 462 U.S. at 463, 103 S.Ct. at 2510 (O’Connor, J., dissenting) (citation omitted) (courts should exercise “ ‘deliberate restraint’ ” before finding an undue burden “ ‘in view of the respect that properly should be accorded legislative judgments’ ”); id. at 465, 103 S.Ct. at 2511.
At the outset, it is apparent that two factors imposed a low ceiling on any showing that the plaintiffs could have made. First, as the district court found, the “vast majority” of married women voluntarily inform their husbands before seeking an abortion. Planned Parenthood v. Casey, 744 F.Supp. 1323, 1360 (E.D.Pa.1990). Indeed, in the trial testimony on which the district court relied, the plaintiffs’ witness stated that in her experience 95% of married women notify their husbands. App. at 701. Second, the overwhelming majority of abortions are sought by unmarried women.3 Thus, it is immediately apparent that Section 3209 cannot affect more than about 5% of married women seeking abortions or an even smaller percentage of all women desiring abortions.
The plaintiffs failed to show even roughly how many of the women in this small group would actually be adversely affected by Section 3209. As previously noted, Section 3209 contains four significant exceptions. These exceptions apply if a woman certifies that she has not notified her husband because she believes4 that (1) he is not the father of the child, (2) he cannot be found after diligent effort, (3) the pregnancy is the result of a spousal sexual assault that has been reported to the authorities, or (4) she has reason to believe that notification is likely to result in the infliction of bodily injury upon her. If Section 3209 were allowed to take effect, it seems safe to assume that some percentage of the married women seeking abortions without notifying their husbands would qualify for and invoke these exceptions. The record, however, is devoid of evidence showing how many women could or could not invoke an exception.
Of the potentially affected women who could not invoke an exception, it seems safe to assume that some percentage, despite an initial inclination not to tell their husbands, would notify their husbands without suf*723fering substantial ill effects. Again, however, the record lacks evidence showing how many women would or would not fall into this category. Thus, the plaintiffs did not even roughly substantiate how many women might be inhibited from obtaining an abortion or otherwise harmed by Section 3209.5 At best, the record shows that Section 3209 would inhibit abortions “ ‘to some degree’ ” or that “some women [would] be less likely to choose to have an abortion by virtue of the presence” of Section 3209. Thornburgh, 476 U.S. at 828, 106 S.Ct. at 2214 (O’Connor, J., dissenting), quoting Akron, 462 U.S. at 464, 103 S.Ct. at 2510 (O’Connor, J., dissenting).6 And even with respect to these women, the plaintiffs did not show that the impact of Section 3209 would be any greater or any different from the impact of the notice requirement upheld in Matheson. Consequently, the plaintiffs failed to prove that Section 3209 would impose an undue burden.
Needless to say, the plight of any women, no matter how few, who may suffer physical abuse or other harm as a result of *724this provision is a matter of grave concern. It is apparent that the Pennsylvania legislature considered this problem and attempted to prevent Section 3209 from causing adverse effects by adopting the four exceptions noted above. Whether the legislature’s approach represents sound public policy is not a question for us to decide. Our task here is simply to decide whether Section 3209 meets constitutional standards. The first step in this analysis is to determine whether Section 3209 has been shown to create an undue burden under Supreme Court precedent, and for the reasons just explained it seems clear that an undue burden has not been established.
B. This conclusion is not undermined (and may indeed be supported) by the portion of Justice O’Connor’s opinion in Hodg-son regarding the constitutionality of the two-parent notice requirement without judicial bypass. The majority in this case interprets Justice O’Connor’s opinion to mean that this requirement imposed an undue burden and did not serve a “compelling” interest. Majority opin. at 696. I interpret Justice O’Connor’s opinion differently. I do not read her opinion to mean that the two-parent notice requirement without judicial bypass constituted an undue burden. Rather, I interpret her opinion to mean that this requirement was unconstitutional because it was not reasonably related to a legitimate state interest. Thus, I do not believe that her opinion (or the Court’s holding) supports the majority's conclusion in the present case that the spousal notification requirement in Section 3209 imposes an undue burden.
In Hodgson, Justice Stevens wrote the lead opinion discussing the unconstitutionality of the two-parent notification requirement without judicial bypass, and Justice O’Connor joined most of Justice Stevens’ opinion (see 110 S.Ct. at 2949 (O’Connor, J., concurring). Thus, in interpreting Justice O’Connor's position, it is helpful to begin with the relevant portions of Justice Stevens’ opinion.
Two portions of Justice Stevens’ opinion, Parts III and VII, are most important for present purposes. In Part III, Justice Stevens discussed the applicable constitutional standard. Nowhere in this portion of his opinion (or indeed in any portion of his opinion) did Justice Stevens make reference to “strict,” “exacting,” or “heightened” scrutiny or any of the terminology associated with that level of review. Instead, he concluded that the statute failed to satisfy even the least demanding standard of review. He wrote (110 S.Ct. at 2937): “Under any analysis, the ... statute cannot be sustained if the obstacles it imposes are not reasonably related to legitimate state interests.”
In Part VII of his opinion, Justice Stevens explained (id. at 2945) why the two-parent notice requirement did not “reasonably further any legitimate state interest.” Thus it seems clear that Justice Stevens’ opinion concluded that the two-parent notice requirement without judicial bypass was unconstitutional because it failed some variant of the rational relationship test.
In my view, Justice O’Connor’s opinion in Hodgson did not subject this requirement to a more exacting level of scrutiny. Although Justice O’Connor did not join Part III of Justice Stevens’ opinion (in which he discussed the general constitutional standard that he applied), Justice O’Connor wrote as follows (110 S.Ct. at 2949-50 (emphasis added)):
It has been my understanding in this area that “[i]f the particular regulation does not ‘unduly burde[n]’ the fundamental right, ... then our evaluation of that regulation is limited to our determination that the regulation rationally relates to a legitimate state purpose.” _ It is with that understanding that I agree with Justice Stevens’ statement “that the statute cannot be sustained if the obstacles it imposes are not reasonably related to legitimate state interests.”
I interpret this to mean that Justice O’Connor agreed with Justice Stevens that the challenged statute should be judged under the rational relationship test. I do not think that she would have expressed general agreement with Justice Stevens’ statement of the governing legal standard if she believed that the statute imposed an *725undue burden and was thus required to satisfy an entirely different legal standard. I also do not think that she would have concluded that the statute created an undue burden without explaining the basis for that conclusion. Moreover, Justice O’Con-nor joined Part VII of Justice Stevens’ opinion, in which, as previously noted, Justice Stevens concluded that the two-parent notice requirement without judicial bypass was not “reasonably” related to any “legitimate interest.” I do not think that Justice O’Connor would have joined this portion of Justice Stevens’ opinion if her position regarding the constitutionality of the provision was based on a fundamentally different analysis. Thus, I conclude that Justice O’Connor found the two-parent notice statute unconstitutional under the rational relationship test. This must mean either (a) that she did not believe that this requirement constituted an undue burden or (b) that she did not find it necessary to reach that question because she believed that the requirement could not even pass the rational relationship test. In either event, her position in no way undermines my conclusion that Section 3209 has not been shown to create an undue burden.7
III.
Since Section 3209 has not been proven to impose an undue burden, it must serve a “legitimate” (but not necessarily a “compelling”) state interest. The majority acknowledges that this provision serves a “legitimate” interest, namely, the state’s interest in furthering the husband’s interest in the fetus. See majority opin. at 715, 716. I agree with this conclusion, and I do not think that this point requires extended discussion.
The Supreme Court has held that a man has a fundamental interest in preserving his ability to father a child. Skinner v. Oklahoma, 316 U.S. 535, 541, 62 S.Ct. 1110, 1113, 86 L.Ed. 1655 (1942). The Court’s opinions also seem to establish that a husband who is willing to participate in raising a child has a fundamental interest in the child’s welfare. Michael H. v. Gerald D., 491 U.S. 110, 123, 109 S.Ct. 2333, 2342, 105 L.Ed.2d 91 (1989); Quilloin v. Walcott, 434 U.S. 246, 98 S.Ct. 549, 54 L.Ed.2d 511 (1978); Caban v. Mohammed, 441 U.S. 380, 99 S.Ct. 1760, 60 L.Ed.2d 297 (1969); Stanley v. Illinois, 405 U.S. 645, 92 S.Ct. 1208, 31 L.Ed.2d 551 (1972). It follows that a husband has a “legitimate” interest in the welfare of a fetus he has conceived with his wife.
To be sure, the Supreme Court held in Planned Parenthood of Missouri v. Danforth, 428 U.S. 52, 67-72, 96 S.Ct. 2831, 2840-43, 49 L.Ed.2d 788 (1976), that a potential father may not be given the legal authority to veto an abortion, and thus the Court apparently held that the potential father’s interest was not “compelling.” But the Court did not question the legitimacy of this interest. On the contrary, the Court wrote {id. at 69, 96 S.Ct. at 2841 (emphasis added)): “We are not unaware of the deep and proper concern and interest that a devoted and protective husband has in his wife’s pregnancy and in the growth and development of the fetus she is carrying.” See also id. at 93, 96 S.Ct. at 2852 (White, J., dissenting) (“A father’s interest in having a child — perhaps his only child— may be unmatched by any other interest in his life”). Since a “deep and proper ... interest” appears indistinguishable from a “legitimate” interest, it seems clear that a husband has a “legitimate” interest in the fate of the fetus.
This interest may be legitimately furthered by state legislation. “[Sjtatutory regulation of domestic relations [is] an area *726that has long been regarded as a virtually exclusive province of the States.” Sosna v. Iowa, 419 U.S. 393, 404, 95 S.Ct. 553, 560, 42 L.Ed.2d 532 (1975). See also Moore v. East Cleveland, 431 U.S. 494, 499, 97 S.Ct. 1932, 1935, 52 L.Ed.2d 531 (1977); Scheinberg v. Smith, 659 F.2d 476, 483-94 (5th Cir.1981). Accordingly, Pennsylvania has a legitimate interest in furthering the husband’s interest in the fate of the fetus, as the majority in this case acknowledges.
IV.
The remaining question is whether Section 3209 is “rationally” or “reasonably” related to this interest. Under the rational relationship test, which developed in equal protection cases, “legislation carries with it a presumption of rationality that can only be overcome by a clear showing of arbitrariness and irrationality.” Hodel v. Indiana, 452 U.S. 314, 331-32, 101 S.Ct. 2376, 2386-87, 69 L.Ed.2d 40 (1981). This test does not permit the invalidation of legislation simply because it is “deemed unwise or unartfully drawn.” U.S. Railroad Retirement Board v. Fritz, 449 U.S. 166, 175, 101 S.Ct. 453, 459, 66 L.Ed.2d 368 (1981). Legislation does not violate this test simply because it produces some adverse effects. Id.; Lindsley v. Natural Carbonic Gas Co., 220 U.S. 61, 78, 31 S.Ct. 337, 340, 55 L.Ed. 369 (1911). As the Court wrote in Dandridge v. Williams, 397 U.S. 471, 485-86, 90 S.Ct. 1153, 1161-62, 25 L.Ed.2d 491 (1970):
“The problems of government are practical ones and may justify, if they do not require, rough accommodations — illogical, it may be, and unscientific.” Metropolis Theatre Co. v. City of Chicago, 228 U.S. 61, 68-70 [33 S.Ct. 441, 443, 57 L.Ed. 730 (1913)]....
... [The rational-basis standard] is true to the principle that the Fourteenth Amendment gives the federal courts no power to impose upon the States their views of what constitutes wise economic or social policy.
See also Dallas v. Stanglin, 490 U.S. 19, 25-27, 109 S.Ct. 1591, 1595-96, 104 L.Ed.2d 18 (1989); Cleburne v. Cleburne Living Center, Inc., 473 U.S. 432, 439-40, 105 S.Ct. 3249, 3253-54, 87 L.Ed.2d 313 (1985). Rather, “those challenging the legislative judgment must convince the Court that the legislative facts on which the classification is apparently based could not reasonably be conceived to be true by the governmental decisionmaker.” Vance v. Bradley, 440 U.S. 93, 111, 99 S.Ct. 939, 949, 59 L.Ed.2d 171 (1979). See also Hancock Industries v. Schaeffer, 811 F.2d 225, 238 (3d Cir. 1987).
Even assuming that the rational relationship test is more demanding in the present context than in most equal protection cases, that test is satisfied here. The Pennsylvania legislature could have rationally believed that some married women are initially inclined to obtain an abortion without their husbands’ knowledge because of perceived problems — such as economic constraints, future plans, or the husbands’ previously expressed opposition — that may be obviated by discussion prior to the abortion. In addition, the legislature could have reasonably concluded that Section 3209 would lead to such discussion and thereby properly further a husband's interests in the fetus in a sufficient percentage of the affected cases to justify enactment of this measure. Although the plaintiffs and supporting amici argue that Section 3209 will do little if any good and will produce appreciable adverse effects, the Pennsylvania legislature presumably decided that the law on balance would be beneficial. We have no authority to overrule that legislative judgment even if we deem it “unwise” or worse. U.S. Railroad Retirement Board v. Fritz, 449 U.S. at 175, 101 S.Ct. at 459. “We should not forget that ‘legislatures are ultimate guardians of the liberty and welfare of the people in quite as great a degree as the courts.’ ” Akron v. Akron Center For Reproductive Health, 462 U.S. at 465, 103 S.Ct. at 2511 (O’Connor, J., dissenting), quoting Missouri, K. & T.R. Co. v. May, 194 U.S. 267, 270, 24 S.Ct. 638, 639, 48 L.Ed. 971 (1904). Clearly, the plaintiffs have not shown that “the legislative facts on which [the statute] is apparently based could not reasonably be conceived to be true by the governmental deci-*727sionmaker.” Vance v. Bradley, 440 U.S. at 111, 99 S.Ct. at 949. Thus, Section 3209 is rationally related to a legitimate state interest and may not be invalidated under the Supreme Court’s abortion precedents.8
. Because the plaintiffs made a facial attack on Section 3209, they could not rely on a "worst-case analysis” (Ohio v. Akron Center for Reproductive Health, — U.S. —, 110 S.Ct. 2972, 2981, 111 L.Ed.2d 405 (1990)) or on proof showing only that the provision would impose an undue burden “under some conceivable set of circumstances” (United States v. Salerno, 481 *722U.S. 739, 745, 107 S.Ct. 2095, 2100, 95 L.Ed.2d 697 (1987)). Thus, proof that the provision would adversely affect an unknown number of women with a particular combination of characteristics could not suffice.
.In Thornburgh, Justice O’Connor made clear that a party challenging the constitutionality of a statute must bear the burden of proving that the law imposes an undue burden. After arguing strenuously that the case should be sent back to the district court for "additional factual development” (476 U.S. at 827, 106 S.Ct. at 2213 (O’Connor, J., dissenting), Justice O'Connor repeatedly stated that the appellees, who were challenging the statute, had the burden of proving that individual statutory provisions would impose an undue burden. She discussed whether "appellees could succeed in making the threshold showing of undue burden” {id. at 831, 106 S.Ct. at 2215), whether "appellees [could] establish that the abortion decision [would be] unduly burdened” (id.), and whether the appel-lees "could succeed in establishing an undue burden” (id. at 832, 106 S.Ct. at 2216).
. Since 1973, abortions on unmarried women have consistently exceeded 70% of the national total and at times have surpassed 80%. United States Department of Commerce, Statistical Abstract of the United States 1990 at 71.
. The form prepared by the Pennsylvania Department of Health for use in implementing Section 3209 requires a woman to certify that she has not notified her husband "for the following reason(s)....” (744 F.Supp. at 1359). Moreover, a false statement is punishable (as a third degree misdemeanor) only if the woman did not “believe [the statement] to be true” (18 Pa.Cons.Stat.Ann. § 4904(b) (1983)).
. In considering whether Section 3209 would impose an undue burden, I do not take into account a fact that seems glaringly apparent, i.e., that Section 3209 would be difficult to enforce and easy to evade. Section 3209 does not require a woman to provide any proof of notification other than her own unnotarized statement. Thus, if a woman claimed that she had orally notified her husband in private (the mode and place of notification to be expected in most cases), it would be exceedingly difficult in most cases for the Commonwealth to prove beyond a reasonable doubt that she had not done so.
Proving that a woman violated the law due to a false statement concerning one of the exceptions would also be hard. As noted (see footnote 4, supra ), the Commonwealth would have to prove that the woman did not "believe [the statement] to be true" (18 Pa.Cons.Stat.Ann. § 4904(b) (1983)). Consequently, if a woman certified that she did not notify her husband because he was not the father, the Commonwealth would have to prove that she subjectively believed that the husband was the father. Or, if a woman certified that she did not notify her husband because she had reason to believe that this would lead to the infliction of bodily injury upon her, the Commonwealth would have to prove that the woman subjectively believed that she would not be harmed. It seems likely, therefore, that Section 3209, if allowed to take effect, would be widely evaded and infrequently enforced and would consequently be less likely to produce either the good or bad effects that the opposing parties claim.
. The plaintiffs’ proof may be separated into five categories. First, they offered testimony that a spousal notification requirement would sometimes delay an abortion or necessitate an extra trip to the abortion provider (see 744 F.Supp. at 1360). But as the majority properly concludes in rejecting identical objections to the 24-hour waiting period required by Section 3205(a) (see majority opin. at 706-07), these potential effects do not amount to an undue burden. See Akron, 462 U.S. at 472-74, 103 S.Ct. at 2515-16 (O’Connor, J., dissenting).
Second, the plaintiffs offered testimony that the exceptions in Section 3209 would not cover a case in which a woman did not want to notify her husband for fear that he would retaliate in some way other than the infliction of bodily injury upon her, such as by subjecting her to psychological abuse or abusing their children (see 744 F.Supp. at 1360-62). The plaintiffs, however, do not appear to have offered any evidence showing how many (or indeed that any actual women) would be affected by this asserted imperfection in the statute.
Third, the plaintiffs introduced general evidence about the problem of spouse abuse (see 744 F.Supp. at 1361). They offered widely varying statistics concerning the dimensions of the problem, as well as evidence that battering occurs in all socioeconomic groups and is sometimes fatal. This proof, while documenting the existence of a broad national problem, provides no basis for any estimate of what is relevant here — the impact of Section 3209.
Fourth, the plaintiffs offered evidence that "mere notification of pregnancy is frequently a flashpoint for battering" (see 744 F.Supp. at 1361). This proof indicates when violence is likely to occur in an abusive marriage but provides no basis for determining how many women would be adversely affected by Section 3209. Finally, the plaintiffs offered the opinion of one of their witnesses that most battered women would be psychologically incapable of taking advantage of Section 3209’s fourth exception, i.e., the exception for cases in which the woman has reason to fear that notification will lead to the infliction of bodily harm upon her (see 744 F.Supp. at 1363). However, the plaintiffs failed to show how many of the women potentially affected by Section 3209 (married women seeking abortions without notifying their husbands) are victims of battering. Thus, the opinion offered by their expert, even if taken at face value, merely describes the likely behavior of most of the women in a group of unknown size. Clearly, then, this evidence does not show how many women would be inhibited or otherwise harmed by Section 3209. I cannot believe that a state statute may be held facially unconstitutional simply because one expert testifies that in her opinion the provision would harm a completely unknown number of women.
. In the portion of her opinion concluding that the two-parent notification requirement with judicial bypass was constitutional, Justice O’Con-nor wrote (110 S.Ct. at 2950 (emphasis added)): "In a series of cases, this Court has explicitly approved judicial bypass as a means of tailoring a parental consent provision so as to avoid unduly burdening the minor’s limited right to obtain an abortion.” I interpret this statement to mean that a judicial bypass option prevents a consent requirement (which would otherwise amount to an absolute veto) from creating an undue burden. This statement is therefore fully consistent with my view that Justice O’Connor did not find that an undue burden was created by the two-parent notice requirement without judicial bypass.
. The plaintiffs argue that the district court’s decision may be affirmed on alternative constitutional grounds not adopted by that court, i.e., that Section 3209 violates the rights to marital and informational privacy and equal protection. Because the majority has relied solely on the abortion right in affirming the district court, I do not address these alternative grounds.