Full opinion text
FINDINGS OF FACT AND CONCLUSIONS OF LAW HAMILTON, District Judge. An Indiana law enacted in 1995 requires in almost all cases that at least 18 hours before an abortion can be performed, a woman must be given certain state-mandated information concerning the abortion and alternatives to abortion. See Ind. Code § 16-34-2-1.1. The law specifically requires that medical personnel provide some of this advance information “in the presence” of the pregnant woman. The “in the presence” provision would require most women to make two trips to a clinic in order to obtain an abortion. Plaintiffs in this case are reproductive health care facilities that provide a range of services related to pregnancy and women’s health, including abortions up to 12 weeks of gestation, and a licensed physician who performs abortions. Plaintiffs contend the “in the presence” requirement is unconstitutional because it imposes an undue burden on a woman’s constitutional right to choose to end a pregnancy. The Indiana statute is similar to a Pennsylvania law upheld by the Supreme Courj; against a facial challenge in Planned Parenthood of Southeastern Pennsylvania v. Casey, 505 U.S. 833, 112 S.Ct. 2791, 120 L.Ed.2d 674 (1992), as well as a Wisconsin law upheld by the Seventh Circuit in Karlin v. Foust, 188 F.3d 446 (7th Cir.1999). Both decisions left open the possibility, however, that additional evidence on the effects of such laws could establish an undue burden. In this case, plaintiffs have presented evidence on the effects of such “in the presence” requirements that was not presented to the courts in Casey or Karlin. The additional evidence shows that Indiana’s “in the presence” requirement is likely to impose an undue burden on the ability of many women to exercise their constitutional right to choose to end a pregnancy. The evidence shows that the burden imposed by the “in the presence” requirement is likely to prevent abortions for approximately 10 to 13 percent of Indiana women who would otherwise choose to have an abortion — roughly 1300 to 1700 per year. Plaintiffs have also shown it is highly unlikely that the effects of the law will result from any persuasive effect the state-mandated information might have. There is no evidence from other states or from Indiana showing that requiring such state-mandated information to be provided to a woman in advance of an abortion (whether in person or otherwise) actually persuades women to choose childbirth over abortion. Accordingly, as explained below, the “in the presence” provision imposes an undue burden on a woman’s constitutional right to choose to end a pregnancy. The court is entering a permanent injunction against enforcement of the “in the presence” requirement, which is severable from the other provisions of Public Law 187. This entry sets forth the court’s findings of fact and conclusions of law pursuant to Fed. R.Civ.P. 52. The substance rather than the court’s label shall determine whether a matter is a finding of fact or a conclusion of law. I. The Indiana Statute Indiana has long required physicians performing abortions to obtain the informed consent of their patients, just as they must obtain informed consent for other medical procedures. Informed consent generally requires that the patient be told the general nature of her condition, the proposed treatment or procedure, the expected outcome, the material risks, and the reasonable alternatives to the treatment or procedure. See Ind.Code § 34-18-12-3 (informed consent for purposes of medical malpractice action). Indiana’s “informed consent” requirements for abortions reach well beyond the more general requirements for medical procedures. Abortions in Indiana are criminal unless a number of conditions are satisfied, one of which is that “the woman submitting to the abortion has filed her consent with her physician.” Ind.Code § 16 — 34—2—1 (1)(B). Indiana Public Law 187-1995 (referred to here as “Public Law 187”) added special mandatory disclosure and waiting period provisions for informed consent for abortions. The law requires in almost all cases that certain medical information and information about alternatives to abortion be provided to a woman orally at least 18 hours before she may have an abortion. Some of the medical information must be provided “in the presence of the pregnant woman.” The law was drafted to have gone into effect on September 1, 1995, but it was enjoined from operation first by this court’s temporary restraining order and then by this court’s preliminary injunction. A Woman’s Choice—East Side Women’s Clinic v. Newman, 904 F.Supp. 1434 (S.D.Ind.1995). The central provisions of Public Law 187 state: An abortion shall not be performed except with the voluntary and informed consent of the pregnant woman upon whom the abortion is to be performed. Except in the case of a medical emergency, consent to an abortion is voluntary and informed only if the following conditions are met: (1) At least eighteen (18) hours before the abortion and in the presence of the pregnant woman, the physician who is to perform the abortion, the referring physician or a physician assistant (as defined in IC 25-27.5-2-10), an advanced practice nurse (as defined in IC 25-23-1 — 1(b)), or a midwife (as defined in IC 27-12-2-19) to whom the responsibility has been delegated by the physician who is to perform the abortion or the referring physician has orally informed the pregnant woman of the following: (A) The name of the physician performing the abortion. (B) The nature of the proposed procedure or treatment. (C) The risks of and alternatives to the procedure or treatment. (D) The probable gestational age of the fetus, including an offer to provide: (i) a picture or drawing of a fetus; (ii) the dimensions of a fetus; and (iii) relevant information on the potential survival of an unborn fetus; at this stage of development. (E) The medical risks associated with carrying the fetus to term. (2) At least eighteen (18) hours before the abortion, the pregnant woman will be orally informed of the following: (A) That medical assistance benefits may be available for prenatal care, childbirth, and neonatal care from the county office of family and children. (B) That the father of the unborn fetus is legally required to assist in the support of the child. In the case of rape, the information required under this clause may be omitted. (C) That adoption alternatives are available and that adoptive parents may legally pay the costs of prenatal care, childbirth, and neonatal care. (3) The pregnant woman certifies in writing, before the abortion is performed, that the information required by subdivisions (1) and (2) has been provided. Ind.Code § 16-34-2-1.1. Anyone who knowingly or intentionally performs an abortion in violation of these requirements is subject to criminal penalties. Ind.Code §§ 16-34-2-1, 16-34-2-7. The evidence here shows that only a few women are referred to abortion providers by other physicians. See Ex. 25 (approximately four percent of patients in one clinic were referred by outside physician). Thus, the practical effect of the “in the presence” requirement is to require a woman to make two trips to an abortion clinic in order to obtain an abortion. II. Procedural Background Plaintiffs filed their complaint on August 24, 1995, asserting that Public Law 187 would impose undue burdens on a woman’s constitutional right to choose to end a pregnancy. Defendants are a class consisting of all prosecuting attorneys in the State of Indiana, with Scott C. Newman of Marion County as representative of the class; and the Commissioner of the Indiana Department of Health. On August 25, 1995, plaintiffs filed a verified motion for a preliminary injunction and expedited hearing, asserting that enforcement of Public Law 187 would cause immediate and irreparable harm. After a hearing on August 30, 1995, the court entered a temporary restraining order enjoining enforcement of the challenged law for ten days. The court also issued an order certifying the defendant class. The parties agreed to a short extension of the temporary restraining order, and the court heard evidence on plaintiffs’ motion for a preliminary injunction on October 11-13, 1995. Also on October 13, 1995, defendants filed a motion to certify the interpretation of the medical emergency exception to the Supreme Court of ' Indiana. On November 9, 1995, the court issued a preliminary injunction enjoining enforcement of Public Law 187. See A Woman’s Choice, 904 F.Supp. 1434. The court also granted defendants’ motion to certify the interpretation of the medical emergency exception to the Supreme Court of Indiana. After the state court answered certified questions about the scope of the medical emergency exception, defendants moved to vacate or modify this court’s preliminary injunction. On October 14, 1997, this court modified its preliminary injunction. See A Woman’s Choice—East Side Women’s Clinic v. Newman, 980 F.Supp. 962 (S.D.Ind.1997). The state court’s construction of the medical emergency exception persuaded this court that the law’s enforcement should not be enjoined entirely. However, the court found that the grounds for the preliminary injunction against enforcement of the “in the presence” requirement in Public Law 187 remained valid and that the requirement was severable from the remaining portions of Public Law 187. The court therefore modified its preliminary injunction to permit enforcement of the waiting period and mandatory disclosure provisions of Public Law 187, concluding that, on the record before the court, plaintiffs were not likely to succeed in showing that the mandatory disclosure and waiting period requirements would impose burdens that would actually prevent a substantial number of women from having abortions they would otherwise choose to have. The court also concluded that plaintiffs were still likely to succeed on their challenge to the “in the presence” requirement. The court therefore also modified its injunction against enforcement of that requirement by directing that the state-mandated information could be provided by telephone. Plaintiffs now challenge the constitutionality of only the “in the presence” requirement of Public Law 187. The record at this point includes evidence from the earlier stages of this case, a court trial on November 1-2, 1999, and later written supplements to the evidence by both sides. After post-trial briefing and further evi-dentiary submissions in writing, the court heard additional argument. The issues have been fully briefed and are ripe for decision. III. Legal Framework Governing Abortion Regulations A. The “Undue Burden” Standard “The issue of abortion is one of the most contentious and controversial in contemporary American society. It presents extraordinarily difficult questions that, as the Court recognizes, involve ‘virtually irreconcilable points of view.’ ” Stenberg v. Carhart, 530 U.S. 914, 120 S.Ct. 2597, 2617-18, 147 L.Ed.2d 743 (2000) (O’Connor, J., concurring). Notwithstanding the contention and controversy, some broad principles have been established by the Supreme Court. First, before viability of the fetus, a woman has a right under the United States Constitution to choose to terminate her pregnancy. Stenberg, 120 S.Ct. at 2604, citing Planned Parenthood of Southeastern Pennsylvania v. Casey, 505 U.S. 833, 870, 112 S.Ct. 2791, 120 L.Ed.2d 674 (1992) (controlling joint opinion of O’Con-nor, Kennedy, and Souter, JJ.). Second, “a law designed to further the State’s interest in fetal life which imposes an undue burden on the woman’s decision before fetal viability” is unconstitutional. Stenberg, 120 S.Ct. at 2604, quoting Casey, 505 U.S. at 877, 112 S.Ct. 2791. An “undue burden is ... shorthand for the conclusion that a state regulation has the purpose or effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus.” Id. The standard has two distinct and independent prongs, based on the purpose of the law and the effect of the law. Either may be sufficient to find an undue burden. Casey and Stenberg together establish the' “undue burden” standard as the controlling test for evaluating state regulations of abortions. The unadorned language of the undue burden test — determining whether a state law has “the purpose or effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus” — does not draw sharp lines on the legal map. This court’s task in this case is to apply the undue burden test to the evidence in this case. For purposes of this case, the best guide to the meaning of the Supreme Court's language is the Court’s application of the test in Casey itself. At issue in Casey were Pennsylvania statutes that required disclosures of state-mandated information to the woman before she could give informed consent; required a woman to wait 24 hours after such disclosures before obtaining an abortion; required a minor to have parental consent prior to obtaining an abortion; and required a married woman to notify her husband before obtaining an abortion. The controlling joint opinion by Justices O’Connor, Kennedy, and Souter in Casey articulated the undue burden standard and then applied it to those statutory requirements. A majority of the Court held that the requirements of mandatory disclosure, a 24-hour waiting period, and parental consent for minors had not been shown to impose an undue burden on women seeking pre-viability abortions. See 505 U.S. at 885-87, 899-900, 112 S.Ct. 2791 (O’Connor, Kennedy, Souter, JJ.), id. at 967-71, 112 S.Ct. 2791 (Rehnquist, C.J., joined by White, Scalia, Thomas, JJ.). A different majority of the Court struck down the requirement that a married woman notify her spouse before seeking an abortion because the requirement was likely to prevent a significant number of women from obtaining an abortion. See 505 U.S. at 893, 112 S.Ct. 2791 (O’Connor, Kennedy, Souter, JJ., joined by Blackmun and Stevens, JJ.). Pennsylvania’s spousal notification requirement was the only provision for which the Court found sufficient evidence of an undue burden on a woman’s right to choose to terminate a pregnancy. The portion of the joint opinion striking down that requirement therefore provides the surest guide to the type of showing required to meet the undue burden standard in a facial challenge. See Karlin v. Foust, 188 F.3d at 480 (acknowledging instructive significance of this portion of Casey); see also Casey, 505 U.S. at 920 n. 6, 112 S.Ct. 2791 (Stevens, J.) (“The meaning of any legal standard can only be understood by reviewing the actual cases in which it is applied.”). When the Casey Court struck down the spousal notification law, the record did not include statistical evidence of the effects of any similar spousal notification requirements in other states. The record included substantial evidence and findings of fact on the frequency of domestic violence and the role that pregnancy often plays as a flashpoint for such violence against women and their children. See 505 U.S. at 887-92, 112 S.Ct. 2791. The Court also considered “limited research” on spousal notification requirements that involved “samples too small to be representative.” Id. at 892, 112 S.Ct. 2791. That research also supported findings of fact indicating that, when married women do not notify their husbands of their decisions to have an abortion, in most cases the pregnancy is the result of an extramarital affair or, if the husband is the father, the husband and wife are experiencing marital difficulties, often accompanied by violence. Id. The Court concluded that the research and the district court findings “reinforce what common sense would suggest.” Id. at 892-93, 112 S.Ct. 2791. Although spouses in healthy marriages discuss such important and intimate decisions, “there are millions of women in this country who are the victims of regular physical and psychological abuse at the hands of their husbands. Should these women become pregnant, they may have very good reasons for not wishing to inform their husbands of their decision to obtain an abortion.” Id. at 893, 112 S.Ct. 2791. The fact that a law may make abortions more difficult or expensive to obtain does not, the joint opinion in Casey emphasized, necessarily show that the law imposes a substantial obstacle. Id. at 893-94, 112 S.Ct. 2791. What made the difference for the Court was the evidence supporting the following conclusion about the spousal notification law: “We must not blind ourselves to the fact that the significant number of women who fear for their safety and the safety of their children are likely to be deterred from procuring an abortion as surely as if the Commonwealth had outlawed abortion in all cases.” Id. at 894, 112 S.Ct. 2791. The defendants in Casey argued that the spousal notice requirement could not be an “undue burden” because it would affect only about one percent of women seeking abortions. (About 20 percent of women seeking abortions were married, the evidence showed, and 95 percent of those women notified their husbands on their own.) The Court rejected the argument: “The analysis does not end with the one percent of women upon whom the statute operates; it begins there. Legislation is measured for consistency with the Constitution by its impact on those whose conduct it affects.” 505 U.S. at 894, 112 S.Ct. 2791. The Court recognized the target of the legislation as married women seeking abortions who did not wish to notify their husbands of their intentions and who did not qualify for one of the exceptions to the requirement. The Court found that the prospect of domestic violence would mean that, “in a large fraction of the cases in which [the requirement] is relevant, it will operate as a substantial obstacle to a woman’s choice to undergo an abortion. It is an undue burden, and therefore invalid.” Id. at 895, 112 S.Ct. 2791. The Court did not quantify more precisely the “large fraction” or its denominator. The portion of the joint opinion in Casey addressing the mandatory disclosure and waiting period requirements in the Pennsylvania law is also highly instructive here, of course, especially since the Indiana law was modeled on the Pennsylvania law upheld in Casey. The joint opinion recognized that the law would impose some burden on women seeking abortions in terms of increased cost and delay, but the Court was not persuaded that those burdens were so great as to be “undue burdens.” See id. at 886-87. The Court acknowledged the district court’s findings that delays would require two trips to a clinic, thus increasing exposure of women to “the harassment and hostility” of protesters at clinics, and that “for those women who have the fewest financial resources, those who must travel long distances, and those who have difficulty explaining their whereabouts to husbands, employers, or others, the 24-hour waiting period will be ‘particularly burdensome.’ ” 505 U.S. at 886, 112 S.Ct. 2791, quoting 744 F.Supp. 1323, 1352 (E.D.Pa.1990). The Court described these findings as “troubling in some respects,” but inadequate to demonstrate that the waiting period imposed an “undue burden.” 505 U.S. at 886, 112 S.Ct. 2791. The Court concluded: “Hence, on the record before us, and in the context of this facial challenge, we are not convinced that the 24-hour waiting period constitutes an undue burden.” Id. at 887, 112 S.Ct. 2791. That conclusion in Casey left the door open for possible future challenges, such as the one here, to similar state laws modeled after the Pennsylvania statute if parties challenging such laws could come forward with evidence that would meet the “undue burden” standard. See Planned Parenthood of Southeastern Pennsylvania v. Casey, 510 U.S. 1309, 1313, 114 S.Ct. 909, 127 L.Ed.2d 352 (1994) (Souter, Circuit Justice) (denying application for a stay of Court of Appeals’ refusal on remand to consider additional evidence on the likely effects of Pennsylvania’s statute, but noting that “litigants are free to challenge similar restrictions in other jurisdictions, as well as these very provisions as applied”); Karlin v. Foust, 188 F.3d at 484 (“We conclude plaintiffs are not precluded from challenging a waiting period provision nearly identical in all respects to the one upheld in Casey.”). The parties here have recognized, as the Supreme Court and the Seventh Circuit have taught, that application of the effects prong of the undue burden standard to similar laws could lead to different results in different states. Such different results may occur either because new evidence is presented regarding the actual effects of such laws, or perhaps because of demographic or geographic factors unique to a particular state. Thus, the Supreme Court’s decision in Casey did not control the validity of a similar law in Wisconsin where different evidence was presented. See Karlin, 188 F.3d at 485 (recognizing that Casey’s decision to uphold Pennsylvania law did not control constitutionality of Wisconsin law). Nor should the Wisconsin case control the validity of Indiana’s law in a case where additional and different evidence has been presented. B. Burden of Proof for a Facial Challenge Plaintiffs have brought a facial challenge to the “in the presence” requirement before it could take effect. In bringing a facial challenge, “a party seeks to vindicate not only [her] own rights, but those of others who may also be adversely impacted by the statute in question.” Chicago v. Morales, 527 U.S. 41, 55 n. 22, 119 S.Ct. 1849, 144 L.Ed.2d 67 (1999). “When a facial challenge is successful, the law in question is declared to be unenforceable in all its applications, and not just in its particular application to the party in suit.” Id. at 74, 119 S.Ct. 1849 (Scalia, J. dissenting) (emphasis in original). In a facial challenge, those challenging an abortion regulation do not have to wait for the law to take effect and cause them harm. Since Casey was decided in 1992, the lower federal courts have debated whether facial challenges to laws regulating abortions should be governed by the “undue burden” standard as the Court actually applied it in Casey or by the so-called “Salerno standard,” which requires a demonstration “that no set of circumstances exists under which the [statute] would be valid.” See United States v. Salerno, 481 U.S. 739, 745, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987) (upholding federal Bail Reform Act against facial challenge under Eighth Amendment). The Court in Casey did not expressly overrule or reject the Salerno standard, but the Court did not apply the standard in evaluating the facial challenge to Pennsylvania’s abortion regulation. The joint opinion in Casey found that a woman’s right to choose to terminate or continue her pregnancy must be protected from “undue burdens” imposed by a state regulation. Casey, 505 U.S. at 876, 112 S.Ct. 2791. The joint opinion further held that a regulation imposes an undue burden if it operates as a substantial obstacle to “a large fraction” of the women “for whom the law is a restriction.” Id. at 895, 112 S.Ct. 2791. Applying this standard, the Court struck down the statute’s spousal notification requirement, finding that the provision was likely to pose a substantial obstacle to a significant fraction of the women affected by the restriction. Id. at 893-97, 112 S.Ct. 2791. The apparent tension between Casey and the Salerno standard for facial challenges has resulted in a circuit split regarding the proper standard for facial challenges to abortion regulations. See Hope Clinic v. Ryan, 195 F.3d 857, 865 (7th Cir.1999) (,en bane ) (“Courts of appeals are divided on the question whether Salerno applies to abortion legislation.”), vacated, 530 U.S. 1271, 120 S.Ct. 2739, 147 L.Ed.2d 1002 (2000); Carhart v. Stenberg, 192 F.3d 1142, 1149 (8th Cir.1999) (“In considering a challenge to the facial validity of an abortion regulation, we follow the standard set out in Casey.”), aff'd, 530 U.S. 914, 120 S.Ct. 2597, 147 L.Ed.2d 743. In ruling on plaintiffs’ motion for a preliminary injunction, this court concluded that “Casey effectively displaced Salerno’s, application to abortion laws.” A Woman’s Choice, 904 F.Supp. at 1448, citing Planned Parenthood, Sioux Falls Clinic v. Miller, 63 F.3d 1452, 1458 (8th Cir.1995) (“We choose to follow what the Supreme Court actually did — rather than what it failed to say — and apply the undue-burden test.”). This court adheres to that view, follows what the Supreme Court has actually done in Casey and Stenberg, and applies the undue burden standard as applied in those cases. IV. The Effects Prong of the Undue Burden Test In applying the effects prong of the undue burden test, the question is whether the “in the presence” requirement of Public Law 187 is likely to impose an “undue burden” on Indiana women who seek to have abortions. In terms of Casey, would the law operate to place a “substantial obstacle” in the path of “a large fraction” of the women for whom the law operates as a restriction? See 505 U.S. at 895, 112 S.Ct. 2791. The best available guide to answer this question in this facial challenge comes from the post-Case?/ experience in other states that have implemented “two-trip” requirements, as well as from the experience in Indiana and other states that have implemented waiting period and mandatory disclosure requirements without requiring two trips. Mississippi has been studied most closely, but evidence in this record also pertains to Utah, Louisiana, North Dakota, and Indiana itself. In applying the undue burden test, it is important to distinguish between two factual questions. The first is whether “two-trip” laws have a significant effect on abortion rates and practices. That issue is the subject of extensive statistical evidence and debate. As explained below, the answer is yes. The court finds as a fact that enforcement of the “in the presence” requirement in Indiana, which effectively requires two trips to an abortion clinic, (a) is likely to cause a significant decline of approximately 10 to 13 percent in the rate for Indiana women having abortions, which amounts to approximately 1300 to 1700 per year; (b) is likely to cause significant increases in travel out-of-state to have abortions; and (c) is likely to cause a significant increase in the proportion of second trimester abortions, which are both riskier and more expensive than earlier abortions. Proof that the law is likely to cause a decrease in the abortion rate is not sufficient by itself, however, to show that the “in the presence” requirement would impose an undue burden on women. Putting aside for the moment the increase in second trimester abortions and the changes in interstate travel for abortions, if the effect of the law, with its combination of state-mandated information and the required waiting period, were to persuade women to change their minds about whether to have an abortion, that effect would be constitutional under Casey. Casey upheld the Pennsylvania waiting period and mandatory disclosure law precisely because the state was entitled to take some measures to try to persuade women not to have abortions. See 505 U.S. at 883, 112 S.Ct. 2793. The second factual issue, therefore, is whether the expected reduction in abortion rates would result from any persuasive effect the law might have, or whether the reduction would result instead from the law posing substantial obstacles for women seeking abortions. That question cannot be answered by examining only the abortion rate statistics in states requiring two trips to clinics. However, states that require mandatory disclosures and waiting periods without requiring two trips have not shown any significant reduction in abortion rates resulting from those requirements. That evidence indicates that such laws have no detectable persuasive effects. There is also a complete lack of any other evidence tending to show that mandatory disclosures and waiting periods have any persuasive effect. In addition, the evidence shows that a two-trip law is likely to increase the number of women leaving the enacting state to obtain abortions in other states and to decrease the number of women from other states coming to the enacting state to have abortions, and to increase the frequency of later abortions. The sum of this evidence, and the absence of evidence of any persuasive effect, shows convincingly that the predicted reduction in abortion rates would result not from persuasion but from restrictions posing a substantial obstacle for some women’s ability to obtain abortions. Accordingly, the court further finds as a fact that the likely effects of the “in the presence” requirement would not be the result of any persuasive effect of the state-mandated information or the required waiting period. The likely effects would instead result from the burdens that the “in the presence” requirement would impose on women. The court explains next the evidence supporting these findings on the reduction in rates and other effects of the law, as well as the reasons for the reduction. A. Effects of “In the Presence” Requirements on Abortion Rates 1. The Evolution of the Data The parties have submitted extensive evidence relevant to the effects prong of the undue burden test. In a sense, the litigation process has produced an ongoing dialogue among expert statisticians and judges in this case and in Karlin v. Foust. From the outset of this case, plaintiffs have relied primarily on Dr. Stanley Hen-shaw, a statistician with the Alan Gutt-macher Institute. After this court’s preliminary injunction decision in 1995, Dr. Henshaw also testified in the case challenging the similar Wisconsin law in Kar-lin. The State of Wisconsin in Karlin relied on Dr. Peter Uhlenberg, a professor of sociology at the University of North Carolina, to criticize and rebut Dr. Henshaw’s work. Judge Crabb reviewed the work of both experts in her decision. See Karlin, 975 F.Supp. at 1215-18. Judge Crabb relied on Dr. Henshaw’s data and analysis to find that the Mississippi two-trip that took effect in 1992 law caused a statistically significant reduction in the rate of abortions for Mississippi residents. Id. at 1217. However, Judge Crabb was not convinced by the evidence offered in that case that the reduction was caused by burdens the law imposed rather than by a possible persuasive effect. Id. at 1217-18. The Seventh Circuit did not disturb either finding. See 188 F.3d at 487. Thus, one way of looking at the present case is that it turns first on whether Indiana and its experts have undermined the basis for expecting the “in the presence” requirement to cause a reduction in the abortion rate, and second, if not, it turns on whether plaintiffs have come forward with sufficient evidence in this case to enable this court to reach a reliable conclusion about whether the reduction would result from persuasive effects or from burdens on women seeking abortions. For the trial of this case, both Dr. Hen-shaw and Dr. Uhlenberg refined and supplemented their work. At trial the state also offered evidence from Dr. Lee Jen Wei, a professor of biostatistics at Harvard. At each step of the process, the dialogue has gone a step or two further as Dr. Henshaw and Dr. Uhlenberg (with the fresh support from Dr. Wei) have responded to criticisms from one another and from the courts. This process continued even beyond the trial as the court held the record open for further submissions of written evidence from Dr. Henshaw and Dr. Uhlenberg. As a result of this ongoing process of analysis, this court cannot assume that the courts’ and the experts’ earlier adoptions of or criticisms of Dr. Henshaw’s and Dr. Uhlenberg’s findings are still valid. As shown below, for example, in response to very specific criticisms of his study of Mississippi’s experience, Dr. Henshaw has conducted further analyses to test those criticisms. The court now turns to the evidence concerning several states, beginning with Mississippi. 2. Dr. Henshaw’s Mississippi Studies After Casey was decided in Jurie 1992, a similar Mississippi law took effect on August 8, 1992. The law requires a woman seeking an abortion to receive certain information in person at least 24 hours before the abortion. Dr. Henshaw and colleagues studied abortion data from Mississippi and other states to evaluate the effects of the law on women seeking abortions. The results of that study were published in the Journal of the American Medical Association in 1997. Ex. 224 (“JAMA article”). That study was more detailed, precise, and reliable than the preliminary study presented to this court in 1995 as part of the preliminary injunction proceedings. See 904 F.Supp. at 1454-55. (It also appears that the JAMA article may not have been available to the Seventh Circuit in Karlin v. Foust. The court’s opinion in Karlin refers to a study of the number of abortions performed in Mississippi in the seven months before the law took effect and five months after it took effect, see 188 F.3d at 486, which was the preliminary study this court also considered. See Ex. 3.) The analysis in the JAMA article used both a time series model looking at the twelve months before and the twelve months after the law took effect, and regression analyses using data from the three years before the law took effect and the two years after. The JAMÁ article also introduced comparisons to Georgia and South Carolina for both the time series and the regression analy-ses. The Mississippi study in the JAMA article shows that when the twelve months before and after the new Mississippi law took effect are compared: (1) the total rate of abortions for Mississippi residents (regardless of where they were obtained) decreased by approximately 16 percent; (2) the proportion of Mississippi residents traveling to other states to obtain abortions increased by approximately 37 percent; and (3) the proportion of second trimester abortions for Mississippi residents increased by approximately 40 percent. Ex. 224 at 655. (The state’s expert Dr. Uhlenberg replicated these results, but criticized their reliability for reasons discussed below. See Ex. 301 ¶ 9.) To test the significance of these changes, Dr. Henshaw and his colleagues compared the Mississippi data to data from Georgia and South Carolina, which did not impose similar changes in their abortion laws. All three states are southern states and had comparable abortion laws, apart from the law being evaluated. All three also had relatively reliable abortion data, including data from neighboring states under reciprocal agreements. Thus, the authors could account for women who left their home states to have abortions in neighboring states. See Ex. 224 at 654. The authors also noted that all three states had large non-white populations, which allowed race-specific analysis of data. Id. The time series analysis showed that Mississippi residents obtained 16 percent fewer abortions in the twelve months after the law took effect as compared to the twelve months before. Abortions during that same period declined by three percent in Georgia and by five percent in South Carolina. Ex. 224 at 655 (Table 1). By comparing the changes in rates, the researchers found that the decrease in Mississippi was 14 percent as compared to Georgia and 12 percent as compared to South Carolina. The confidence intervals for the Mississippi changes as compared to the changes in the other states showed that the changes in Mississippi were statistically significant, meaning they were unlikely to have resulted from random fluctuations. Id. The researchers also broke down the data for all three states by age and race. The declines in Mississippi rates were greatest for white adults and white teenagers, indicating decreases of 24 and 22 percent respectively. Ex. 224 at 656 (Table 2). The decline for non-white adults in Mississippi was smaller, and there was no decline for non-white teenagers. The increase in second trimester abortions in Mississippi also withstood comparison to the other states. Georgia and South Carolina experienced essentially no change in the rate of second trimester abortions. The rate in Mississippi increased by a very substantial 40 percent. The difference was statistically significant. Ex. 224 at 655 (Table 1), 657. In addition to the comparison of rates in the twelve months before and after the Mississippi law took effect, Dr. Henshaw and his colleagues also conducted regression analyses for the abortion rates in Mississippi, South Carolina, and Georgia, using data from January 1989 through December 1994. The regression analyses showed that overall abortion rates declined between 10 and 13 percent in Mississippi after the two-trip law took effect, as compared to South Carolina and Georgia. Ex. 224 at 656 & Table 4. The decrease was statistically significant. When results were broken down by age and race, the decrease in Mississippi was greater for white adults and white teenagers than for all women. Results from the regression analyses for non-whites were consistent with the results of the twelve-month before/twelve-month after comparison but were not statistically significant by themselves. Ex. 224 at 656. In the JAMA article, Dr. Henshaw and his colleagues reached the conclusion that the Mississippi two-trip law caused a statistically significant reduction in the rate of abortions for women residing in Mississippi. That conclusion is consistent with Judge Crabb’s finding in Karlin based on the preliminary data. 975 F.Supp. at 1217. In the JAMA article, Dr. Henshaw and his colleagues tested a number of alternative explanations for the decrease in abortion rates that they observed in Mississippi. They considered whether the decrease was part of a long-term trend of declining abortion rates in Mississippi. In fact, however, there was no such trend. Abortion rates had risen significantly in Mississippi in the years before the law took effect. Dr. Henshaw also examined data showing a two percent nationwide drop in abortion rates, which was not sufficient to account for the much larger decrease in Mississippi. See Ex. 205 ¶ 2(f). Dr. Hen-shaw also considered whether changes in prices or the number of providers of abortions could explain the decrease. He found no support for those explanations, id., which is important in light of the reported closing of a clinic discussed in the Karlin opinions. He also found no significant change in the number of women of childbearing age in Mississippi. Id. Furthermore, by including in the JAMA article the comparison of Mississippi’s experience to experience in South Carolina and Georgia, Dr. Henshaw and his colleagues also provided a further means for accounting for other trends and factors that might have affected abortion rates in Mississippi and those other states. See Ex. 224. 3. Criticisms of Dr. Henshaw’s Mississippi Studies Defense experts Dr. Uhlenberg and Dr. Wei both opined that the evidence did not show the Mississippi two-trip law had any statistically significant effect on abortion rates. The court considers first Dr. Uh-lenberg’s criticisms and then Dr. Wei’s. Consideration of the defense experts’ numerous criticisms requires delving into some fairly intricate statistical and data quality issues. Dr. Uhlenberg criticized Dr. Henshaw’s Mississippi study because many out-of-state abortions for Mississippi women (especially in Louisiana) were not reported and included in the data. Ex. 301 ¶ 9. In writing the JAMA article, however, Dr. Henshaw and his colleagues made adjustments on the generous assumption that 1,000 abortions per year for Mississippi women in Louisiana were not reported to Mississippi state officials for inclusion in the data. See Ex. 224 at 654 (referencing estimate). They still found a 10 percent decrease in abortion rates for Mississippi residents attributable to the law, which was still statistically significant. Tr. at 52 (Henshaw). In the JAMA article Dr. Hen-shaw and his colleagues also tested the sensitivity of their conclusions to changes in out-of-state abortions. They assumed for purposes of argument that there was an increase in travel to Louisiana for abortions equal to the sum of increase in travel to Alabama and Tennessee. Under that generous assumption, the overall drop in abortion rate would be 13 percent instead of 16 percent, but still statistically significant. Ex. 224 at 657; see also Ex. 206 ¶ 24 (same). In response to Dr. Uhlenberg’s criticism, Dr. Henshaw also ran additional analyses of the data in which he excluded southern Mississippi counties. Dr. Hen-shaw took this approach because any missing data from Louisiana and any data problems resulting from inaccurate reports from, or the closing of, a particular clinic in southern Mississippi would have had their greatest effects on data on abortions obtained by residents of those counties. When the data Dr. Uhlenberg had criticized were excluded entirely from the analysis, Dr. Henshaw found comparable statistically significant results for the effect of the law in the rest of the state. Henshaw Supp. Dep. 10-13, 26 & Exs. 255 & 256. Thus, the criticisms based on possible data problems from Louisiana or the accuracy of reports from the southern-most counties in Mississippi do not undermine Dr. Hen-shaw’s conclusions. As a further test of Dr. Henshaw’s methods and conclusions, Dr. Uhlenberg tried to determine whether he could show a similar statistical effect attributable to a random month, not tied to enactment of Mississippi’s two-trip law. Specifically, Dr. Uhlenberg selected June 1992 for his test, which was two months before the two-trip law took effect and in which there was an unusual one-month decrease in reported abortions in the Mississippi state data. Using Dr. Henshaw’s regression calculations, Dr. Uhlenberg found a significant drop as of June 1992. Tr. 123. Dr. Uhlenberg speculated that a change in Mississippi standards for abortion clinics effective in June 1992 might have affected the number of reported abortions. Id. The criticism is not persuasive. Dr. Henshaw noted that if the change of law had a significant effect in August 1992, one would also expect to detect a statistically significant effect with respect to other months close in time to August 1992. The reason is simply that, when data are compared over a significant period of time, a shift of one or two or three months may not be large enough to dilute the measured effect. See Tr. 42-43, 54 (Henshaw); Ex. 206 ¶ 27. In addition, Dr. Uhlenberg’s speculation about the effects of a change in clinic standards affecting the overall rate of reported abortions was based primarily on a legal dispute involving one abortion provider who operated a clinic in Gulfport, along the coast of the Gulf of Mexico in southern-most Mississippi. See Ex. 300 ¶¶ 22-23; Uhlenberg Supp. Dep., passim. Dr. Uhlenberg’s reliance on that situation is not supported by the evidence. First, Dr. Uhlenberg’s information about that situation consists of conflicting affidavits from two sides in a bitter lawsuit in Mississippi. Neither Dr. Uhlenberg nor this court can make a reliable assessment of the credibility of either side’s view of the facts in that case. See Uhlenberg Supp. Dep. 37 (admitting he could not tell who was telling the truth in the dispute); id. at 44-45 (same). Second, and more compelling for present purposes, Dr. Henshaw analyzed the Mississippi data without taking into account the southern-most counties in Mississippi. For purposes of that analysis, he assumed Dr. Uhlenberg was right in doubting the reliability of the data from those counties. Dr. Henshaw’s analysis excluding those counties again showed a substantial statistically significant decline in the rest of the state. Henshaw Supp. Dep. 10-13 and Exs. 255 & 256. Third, Dr. Henshaw also looked more closely at Dr. Uhlenberg’s hypothesis that the clinic standards law might account for a drop in the abortion rate in June 1992. Dr. Henshaw examined county-by-county reports of abortions from Mississippi for the period 1990-1993, which are in the record as Exhibits 250, 251, 252, and 253 to his supplemental deposition. Based on his review of those reports', Dr. Henshaw found two months — April 1991 and June 1992 — for which there appeared anomalies best explained as reporting errors, both in DeSoto County (adjacent to Memphis, Tennessee). For the years 1990 to 1993, the reports show that residents of DeSoto County generally obtained 10 to 20 abortions per month with a few months above or below that range. The report for June 1992, however, shows zero abortions that month for residents of DeSoto County and a substantial drop in abortions for women from other states. See Ex. 252 (reporting 10 abortions in May for DeSoto County residents, zero in June, and 17 in July; reported abortions for out-of-state residents were 180 in May, 42 in June, and 171 in July). To account for those errors, Dr. Hen-shaw compared abortion numbers in June 1992 for the entire state, for the northern counties, and for the state without the northern counties, and he compared June 1992 to Junes of 1990, 1991, and 1993. The results, shown in Exhibit 257, reflect a dramatic fluctuation in the northern counties from a high of 141 abortions for residents of those counties in June 1991, down to a low of 45 in June 1992 (a 68 percent drop), and an increase to 83 in June 1993 (an 84 percent increase over 1992). These problems in the data for June 1992 tend to undermine Dr. Uhlenberg’s hypothesis that the change in clinic standards actually accounted for any observed reduction in the abortion rate. Even more important, however, because the data problems would give pause to any cautious researcher, Dr. Henshaw tested his regression methods and analysis if both the northern-most and the southern-most counties were excluded. He still found a substantial statistically significant drop in abortion rates resulting from the two-trip law in August 1992. He also tried excluding the data from both April 1991 and June 1992 in his calculations, in light of the reasons for doubting the accuracy of the two reports for those months. When those months were excluded entirely, Dr. Henshaw found even larger statistical effects for the two-trip law. Henshaw Supp. Dep. 23-26; Ex. 256 (Regression 2) (coefficient of absolute number of abortions changed from -91.18 to -115.68, and p-value changed from 0.0263 to 0.0046); see also Henshaw Supp. Dep. 12-13 (explaining meaning of coefficients and p-values). Thus, Dr. Henshaw ran to ground those criticisms by Dr. Uhlenberg, and he showed convincingly that they did not affect the overall reliability of his conclusions. These additional studies are important because they are new. Not only do they address Dr. Uhlenberg’s criticisms, but they address the principal problem identified by Judge Crabb and the Seventh Circuit in Karlin v. Foust with respect to whether the two-trip law caused abortion rates in Mississippi to decrease. The Seventh Circuit noted in its opinion that Dr. Henshaw’s initial study was compromised by the combination of missing data from Louisiana and a problem resulting from either the closure of “the sole abortion facility” in southern Mississippi in early 1992 or that clinic’s failure to report abortions for some period. See Karlin, 188 F.3d at 487 (closed); Karlin, 975 F.Supp. at 1216 (either closed or stopped reporting). Dr. Henshaw’s further analysis here shows that even if one excludes the southern-most counties from the analysis— where the effects of missing Louisiana data and the closure of a clinic in Gulfport would obviously be greatest — the overall effect of the two-trip law in August 1992 remains substantial and statistically significant. Dr. Uhlenberg also tested Dr. Hen-shaw’s conclusions by studying birth rates in Mississippi. If the Mississippi two-trip law had the effect of preventing abortions, one would expect to see an increase in the number of births in the state, all other things being equal. Dr. Uhlenberg found no such increase. Ex. 300, ¶¶ 14-16. Although this criticism of Dr. Henshaw’s analysis has some superficial appeal, it is not persuasive. The abortion rate is much lower than the birth rate in Mississippi. See, e.g., Ex. 201 at 6 (in 1992, Mississippi had 177 abortions for every 1000 live births). Thus, a 10 to 13 percent decrease in the abortion rate would be expected to produce a much smaller percentage change in the birth rate. In addition, some proportion of the pregnancies would end in miscarriages in any event, although Dr. Uhlenberg made an adjustment for that fact. Ex. 300, Table 5. Also, Dr. Uhlenberg concentrated his analysis on births to unmarried white women in Mississippi. See Ex. 301, ¶¶ 2-5. (Dr. Hen-shaw’s data indicated the law had stronger effects on white women than on non-white women, and abortions were much more common among unmarried women than among all women. See Ex. 201 at 22 (in 1992, 85 percent of abortions in Mississippi were performed for unmarried women). Thus, any effect on the birth rate should have been most readily observable among unmarried white women.) By looking for changes in births to unmarried white women, however, Dr. Uhlenberg did not account for the proportion of single pregnant women who marry before they give birth. The evidence indicated that the proportion would be about 29 or 30 percent. See Tr. 373. In addition, Dr. Uhlenberg developed his projected number of births from comparisons to birth rates in neighboring states that simply did not correlate well with Mississippi’s birth rate at times other than those being examined and was not a reliable basis for the projection. Tr. 45-46. In fact, Exhibit 240 shows a small but noticeable increase in births among unmarried white women between 1992 and 1994. Dr. Henshaw and his colleagues also looked at birth rates as part of their study reported in the JAMA article. Looking at the relevant time period, they observed that birth rates declined in Mississippi, South Carolina, and Georgia, but the decline was the smallest in Mississippi. The data on birth rates were inconclusive, but they were not inconsistent with the hypothesis that the Mississippi law caused an increase in the number of unintended pregnancies carried to term. See Ex. 224 at 657-58; see also id. at 655 (Figures 1 & 2 and Table 1). Thus, when adjustments are made for miscarriages and for marriages before the birth of a baby, the predicted change in the birth rate is small compared to the degree of random fluctuations in birth rates. As a result, random fluctuations could easily mask such effects of changes in the abortion rate. See Ex. 206 ¶ 28. The birth data are not inconsistent with Dr. Henshaw’s conclusions. The absence of a detectable effect in the birth rate does not undermine the validity of Dr. Hen-shaw’s finding that the Mississippi two-trip law caused a statistically significant reduction in the rate of abortions. Dr. Uhlenberg also criticized Dr. Hen-shaw’s conclusions because of the different results found for whites and non-whites in the Mississippi data. As noted above, Dr. Henshaw and his colleagues found greater decreases in abortion rates for whites in Mississippi. The decrease for non-white adults as shown through the regression analyses was smaller and not statistically significant by itself. There was no decrease for non-white teenagers. See Ex. 224. Dr. Uhlenberg used annual data for non-white women in Mississippi and found a slight increase in the number of abortions for that group after the Mississippi law took effect, although that result may well be an artifact of the particular time periods chosen, which used longer time periods and introduced more potentially confounding variables. See Ex. 300 ¶¶ 9-13 & Table 4. Dr. Uhlenberg found no apparent explanation for the different results by race, and he argued that this anomaly undermines Dr. Henshaw’s analysis and tends to show the decrease observed by Dr. Henshaw was an artificial phenomenon rather than a true effect of the Mississippi law. See also Tr. 116. Dr. Henshaw and his colleagues had also considered the issue of race. In the JAMA article, they wrote that the differences by race were “more difficult to interpret.” Ex. 224 at 657. They suggested two possible explanations in the article, but neither is especially persuasive. It is clear, however, that there are substantial underlying differences by race with respect to both birth rates and abortion rates in Mississippi. Figure 2 of Exhibit 224 shows, for example, birth rates of 85 to 90 births per 1000 non-white women, but 58 to 62 births per 1000 white women. Figure 1 shows abortion rates of 14 to 17 abortions per 1000 non-white women, but 8 to 11 abortions per 1000 white women. With differences that large in these underlying rates, the differences in measured effects by race do not undermine the conclusion that the law had significant effects on the overall rate of abortions. Dr. Wei agreed with Dr. Uhlenberg on some of the criticisms discussed above. He also added some criticisms of his own and carried out some statistical analyses that persuaded him that the Mississippi data do not show a statistically significant decline in the abortion rate attributable to the two-trip law taking effect. Dr. Wei criticized Dr. Henshaw for failing to take account of “the declining national trend in the number of abortions.” Ex. 303 ¶ 5. However, Dr. Henshaw’s use of South Carolina and Georgia as comparator states in the JAMA article provided a reasonable means for accounting for such changes. See Ex. 224 at 656-58. In addition, Mississippi had experienced increases in its relatively low abortion rate while the national average rate was declining slightly. The modest downward national trend does not undermine Dr. Henshaw’s conclusions. Dr. Wei also criticized Dr. Henshaw for failing to account for changing beliefs regarding abortion. See Ex. 303 ¶ 5. Again, however, the use of comparator states provided a mechanism for accounting for such changes (if they had actually occurred) on a national or regional level. Moreover, Dr. Wei and the defendants have not come forward with any actual evidence of such changes at any relevant times. It is always possible to imagine other possible confounding factors. Without some reason to believe the factors are real, the mere possibility does not undermine Dr. Hen-shaw’s results. Dr. Wei also cited an article by Meier and others that he described as providing “scientific evidence that state-level restrictions have no effect on the incidence of abortion.” Ex. 303 ¶ 6, citing Meier, et al., The Impact of State-Level Restrictions on Abortion, 1996 Demography 307. Dr. Wei seriously overstated the conclusions of that article, which is in the record as Exhibit 328. The Meier paper studied 23 separate state policy actions in the wake of Webster v. Reproductive Health Services, 492 U.S. 490, 109 S.Ct. 3040, 106 L.Ed.2d 410 (1989). The authors found no evidence that the policies they studied had an appreciable impact on abortion rates. Ex. 328 at 311. The authors recognized, however, as Dr. Wei did not, that their findings did “not refute the possibility that state policies can affect the rate of induced abortions.” Id. They also recognized that state Medicaid policies affect abortion rates. The authors did not purport to have come forward with “scientific evidence that state-level restrictions have no effect on the incidence of abortion,” without qualification. Even more relevant for this case, Meier and his colleagues also recognized the importance of the Casey decision. They cited “early anecdotal or case analysis” suggesting that the waiting period and mandatory disclosure laws have an effect. Id. The authors also suggested more systematic study of the effects of such laws as later data became available. Id. Dr. Henshaw’s JAMA article is precisely such an effort, and it provides substantial proof of such effects. Dr. Wei also identified a number of additional possible confounding and intervening variables, including changes in abortion pricing, levels of anti-abortion activity “perceived as harassment” at clinics, negative publicity, the quality of services, changed service availability, the reputation of providers, the scope and funding of abstinence programs, the scope and availability of services providing alternatives to abortion, differences in geographic proximity of abortion services, misreporting and underreporting of abortion data, changing laws on abortion facilities, changes in Medicaid financing for abortion, and alterations in reimbursement for abortions by private insurance carriers. Ex. 303 ¶ 7. Dr. Wei claimed: “None of these were accounted for by Dr. Henshaw in examining the impact of Mississippi’s waiting period and informed consent law.” Plaintiffs have objected to Dr. Wei’s testimony on these points, arguing that he is not qualified to offer an admissible expert opinion on the topic of potential confounding factors affecting studies of abortion rates. Dr. Wei had not studied abortion rates or the subject of abortion more generally before he began working on this case. However, he is experienced in statistical analysis in a wide range of areas relating to human health and medical practice. He testified that when he begins to do statistical work in a new field (independent of litigation work, which is a small part of his work), he immerses himself in the literature of the field to help him get a feel for what sorts of factors are most likely to affect and possibly to obscure the data. The court credits that testimony and overrules plaintiffs’ objections to Dr. Wei’s testimony on that score. Plaintiffs’ criticism goes to the weight of his testimony, not its admissibility. However, Dr. Wei’s testimony on potential confounding factors is entitled to little weight. Dr. Wei’s criticism of Dr. Henshaw for failing to account for potential confounding factors simply is not accurate. As noted above, Dr. Henshaw and his colleagues considered many of these factors, and they used methods designed to minimize and control for such effects. For example, they considered changes in prices, changes in the number of providers, and changes in the population of women of child-bearing age. They found no significant changes in these factors. In addition, they used comparisons to Georgia and South Carolina, which share important similarities with Mississippi, as a useful control for other factors, including those that might have affected abortion rates more broadly. As for the other potential confounding factors identified by Dr. Wei, there is simply no evidence that any of these factors changed in Mississippi during the time period studied by Dr. Henshaw. Dr. Henshaw and his colleagues made a reasonable and serious effort to account for likely confounding variables. Their effort was easily sufficient to render their analysis relevant and admissible. See generally Bazemore v. Friday, 478 U.S. 385, 400, 106 S.Ct. 3000, 92 L.Ed.2d 315 (1986) (“Normally, failure to include variables will affect the analysis’ probativeness, not its admissibility.”); Adams v. Ameritech Services, Inc., 231 F.3d 414, 422-28 (7th Cir.2000) (reversing district court finding that statistical reports were inadmissible). It is almost always possible for someone disappointed with the results of statistical analysis to hypothesize one more independent variable that might explain the results. Such a bare hypothesis is not sufficient to undermine the analysis. See Allen v. Seidman, 881 F.2d 375, 380 (7th Cir.1989) (defendant’s “attack on the plaintiffs’ statistical case amounts to a contention that unless a plaintiff