This case requires us to address the scope of congressional power under Section 5 of the Fourteenth Amendment to abrogate the immunity of the states from suit in federal court which the states would otherwise enjoy under the Eleventh Amendment and Supreme Court precedent. At issue is whether the creation of a private cause of action against a state for money damages under the personal medical leave provision of the Family and Medical Leave Act, 29 U.S.C. § 2612(a)(1)(D), validly abrogates that immunity as an exercise of Congress’s Section 5 powers.
Following the analytic framework suggested by two recent Supreme Court decisions, Kimel v. Florida Board of Regents, 528 U.S. 62, 120 S.Ct. 631, 145 L.Ed.2d 522 (2000), and Board of Trustees of the University of Alabama v. Garrett, 531 U.S. 356, 121 S.Ct. 955, 148 L.Ed.2d 866 (2001), we hold that the FMLA’s personal medical leave provision, 29 U.S.C. § 2612(a)(1)(D) (affording leave for serious personal health conditions), insofar as it authorizes private suits against states, does not validly abrogate the states’ immunity.1 Every circuit court which has addressed the personal medical leave provision of the FMLA in this context has held that that provision does not abrogate the immunity of the state as employer in the face of the Eleventh Amendment. Our holding is narrow: the present legislative record does not demonstrate that the personal medical leave provision of the FMLA is an appropriate response necessary to remedy or prevent unconstitutional gender discrimination practiced by the states as employers.
I.
Stephen Laro was employed by the State of New Hampshire as a computer specialist for the New Hampshire Retirement System. In early. 1998, he had heart bypass surgery. Because of his medical condition, following his surgery he requested and received leave under the FMLA, which leave began on March 6, 1998. Laro’s physician provided the State with a certification which said that Laro’s condition required him to be out of work for at least eight weeks, or until at least May 3, 1998. Apparently his employer understood that to mean he requested leave only until that day. When he did not return to work as of May 5, 1998, his employer inquired, and Laro explained that his physician had not yet cleared him to return to work. On May 8, 1998, the State wrote to Laro, informing him that his FMLA leave would expire as of May 29, 1998. Laro replied that he would not need any more time than that, and on May 18, 1998, he provided his employer with a letter from his physician authorizing his immediate return to work. Laro’s employer then told him that before returning to work he had to meet with his supervisors, and asked him to schedule an appointment. At this time, Laro expected to return to work on Thursday, May 21, before the expiration of the twelve week FMLA period. Instead, he was given a termination letter, dated May 21, 1998, and setting an effective termination date of May 29, 1998. The termination letter stated that Laro had exhausted his accumulated leave bal-*5anees and that he was unable to meet the New Hampshire Retirement System’s attendance requirements. Other than that statement there is no explanation in the record for New Hampshire’s about-face on Laro’s ability to return to work.
Laro sued for monetary damages in federal court, claiming that the state had violated 29 U.S.C. § 2615 by terminating his employment before the expiration of the twelve week period of unpaid leave guaranteed under the FMLA. The state moved to dismiss, asserting its immunity under the Eleventh Amendment. The district court agreed, and dismissed the action. Laro now appeals, joined by the United States as intervenor.
II.
The Eleventh Amendment provides that “[t]he Judicial power of the United States shall not be construed to extend to any suit in law or equity, commenced or prosecuted against one of the United States by Citizens of another State, or by Citizens or Subjects of any Foreign State.” U.S. Const, amend. XI. Though limited by its express terms, the Amendment has been construed broadly by the Supreme Court, such that its “ultimate guarantee” is “that nonconsenting States may not be sued by private individuals in federal court.” Garrett, 121 S.Ct. at 962.
Congress may abrogate this Eleventh Amendment immunity when it both “unequivocally intends to do so” and “act[s] pursuant to a valid grant of constitutional authority.” Id. (citing Kimel, 528 U.S. at 73, 120 S.Ct. 631).2 Here, the State properly concedes that Congress did intend to abrogate the states’ immunity. See 29 U.S.C. § 2617(a)(2) (extending private right of action for damages to employees against “any employer (including a public agency)”); 29 U.S.C. § 2611(4)(A)(iii) (defining “employer” to include any “public agency” and cross-referencing 29 U.S.C. § 203(x) (defining “public agency” to include “the government of a State or political subdivision thereof’)). The question, then, is whether Congress, in subjecting the states to suit in federal court for money damages under the personal medical leave provision of the FMLA, acted appropriately pursuant to a valid grant of constitutional authority.
The Supreme Court has held that Congress may not properly base its abrogation of Eleventh Amendment immunity upon the powers enumerated in Article I. Garrett, 121 S.Ct. at 962; Kimel, 528 U.S. at 79, 120 S.Ct. 631; Seminole Tribe of Fla. v. Florida, 517 U.S. 44, 72-73, 116 S.Ct. 1114, 134 L.Ed.2d 252 (1996). Where Congress acts pursuant to a valid exercise of its power under Section 5 of the Fourteenth Amendment, however, it may subject nonconsenting states to suit in federal court. Garrett, 121 S.Ct. at 962 (“[T]he Eleventh Amendment, and the principle of state sovereignty which it embodies, are necessarily limited by the enforcement provisions of § 5 of the Fourteenth Amendment.”) (quoting Fitzpatrick v. Bitzer, 427 U.S. 445, 456, 96 S.Ct. 2666, 49 L.Ed.2d 614 (1976)); Seminole Tribe, 517 U.S. at 59, 116 S.Ct. 1114 (“[T]hrough the Fourteenth Amendment, federal power extended to intrude upon the province of the Eleventh Amendment and therefore [ ] § 5 of the Fourteenth Amendment allowed Congress to abrogate the immunity from suit guaranteed by that Amendment.”). That is so because the Fourteenth Amendment expressly empowers Congress to enforce its provisions against the states. See Seminole Tribe, 517 U.S. at 59, 116 S.Ct. *61114 (“[T]he Fourteenth Amendment, by expanding federal power at the expense of state autonomy, [ ] fundamentally altered the balance of state and federal power struck by the Constitution.”); Ex parte Virginia, 100 U.S. 339, 346, 25 L.Ed. 676 (1879) (“The prohibitions of the Fourteenth Amendment are directed to the States, and they are to a degree restrictions of State power. It is these which Congress is empowered to enforce, and to enforce against State action.... Such enforcement is no invasion of State sovereignty. No law can be, which the people of the States have, by the Constitution of the United States, empowered Congress to enact.”).
Section 1 of the Fourteenth Amendment provides, in relevant part:
No state shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any State deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws.
U.S. Const, amend. XIV. Section 5 of the Fourteenth Amendment grants Congress the power to enforce the substantive guarantees contained in Section 1 by enacting “appropriate legislation.”3 Id. This provision embodies an affirmative grant of power to Congress. See City of Boerne v. Flores, 521 U.S. 507, 517, 117 S.Ct. 2157, 138 L.Ed.2d 624 (1997) (“All must acknowledge that § 5 is ‘a positive grant of legislative power’ to Congress.”) (quoting Katzenbach v. Morgan, 384 U.S. 641, 651, 86 S.Ct. 1717, 16 L.Ed.2d 828 (1966)).
Section 5 of the Fourteenth Amendment, then, affords Congress a broad remedial and prophylactic power to enact legislation restricting state action, so long as the legislation reflects a rational means, under the standard of McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 421, 4 L.Ed. 579 (1819), to realize the end of enforcing the Fourteenth Amendment’s substantive prohibitions. See City of Boerne, 521 U.S. at 517-18, 117 S.Ct. 2157 (“Whatever legislation is appropriate, that is, adapted to carry out the objects the amendments have in view, whatever tends to enforce submission to the prohibitions they contain, and to secure to all persons the enjoyment of perfect equality of civil rights and the equal protection of the laws against State denial or invasion, if not prohibited, is brought within the domain of congressional power.”) (quoting Ex pante Virginia, 100 U.S. at 345-46); Katzenbach v. Morgan, 384 U.S. at 650, 86 S.Ct. 1717 (“By including § 5 the draftsmen sought to grant to Congress, by a specific provision applicable to the Fourteenth Amendment, the same broad powers expressed in the Necessary and Proper Clause, Art. I, § 8, cl. 18.”); id. at 651, 86 S.Ct. 1717 (“[T]he McCulloch v. Maryland standard is the measure of what constitutes ‘appropriate legislation’ under § 5 of the Fourteenth Amendment.”); see also South Carolina v. Katzenbach, 383 U.S. 301, 324-27, 86 S.Ct. 803, 15 L.Ed.2d 769 (1966) (holding that under the enforcement clause of the Fifteenth Amendment, Congress may employ “any rational means to effectuate [its] constitutional prohibitions,” and citing the standard of McCulloch v. Maryland as “[t]he basic test to be applied in a case involving § 2 of the Fifteenth Amendment”); James Everard’s Breweries v. Day, 265 U.S. 545, 558-59, 44 S.Ct. 628, 68 L.Ed. 1174 (1924) (legislation under the enforcement clause of the Eighteenth Amendment assessed under the standard *7of McCulloch v. Maryland). Indeed, by its express terms, the Necessary and Proper Clause applies to enumerated powers added to the Constitution after its adoption. See U.S. Const. Art. I, § 8, cl. 18 (granting Congress the power “[t]o make all Laws which shall be necessary and proper for carrying into Execution the foregoing Powers, and all other Powers vested by this Constitution in the Government of the United States, or in any Department or Officer thereof’) (emphasis added).
The scope of Section 5 of the Fourteenth Amendment enables Congress to do more than simply prohibit what the Amendment itself already prohibits; Congress may enact legislation which is rationally tailored to prevent or deter the incidence of violations of the Amendment. See, e.g., Garrett, 121 S.Ct. at 963 (“Congress is not limited to mere legislative repetition of this Court’s constitutional jurisprudence.”); Kimel, 528 U.S. at 81, 120 S.Ct. 631 (“Congress’ § 5 power is not confined to the enactment of legislation that merely parrots the precise wording of the Fourteenth Amendment. Rather, Congress’ power ‘to enforce’ the Amendment includes the authority both to remedy and to deter violation of rights guaranteed thereunder by prohibiting a somewhat broader swath of conduct, including that which is not itself forbidden by the Amendment’s text.”); City of Boerne, 521 U.S. at 518, 117 S.Ct. 2157 (“Legislation which deters or remedies constitutional violations can fall within the sweep of Congress’ enforcement power even if in the process it prohibits conduct which is not itself unconstitutional and intrudes into ‘legislative spheres of autonomy previously reserved to the States.’ ”) (quoting Fitzpatrick, 427 U.S. at 455, 96 S.Ct. 2666).
Indeed, the enforcement clause delegates to Congress in the first instance the responsibility to determine what prophylactic measures are necessary to enforce most effectively the Amendment’s substantive provisions. See, e.g., City of Boerne, 521 U.S. at 536, 117 S.Ct. 2157 (“It is for Congress in the first instance to ‘deter-min[e] whether and what legislation is needed to secure the guarantees of the Fourteenth Amendment,’ and its conclusions are entitled to much deference.”) (quoting Katzenbach v. Morgan, 384 U.S. at 651, 86 S.Ct. 1717); Katzenbach v. Morgan, 384 U.S. at 648, 86 S.Ct. 1717 (“It is the power of Congress which has been enlarged. Congress is authorized to enforce the prohibitions by appropriate legislation. Some legislation is contemplated to make the amendments fully effective.”) (quoting Ex parte Virginia, 100 U.S. at 345); see also South Carolina v. Katzenbach, 383 U.S. at 326, 86 S.Ct. 803 (enforcement clause of the Fifteenth Amendment indicates that Congress was to be “chiefly responsible for implementing” the rights created in the Amendment).
Because the Fourteenth Amendment affords this role to Congress, courts consider congressional conclusions with considerable deference. See, e.g., Katzenbach v. Morgan, 384 U.S. at 653, 86 S.Ct. 1717 (“It was for Congress ... to assess and weigh the various conflicting considerations .... It is not for us to review the congressional resolution of these factors. It is enough that we be able to perceive a basis upon which the Congress might resolve the conflict as it did.”); James Everard’s Breweries, 265 U.S. at 560, 44 S.Ct. 628 (“In enacting [enforcement] legislation, Congress has affirmed its validity. That determination must be given great weight; this Court by an unbroken line of decisions having steadily adhered to the rule that every possible presumption is in favor of *8the validity of an act of Congress until overcome beyond rational doubt.”) (internal quotation marks omitted).
Nevertheless, this congressional power is not without limits; congressional power under Section 5 “is but a limited authority [ ] extending only to a single class of cases,” though within those limits it is “complete.” Ex parte Virginia, 100 U.S. at 348; see also City of Boerne, 521 U.S. at 518, 117 S.Ct. 2157 (“[A]s broad as the congressional enforcement power is, it is not unlimited.”) (quoting Oregon v. Mitchell, 400 U.S. 112, 128, 91 S.Ct. 260, 27 L.Ed.2d 272 (1970) (opinion of Black, J.)). The limits are those “prescribed” in the Constitution itself. See Gibbons v. Ogden, 22 U.S. (9 Wheat.) 1, 196, 6 L.Ed. 23 (1824). That is, the ends of appropriate enforcement legislation under Section 5 must be the enforcement of the substantive guarantees of the Fourteenth Amendment. See Kimel, 528 U.S. at 81, 120 S.Ct. 631 (the language of Section 5 limits congressional power to enacting that legislation appropriate “to enforce” the provisions of the Amendment); accord City of Boerne, 521 U.S. at 519, 117 S.Ct. 2157.
Consequently, despite Congress’s initial responsibility to gauge appropriate enforcement legislation, federal courts retain the power to review that legislation to ensure that it pursues appropriate ends. See City of Boerne, 521 U.S. at 536, 117 S.Ct. 2157 (“[T]he courts retain the power ... to determine if Congress has exceeded its authority under the Constitution.”); see also James Everard’s Breweries, 265 U.S. at 559, 44 S.Ct. 628 (“What [the Court] may consider is whether that which has been done by Congress has gone beyond the constitutional limits upon its legislative discretion.”). The unconstitutional exercise of enforcement powers occurs where Congress attacks “evils” not comprehended by the amendments. See South Carolina v. Katzenbach, 383 U.S. at 326-27, 86 S.Ct. 803.
Two recent Supreme Court cases present a framework for assessing whether legislation enacted under Section 5 falls within appropriate bounds in the face of a claim of Eleventh Amendment immunity. See Garrett, 531 U.S. 356, 121 S.Ct. 955, 148 L.Ed.2d 866; Kimel, 528 U.S. 62, 120 S.Ct. 631, 145 L.Ed.2d 522. The first step of this Eleventh Amendment analysis requires the court to identify the precise scope of the constitutional right at issue, including as it is reflected in the degree of judicial scrutiny afforded the state statute or action. See, e.g., Garnett, 121 S.Ct. at 963 (“The first step in applying these now familiar principles is to identify with some. precision the scope of the constitutional right at issue.”). Whether the legislation is prophylactic or remedial, the court must then examine whether the legislative means adopted by Congress have a congruence and proportionality to the constitutional injury to be prevented or remedied.4 See, e.g., Kimel, 528 U.S. at 81, 120 S.Ct. 631 (“[Rjecognizing that *9‘Congress must have wide latitude in determining where [the line between measures that remedy or prevent unconstitutional actions and measures that make a substantive change in governing law] lies,’ we [have] held that ‘[t]here must be a congruence and proportionality between the injury to be prevented or remedied and the means adopted to that end.’ ”) (quoting City of Boerne, 521 U.S. at 520, 117 S.Ct. 2157). We turn to applying this framework to the FMLA.
III.
The Family and Medical Leave Act of 1993, 29 U.S.C. § 2601 et seq., entities eligible employees to twelve workweeks of unpaid leave per year in four specified situations: (1) the birth of a child, (2) the adoption of a child or placement of a foster child, (3) the need to care for a parent, child, or spouse with a serious health condition, and (4) the inability to work due to the employee’s own serious health condition. 29 U.S.C. § 2612(a)(1).5 The only issue before us in this case concerns the availability of private damages actions against states under the final provision, 29 U.S.C. § 2612(a)(1)(D), which affords an employee the right to unpaid leave for inability to work due to his or her own serious health condition.6
*10The first step in assessing the validity of this personal medical leave provision as appropriate enforcement legislation under Section 5 in the face of Eleventh Amendment immunity is to determine the substantive guarantee of the Fourteenth Amendment it is designed to protect. The FMLA was explicitly enacted in part pursuant to congressional power under Section 5 of the Fourteenth Amendment to address gender-based discrimination (as well as disability-based discrimination), in addition to Congress’s Aticle I, Section 8 powers under the Commerce Clause.
The findings section of the statute provides, in pertinent part:
(5) due to the nature of the roles of men and women in our society, the primary responsibility for family caretaking often falls on women, and such responsibility affects the working lives of women more than it affects the working lives of men; and
(6) employment standards that apply to one gender only have serious potential for encouraging employers to discriminate against employees and applicants for employment who are of that gender.
29 U.S.C. § 2601(a)(5)-(6). Therefore, the statute provides that it is the purpose of the Act:
(4) to accomplish [the statutory goals] in a manner that, consistent with the Equal Protection Clause of the Fourteenth Amendment, minimizes the potential for employment discrimination on the basis of sex by ensuring generally that leave is available for eligible medical reasons (including maternity-related disability) and for compelling family reasons, on a gender-neutral basis; and
(5) to promote the goal of equal employment opportunity for women and men, pursuant to such clause.
29 U.S.C. § 2601(b)(4)-(5).
The question, then, is whether the personal medical leave provision of the FMLA appropriately enforces the Fourteenth Amendment’s guarantee against gender-based discrimination by the states. The scope of constitutional protection against gender-based discrimination under the Fourteenth Amendment is broader than that against mere arbitrary classifications. Where gender-based discrimination is at issue, a state must justify its discrimination by a showing of considerably more than mere rationality. See, e.g., United States v. Morrison, 529 U.S. 598, 620, 120 S.Ct. 1740, 146 L.Ed.2d 658 (2000). Gender-based discrimination violates equal protection unless it “serves important governmental objectives and ... the discriminatory means employed are substantially related to the achievement of those objectives.” United States v. Virginia, 518 U.S. 515, 533, 116 S.Ct. 2264, 135 L.Ed.2d 735 (1996) (internal quotation marks omitted).
The personal medical leave provision of the FMLA at issue does not directly address gender discrimination on its face, but may serve a prophylactic function to prevent or deter unconstitutional state actions. Cf. Hundertmark v. Florida Dep’t of Transp., 205 F.3d 1272, 1276 (11th Cir.2000) (Equal Pay Act facially targets gender discrimination and therefore validly abrogates the states’ Eleventh Amendment immunity under Section 5). The issue, then, for Eleventh Amendment purposes, is whether there is “congruence and proportionality” between “the potential for employment discrimination on the basis of sex” by a state and the FMLA’s provision of twelve-weeks leave to employees because of their own personal health problems. We look only to the gender-based discrimination concerns of Congress.7
*11The Supreme Court has not recently-addressed the issue of congressional power to override a state’s Eleventh Amendment immunity through legislation designed to prevent unconstitutional gender-based discrimination in employment.8 But see Fitzpatrick v. Bitzer, 427 U.S. 445, 455, 96 S.Ct. 2666, 49 L.Ed.2d 614 (1976). Rather, the two employment statutes it has considered in this context address categories of discrimination — age discrimination in Ki-mel and disability discrimination in Garrett — where the courts impose a lesser burden on a state to justify discriminatory practices. See Kimel, 528 U.S. at 83-84, 120 S.Ct. 631; Garrett, 121 S.Ct. at 963-64. Even with the heightened standard of review for gender-based discrimination, however, we do not find that the legislative history sufficiently ties the FMLA’s personal medical leave provision to the prevention of gender-based discrimination by the states to survive Eleventh Amendment scrutiny. In the absence of such history, we must affirm.
Defending the personal medical leave provision within the context of the entire Act, Laro and the intervenor United States argue that the FMLA as a whole was motivated to prevent gender discrimination against both men and women, relying upon the following rationale. First, they say, the Act prevents discrimination against men; without the Act’s protection, men might be deprived of leave opportunities and thereby denied the opportunity to take equal responsibility for their families. In turn, the absence of leave opportunities for men, they say, serves to reinforce general stereotypes that women are the proper caregivers, both for children and for other family members, and thereby forces women back into those stereotypical roles. Such stereotypes, so reinforced, deter employers from hiring women or promoting them to positions of equal responsibility vis-a-vis men. The existence of such a reinforcing dynamic was demonstrated, they say, by experience with respect to state law maternity provisions: when maternity leave for women was mandated by state law, employers reacted by refusing to hire or promote women. Thus, legislation that solely protected women in an effort to eliminate discrimination perversely gave employers an economic incentive to discriminate against women. For these reasons, they say, the FMLA insists on gender-neutral leave provisions. For present purposes we accept all of that to be true.
In defending the personal medical leave provision of the FMLA as validly abrogating the states’ Eleventh Amendment immunity on this rationale, however, Laro and the United States must demonstrate *12how this particular provision contributes to the broader purposes they attribute to the Act as a whole. In doing so, the problems for Laro and the United States are twofold. First, in order for this particular provision to validly abrogate a state’s Eleventh Amendment immunity, it must be linked through some nexus not just to such gender-based problems in society at large, but specifically to unconstitutional gender discrimination by states in their capacity as employers. Second, establishing this connection is particularly difficult for the personal medical leave provision at issue, which, on its face, only bears on these concerns indirectly (likely this is why the plaintiff sought to tie his case to the other provisions of the FMLA). The argument that this provision validly abrogates New Hampshire’s Eleventh Amendment immunity founders on this lack of congruence between the personal medical leave provision at issue here and the prevention of gender-based discrimination by states as employers, because Congress has not found the states to have engaged in the specific gender-based discriminatory practices this provision was designed to prevent.
One rationale advanced to defend the provisions — that they serve to counteract gender stereotyping on family roles — is misplaced as to the personal medical leave provision at issue in this case. The argument is not without sense. It proceeds as follows: employers perceive that the primary responsibility for caretaking of family members in our society tends to fall on women; therefore, they presume women will inevitably need to take leave for such caretaking, and hence will either be less likely to hire women, or else will afford such leave to women but not to men, thus reinforcing gender roles. Therefore, the argument concludes, Congress might reasonably seek to break the cycle of stereotyping and discrimination by requiring gender-neutral leave policies for family care. This may all be true, but this argument does not go to the need for a 'personal medical leave provision. Rather, it provides a rationale for the parental and family-care leave provisions found in the first three subsections of the FMLA, which are not at issue in this case.
Attention to the legislative history reveals that Congress’s primary motivation for including the personal medical leave provision contained in subsection (D) was to protect families from the economic dislocation caused by a family member losing his or her job due to a serious medical problem. See S.Rep. No. 103-3, at 11 (1993), reprinted in 1993 U.S.C.C.A.N. 3, 13-14 (“The fundamental rationale for [a personal medical leave] policy is that it is unfair for an employee to be terminated when he or she is struck with a serious illness and is not capable of working. Job loss because of illness has a particularly devastating effect on workers who support themselves and on families where two incomes are necessary to make ends meet or where a single parent heads the household.”); id. at 12 (example of Frances Wright); see also H.R.Rep. No. 101-28(1), at 23 (1990) (“The temporary medical leave requirement is intended to provide basic, humane protection to the family unit when it is most in need of help. It will also help reduce the societal cost born by government and private charity.”). This concern clearly goes to Congress’s power under the Commerce Clause and not Section 5.
A secondary motivation that appears in the legislative history is a concern to protect workers who were temporarily disabled by serious health problems from discrimination on account of their medical condition. See S.Rep. No. 103-3, at 12 (citing testimony of Ms. Barbara Hoffman, Vice President of the National Coalition for Cancer Survivorship, stating that a *13quarter of all cancer survivors face “some form of employment discrimination” and that “such discrimination against qualified employees costs society millions of dollars in lost wages, lost productivity and needless disabilities payments”); H.R. Rep. 101-28(1), at 23 (“[A] worker who has lost a job due to a serious health condition often faces future discrimination in finding a job which has even more devastating consequences for the worker and his or her family.”); see also Kazmier v. Widmann, 225 F.3d 519, 527 n. 31 (5th Cir.2000). Garrett has effectively disposed of that disability rationale as sufficient basis to overcome Eleventh Amendment immunity.
Nevertheless, the United States and Laro contend that the prevention of gender-based discrimination also motivated Congress in enacting the personal medical leave provision of the FMLA, and hence that provision does validly abrogate the states’ Eleventh Amendment immunity. They suggest that the provision counters gender-based discrimination in two respects: first, it prevents discrimination directly on the basis of pregnancy; and, second, it undermines incentives to discriminate against women indirectly on the basis of stereotypes about women due to pregnancy.9 These claims are not irrational. Nonetheless, where the connection between the provision at issue and gender-based discrimination is indirect at best, it is incumbent on Congress either to establish a clear link to the prevention of unconstitutional gender discrimination or to identify problematic state practices to which the provision responds. Here there is no indication that Congress found such a problem on the part of states as employers.
The United States and Laro advance the following argument, attempting to connect the personal medical leave provision to the need to combat gender discrimination on the part of state employers. Employers might assume that women will require greater accommodation at work as a result of pregnancy-related disability, and therefore be more reluctant to hire women or place them in positions of equal responsibility. The personal medical leave provision, they say, by affording medical leave for any “serious health condition,” mitigates this incentive to discriminate because it places men and women on equal footing with respect to medical leave. See S.Rep. No. 102-68, at 35 (1991) (“Because the bill treats all employees who are temporarily unable to work due to serious health conditions in the same fashion, it does not create the risk of discrimination against pregnant women posed by legislation which provides job protection only for pregnancy related disability.”). There is no showing, however, that establishes any nexus between gender-neutral medical leave for one’s own health conditions and *14the prevention of discrimination on the basis of gender on the part of states as employers.
We understand the problem in these terms. In 1976, the Supreme Court held that discrimination on the basis of pregnancy was not gender-based discrimination within the meaning of Title VII. See General Elec. Co. v. Gilbert, 429 U.S. 125, 97 S.Ct. 401, 50 L.Ed.2d 343 (1976).10 Congress reacted to this decision, amending the definition section of Title VII such that its prohibition of discrimination on the basis of sex encompassed pregnancy discrimination. See Pregnancy Discrimination Act, Pub.L. No. 95-555, 92 Stat. 2076, 42 U.S.C. § 2000e(k); see also Newport News Shipbuilding and Dry Dock Co. v. EEOC, 462 U.S. 669, 676, 103 S.Ct. 2622, 77 L.Ed.2d 89 (1983). . Congress’s conclusion that gender and pregnancy were sufficiently related that barring pregnancy discrimination served the end of preventing gender-based discrimination was quite rational; indeed, as the Supreme Court recently suggested, the potential for pregnancy is one immutable characteristic which distinguishes men from women and consequently has definite real life consequences. See Nguyen v. INS, 531 U.S. -, 121 S.Ct. 2053, 2061, 150 L.Ed.2d 115 (2001). In enacting the PDA, Congress was expressly concerned with the issues of gender-based discrimination now advanced in defense of subsection (D) of the FMLA. See H.R.Rep. No. 95-948, at 6-7 (1978), U.S. Code Cong. & Admin. News at 4749, 4754-55 (“[T]he consequences of [ ] discriminatory employment policies on pregnant women and women in general has historically had a persistent and harmful effect upon their careers. Women are still subject to the stereotype that all women are marginal workers. Until a woman passes the child-bearing age, she is viewed by employers as potentially pregnant. Therefore, the elimination of discrimination based on pregnancy in these employment practices in addition to disability and medical benefits will go a long way toward providing equal employment opportunities for women.... ”).
The PDA affected gender discrimination law in two respects: (1) women can no longer be treated differently in employment because of pregnancy, childbirth, or related medical conditions (or stereotypes about the same); and (2) if an employer chooses to offer benefit programs, then those programs must cover pregnancy, childbirth, and related medical conditions. Medical insurance and leave policies, if offered to employees, have to cover pregnancy. Thus the PDA itself operates to accomplish much of what is now offered as the gender-based rationale for subsection (D) of the FMLA, since it mandates equivalent treatment of all temporarily disabled workers, including those disabled because of pregnancy-related conditions. Indeed, to the extent that states provide leave to male employees for temporary medical conditions that render them incapable of working, but do not provide similar leave to women who are pregnant (or vice ver-sa), Title VII as amended by the PDA affords a federal right to relief.
The PDA did leave a gap in its coverage: those employers who did not offer benefit packages for leave at all. The FMLA closes this gap in coverage by requiring medical leave for pregnancy.11 But there *15is nothing in the legislative history which indicates the states posed this “gap” problem. Indeed, the only direct evidence regarding the actual leave policies of public sector employees in the legislative record suggests that state employers did not fall into the “gap” left by the PDA. See The Family and Medical Leave Act of 1989: Hearings on H.R. 770 Before the Sub-comm. on Labor-Management Relations of the Comm, of Educ. and Labor, 101st Cong. 45, 51 (testimony of Gerald McEn-tee, President of the American Federation of State, County, and Municipal Employees, stating that the union had successfully negotiated leave policies for a vast number of its members with public sector employers).
The argument for subsection (D) is that, at least in theory, these two requirements created a new problem: while the issue of discrimination against pregnant women was solved by mandatory pregnancy leave, this solution had the effect of imposing the costs of these leaves on employers; this led to the concern that these costs would induce employers to respond by not hiring women who had the potential to be pregnant. This new problem, the argument goes, would be solved by requiring medical leave for all employees with serious medical problems, as subsection (D) of the FMLA does. The larger concept that special treatment of one gender can lead to discrimination is reflected in the testimony before the Senate that legislation which provides job protection only for pregnancy related discrimination created the risk of discrimination against pregnant women. See S.Rep. No. 102-68, at 35 (1991). The argument is made that employers, despite the PDA and despite the nondiscrimination in hiring provisions of Title VII, would not hire women who could become pregnant and take leave because they could be more costly.12 As to assessing whether the states as employers posed such risks,13 *16even in light of the requirements of the PDA and Title VII, we think the Eleventh Amendment requires greater information from Congress as to whether such a risk is real.
Only two references in the legislative history are relied on to link such discrimination to actual state practices. The first concerns a general statement made twice at a congressional hearing asserting that public sector leave policies do not vary much from private sector policies. See The Parental and Medical Leave Act of 1986: Hearings on H.R. 4,300 Before the Subcomm. on Labor-Management Relations and the Subcomm. on Labor Standards of the Comm, on Educ. and Labor, 99th Cong. 30; id. at 147. Perhaps this is so, but it is insufficiently informative as to what the flaws were in the states’ practices for Eleventh Amendment purposes. The second piece of information relied upon is an attachment to a House Committee Report, which simply indicates that 28 of the 50 states had not enacted state laws requiring family and medical leaves in the private sector. See H.R.Rep. No. 103-8(1), Attachment B (1993). This tells us little, if anything, about whether the states themselves had not hired women because of the risk of having the costs of pregnancy leave imposed on them, thus indicating a need for personal medical leave for all under the FMLA subsection (D) in order to avoid creating disincentives to hire women. It may be that such a problem exists, but Congress did not find that to be so. If it had, we would have a different case before us.
In order to validly abrogate the states’ Eleventh Amendment immunity, enforcement legislation must have a congruence to the constitutional evil to be prevented. Here, there is no identified link between this particular provision and any pattern of discriminatory stereotyping on the part of states as employers. On this record, the personal medical leave provision of the FMLA does not exhibit a sufficient congruence to the prevention of unconstitutional state discrimination to validly abrogate the states’ Eleventh Amendment immunity. Without more, then, these legislative responses are out of proportion to the preventive objective as to states as employers and cannot be understood to be designed to prevent unconstitutional behavior.14 Cf. City of Boerne, 521 U.S. at 520, 117 S.Ct. 2157 (requiring congruence and proportionality). But see Kazmier v. Widmann, 225 F.3d 519, 533 (5th Cir.2000) (Dennis, J. dissenting); Garrett v. University of Ala. Bd. of Trustees, 193 F.3d 1214, 1220 (11th Cir.1999) (subsequent history omitted) (Cook, J., concurring in part and dissenting in part).
To summarize, the personal medical leave provision of the FMLA, § 2612(a)(1)(D), on its face has no direct connection to preventing unconstitutional gender discrimination by state employers. Cf. Hundertmark, 205 F.3d at 1276. The arguments advanced to connect that particular provision to preventing gender discrimination by states as employers are too attenuated, in the absence of a stronger legislative record, to allow Congress to abrogate a state’s Eleventh Amendment immunity under the auspices of Section 5. Accordingly, we affirm the dismissal of Laro’s complaint. The result we reach is consistent with that of every circuit that *17has addressed the issue with regard to 29 U.S.C. § 2612(a)(1)(D) of the FMLA. See Lizzi v. Alexander, 255 F.3d 128 (4th Cir.2001); Chittister v. Department of Cmty. and Econ. Dev., 226 F.3d 223 (3rd Cir.2000); Townsel v. Missouri, 233 F.3d 1094 (8th Cir.2000); Kazmier, 225 F.3d at 527-29; Sims v. University of Cincinnati, 219 F.3d 559, 566 (6th Cir.2000); Hale v. Mann, 219 F.3d 61, 69 (2d Cir.2000); Garrett v. University of Ala. Bd. of Trustees, 193 F.3d at 1219.
While we hold that the personal medical leave provision of the FMLA does not validly abrogate the Eleventh Amendment immunity of the states as employers from private damages actions, other remedies remain. As was the case in Gairett, see 121 S.Ct. at 968 n. 9, the United States may choose to pursue its own actions against New Hampshire (or other states violating the provisions) to enforce the FMLA and recover damages. See 29 U.S.C. § 2617(b)(2) (empowering the Secretary of Labor to bring civil actions to recover damages for violations of the FMLA). Private parties may also enforce the substantive requirements of the provision against states through injunctive actions against state officials rather than through suits for money damages. See, e.g., Gairett, 121 S.Ct. at 968 n.9. Similarly, New Hampshire may voluntarily choose to provide for state employees the same privately enforceable right to FMLA personal medical leave enjoyed by most private sector employees (and many state employees, under state law), either by consenting to suit in federal court or providing an enforceable state remedy. But on this legislative record, Congress does not have the power to empower a federal court to force New Hampshire to pay damages to an employee for failing to provide such leave through a private enforcement action if the state has not consented or waived its immunity.
Affirmed. Each party to bear its own costs.
. We express no opinion on the states’ immunity from a private cause of action under the other provisions of the -FMLA. We also note that the applicability of the personal medical leave provision to private parties as a valid exercise of Congress’s Article I powers is not in question.
. States may also waive this immunity by consenting to suit, but no one argues here that New Hampshire consented to suit or waived its immunity.
. Section 5 provides in full: "The Congress shall have power to enforce, by appropriate legislation, the provisions of this article.” U.S. Const, amend. XIV.
. In deciding whether legislation is congruent to a constitutional violation, the Court has typically looked to the legislative record to determine if Congress has identified a history and pattern of unconstitutional state action, see Garrett, 121 S.Ct. at 964-67; Kimel, 528 U.S. at 90-91, 120 S.Ct. 631; Florida Prepaid Educ. Expense Bd. v. College Savings Bank, 527 U.S. 627, 640, 119 S.Ct. 2199, 144 L.Ed.2d 575 (1999), although it has suggested that such an inquiry may not be required where the constitutional wrong is otherwise evident, cf. Florida Prepaid, 527 U.S. at 646, 119 S.Ct. 2199 ("Though lack of support in the legislative record is not determinative, identifying the targeted constitutional wrong or evil is still a critical part of our § 5 calculus.”) (citations omitted); accord City of Boerne, 521 U.S. at 530-31, 117 S.Ct. 2157; see also Kimel, 528 U.S. at 91, 120 S.Ct. 631. *9Because we do not consider subsection (D), the FMLA provision at issue here, to be a matter where the targeted constitutional wrong is self-evident, we also look to the legislative record. Given this result, we need not reach the question of whether, as some of the language in Garrett suggests, the Court now considers such legislative history to be more generally required than Kimel and City of Boeme suggest. See Garrett, 121 S.Ct. at 967-68 ("[I]n order to authorize private individuals to recover money damages against the States, there must be a pattern of discrimination by the States which violates the Fourteenth Amendment, and the remedy imposed by Congress must be congruent and proportional to the targeted violation.'1); id. at 969 (Kennedy, J., concurring) ("The predicate for money damages against an unconsenting State in suits brought by private persons must be a federal statute enacted upon the documentation of patterns of constitutional violations committed by the State in its official capacity.”).
. 29 U.S.C. § 2612(a)(1), titled "Entitlement to leave,” provides:
Subject to [29 U.S.C. § 2613], an eligible employee shall be entitled to a total of 12 workweeks of leave during any 12-month period for one or more of the following:
(A) Because of the birth of a son or daughter of the employee and in order to care for such son or daughter.
(B) Because of the placement of a son or daughter with the employee for adoption or foster care.
(C) In order to care for the spouse, or a son, daughter, or parent, of the employee, if such spouse, son, daughter, or parent has a serious health condition.
(D) Because of a serious health condition that makes the employee unable to perform the functions of the position of such employee.
. The State, Laro, and the intervenor United States all argue that the entire FMLA is at issue. We disagree. The constitutional arguments in support of the remaining provisions have greater strength and raise issues (for instance, their implications for family roles) not at stake here. Those other provisions are not tested by this case. Assessing the provisions separately is not unprecedented, see, e.g., Kazmier v. Widmann, 225 F.3d 519, 525 (5th Cir.2000) ("[W]e discern no reason why the provisions of one of the FMLA's subsections could not validly abrogate the States’ Eleventh Amendment immunity even if the provisions of some or all of the remaining subsections fail to do so.”), and may be important where, as here, separate provisions of the same act embody differing remedial purposes. Moreover, the Supreme Court has suggested that different provisions of the same act might fare differently under this Eleventh Amendment analysis, in that it declined to address claims arising under Title II of the ADA in Garrett. See 121 S.Ct. at 960 n. 1.
. It is plain that, under Gan'ett, there is no congruence between the personal medical *11leave provision and congressional authority to abrogate Eleventh Amendment immunity in addressing disability discrimination under Section 5. Gairett held that the needed congruence in order to subject the state to a private damages action for disability discrimination was lacking in the Americans with Disabilities Act. See Garrett, 121 S.Ct. at 966-67. If a direct connection to disability discrimination was insufficient in Garrett, then the indirect connection in the FMLA between disability discrimination and inability to work due to serious health problems is surely insufficient to establish congruence, at least on the current legislative record. The United States concedes this point in its brief.
. The Supreme Court did consider a gender-based justification for legislation under Section 5 in United States v. Morrison, supra, assessing the private suit provisions of the Violence Against Women Act. In that case the Court concluded that the provisions of VAWA at issue went beyond Congress’s Section 5 powers on the ground that the provisions did not act to prevent unconstitutional state action, instead targeting private behavior. 529 U.S. at 626, 120 S.Ct. 1740. Since the Court found that VAWA did not target state action, it did not address the Eleventh Amendment analysis pursued in Kimel and Garrett.
. A third way in which the personal medical leave provision might possibly be justified as responding to a threat of unconstitutional gender discrimination is that it serves to eliminate the risk of discriminatory treatment of men in leave policies. While the legislative record suggests that men may have been treated in a discriminatory fashion with respect to parental leave, see, e.g., S.Rep. No. 103-3, at 14-15 (discussing studies by the Bureau of Labor Statistics highlighting the discrepancy between the availability of maternity and paternity leave), and perhaps leave to care for family members, there is no suggestion that men were disadvantaged in personal medical leave policies, nor, in particular, is there any reason to think from the legislative record that states as employers disadvantaged men in offering medical leave in a way that could be considered discriminatory. Indeed, men are protected from inequitable leave policies offered by either private or public employers in any case by Title VII's prohibition of discrimination in employment on the basis of sex.
. The Supreme Court had earlier held that cost-justified discrimination on the basis of pregnancy was not unconstitutional gender discrimination under the Fourteenth Amend-menl. See Geduldig v. Aiello, 417 U.S. 484, 94 S.Ct. 2485, 41 L.Ed.2d 256 (1974).
. In the FMLA legislative history, Congress discussed the gap left by the PDA with respect *15to employers who denied benefits to all workers. See H.R.Rep. No. 103-8(11), at 11 (1993).
. That risk would seem to be greatest, at least in theory, among employers who had escaped the PDA's leave requirements (because they offered no leave to anyone) but who were now caught by the FMLA’s requirement giving pregnancy leave. But as we noted before, there is no evidence in the legislative record that the states present a "gap” problem.
. There are variations on the argument that neutral leave provisions run the risk of creating incentives to discriminate. One variant (though not raised directly by Laro or the United States) distinguishes between concern for women as potential takers of pregnancy leave and concern for women as likely takers of leave for family caretaking. The contention begins with a fear that employers will be less likely to hire or promote women because they believe that women will be more likely than men to take advantage of the family leave provisions in 29 U.S.C. § 2612(a)(1)(B) and (C). The fact that both men and women are entitled to family leave, thus achieving desired neutrality, only solves part of the problem, the argument goes. Either because of stereotypes or because women actually take leave for family caretaking more often, the claim is, these provisions of the FMLA (by increasing the projected costs of women employees) create new incentives not to hire or promote women. Those incentives are powerful enough, the argument proceeds, to override the incentives not to discriminate provided by Title VII and the PDA. In consequence, according to this theory, the number of men taking leave must be increased through the creation of the personal medical leave provision in subsection (D). This undercuts any presumption that women will be more likely to take leave than men.
This theory as to the effects of the new family leave provisions on the future behavior of employers does not appear to be what concerned Congress in enacting the personal medical leave provision, however. Certainly Congress never made any findings to this effect with regard to states as employers. Without more, this theory does not suffice.
. We pause to reject an argument made by the State. The State suggests that only intentional acts of discrimination may ever be reached by a federal statute under the enforcement clause. We do not read the Elev-enlh Amendment jurisprudence that way, and the argument contravenes the rule that Congress is free under Section 5 to act prophylac-tically as well as remedially. E.g., Kimel, 528 U.S. at 81, 120 S.Ct. 631.
. I acknowledge that Congress's adoption of the personal medical leave provision also reflected concerns for economic dislocation and discrimination related to medical conditions. Those additional concerns do not detract from the significance of Congress’s concern for gender discrimination.