The Toughest Job: Adkins v. Rumsfeld, Gender, Incentives, and the Uniformed Services Former Spouse' Protection Act "THE TOUGHEST JOB": ADKINS V. RUMSFELD, GENDER, INCENTIVES, AND THE UNIFORMED SERVICES FORMER SPOUSES' PROTECTION ACT BRIAN R. DECKER* Not much is written about Jessica Miller. She was a student of psychology, was born in 1982 or early 1983,1 and spent at least part of her childhood in Pike County, Kentucky.2 The last part is assumed because she is said to have married her "childhood sweetheart" and Pike County is the childhood home of her ex-husband, James Blake Miller (Blake).3 Blake has been the subject of much more publicity: a 2004 Los Angeles Times photograph captured a close-up of him: a Marine weary in the midst of the battle for Fallujah, Iraq.4 In the picture, a battered combat helmet frames a face caked with dirt and blood, Blake's eyes are locked in the classic soldier's thousand-yard stare, and a cigarette dangles from his lower lip. When the photo made the pages of a hundred U.S. newspapers and the cover of Time Magazine, Blake became an icon nicknamed "the Marlboro Man."5 After he returned from his tour, Jessica helped diagnose him with post-traumatic stress disorder-which is the only reason the public knows she studied psychology. 6 * Associate, Dechert LLP; J.D. 2007, University of Pennsylvania Law School; A.B. 2004, Stanford University. Thanks to Jennifer Finney, Elizabeth Kimmelman, Carolyn Quam, Amy Wax, and Jeffrey Weil for their help with this article. Any errors are the author's alone. 1 See Matthew B. Stannard, Stressed-out 'Marlboro Man' Files for Divorce, S.F. CHRON., June 27, 2006, at A2. 2 See Matthew B. Stannard, The War Within, S.F. CHRON., Jan. 29, 2006, at A2. 3 1d. 4 1d. 5 Paul Harris, A Soldier's Story, OBSERVER (London), July 2, 2006, at 18. 6 Stannard, supra note 2, at A2; accord David Zucchino, Iconic Marine Is at Home but Not at Ease, L.A. TIMES, May 19, 2006, at Al. Columbia Journal of Gender and Law Nor is much ink devoted to Patricia Ann McCarty. As with Jessica Miller's, everything publicly known about McCarty's life is spelled out between the lines of the career of her more noted ex-husband, Richard John McCarty. They were married in 1957 while he was a second-year medical student.7 Richard became an Army doctor two years later and, over the ensuing seventeen years, followed his orders to Pennsylvania, Hawaii, the District of Columbia, California, Texas, and eventually San Francisco's Presidio Military Reservation, where he became Colonel and Chief of Cardiology.8 Patricia presumably followed. They had at least three children together. 9 Unlike Jessica Miller and Patricia McCarty, Tammy Adkins has not been overshadowed in the media by her husband. Adkins enlisted in the Air Force in 197810 at the age of seventeen. 1 She reached the rank of Technical Sergeant,12 serving in Sacramento, Wyoming, New Mexico, and overseas 3 before retiring on disability in 1999.14 In the early 1980s, she married Roland Wachter, with whom she had a son, Robert, and they lived together in Sacramento and abroad 15 before divorcing in 1994.16 Tammy then married Alvin Ray Adkins, 17 but they too divorced in May 2007.18 7 See McCarty v. McCarty, 453 U.S. 210, 216 (1981). 81d. 9 1d. at 218. 10 Report of U.S. Military Locator, accessed Oct. 9, 2007 (on file with author). 1 See Accurint Comprehensive Public Records Report, accessed Oct. 9, 2007 (on file with author) [hereinafter Public Records Report]. 12 See Report of U.S. Military Locator, supra note 10. 13 See Public Records Report, supra note 11 (including a list of past mailing addresses, among which is an APO address). 14 Complaint for Declaratory Judgment 7.a, Adkins v. Rumsfeld, 370 F. Supp. 2d 426, 428 (E.D. Va. 2004) (No. 04-494), aff'd, 464 F.3d 456 (4th Cir. 2006), cert. denied sub nom. Adkins v. Gates, 127 S. Ct. 2972 (2007) [hereinafter Complaint]. 15 See Public Records Report, supra note 11 (demonstrating that a Roland Wachter, who is a few months older than Tammy and must be the subject of the divorce underlying the Adkins case, and a Robert Wachter, who was born in the late 1980s, shared addresses with Tammy). 16 Complaint, supra note 14, at T 7.a. 17 Public Records Report, supra note 11. 246 [Vol. 17:2 2008] "The Toughest Job" 247 James Blake Miller filed for divorce from Jessica in June 2006,19 thirty years after Richard and Patricia McCarty did the same.20 The McCartys' case went to the Supreme Court on the issue of whether the retirement benefits the Army had conferred on the Colonel could be considered "quasi-community property" under California divorce law so that half of the benefits would go to Patricia.2 1 The Court ruled that, because of a conflict with federal law, it could not,22 prompting Congress to answer with the Uniformed Services Former Spouses' Protection Act (USFSPA), which specifically granted courts the authority to treat military retirement pay as the property of both service member and spouse upon divorce.23 The USFSPA became the subject of litigation in 2004 when a group of fifty-eight divorced retired veterans and a nonprofit company filed suit in Virginia to challenge the constitutionality of certain portions of the Act.24 The lead plaintiff in the suit against the Secretary of Defense was Tammy Adkins, whose disability benefits were reapportioned to Roland Wachter under the Act.25 The complaint in Adkins v. Rumsfeld alleged violations of due process and equal protection rights.26 18 Public Record, CLOVIS NEWS J. (N.M.), May 25, 2007, available at http://www.cnjonline.com/news/clovis21634 article.html/michaelboseman.html. 19 Stannard, supra note 1, at A2. 20 McCarty v. McCarty, 453 U.S. 210, 216 (1981). 21 Id. at 211-18. 22 Id. at 232-36. 23 Pub. L. No. 97-252, tit. X, sec. 1002(a), § 1408(c)(1), 96 Stat. 730, 731-32 (1982) (codified as amended at 10 U.S.C. § 1408(c)(1) (2000)). Property division under the Act is subject to certain limitations: Notably, the service member must have served for ten or more years while married to the spouse or former spouse to whom payments are to be made. 10 U.S.C. § 1408(d)(2). The Act also places a fifty percent maximum on the portion of the retirement pay redistributed to former spouses. § 1408(e)(1). 24 Adkins v. Rumsfeld, 370 F. Supp. 2d 426, 428 (E.D. Va. 2004), aff'd, 464 F.3d 456 (4th Cir. 2006), cert. denied sub nom. Adkins v. Gates, 127 S. Ct. 2972 (2007). The plaintiffs filed suit at the federal courthouse in Alexandria, Virginia, not far down the George Washington Parkway from the Marine Corps War Memorial, which is based on another iconic battlefield photograph taken on Iwo Jima almost sixty years before James Blake Miller became the "Marlboro Man." Nat'l Park Serv., U.S.M.C. War Memorial, http://www.nps.gov/archive/gwmp/usmc.htm (last visited Jan. 28, 2008). 25 See Complaint, supra note 14, at T 7.a. 26 Adkins, 370 F. Supp. 2d at 428. Columbia Journal of Gender and Law The equal protection claim was particularly nuanced. The veterans argued that the USFSPA "discriminates against women in the Armed Forces and in favor of men because former husbands are more likely than former wives to have sources of income other than the divided military retirement pay."' 7 The presence of women in the military in numbers much more substantial than when Congress passed the Act in 1982 N-women who often have non-service member husbands with private jobs and private pensions29-"essentially reverses the polarity of the gender assumptions that Congress employed in legislating the statute., 30 Adkins was the lead plaintiff by an accident of alphabetical order, but her story, as one of only seven female service members out of fifty-nine individual plaintiffs, turned the McCarty paradigm of male service member and female homemaker on its head and made for the most-mentioned example in the media. 31 On appeal, the Fourth Circuit rejected the equal protection claim.32 The court was unconvinced by "isolated statements in the legislative history" using gendered language "because the bulk of the congressional materials used sex-neutral terms. 33 The court further rejected the veterans' due-process claims and affirmed the district court's dismissal on summary judgment.34 The Supreme Court denied certiorari.35 These three women-Miller, McCarty, and Adkins-have had three very different lives. The student, the homemaker, and the sergeant 27 Adkins, 464 F.3d at 468. 28 See infra note 96 and accompanying text. 29 See Alicia H. Munnell with Natalia Zhivan, Earnings and Women's Retirement Security 11 (Ctr. for Retirement Research at Boston Coll., Working Paper No. 3, 2006), available at http://www.bc.edu/centers/crr/papers/wp_2006-12.pdf (reporting data that suggest that fewer retired women have pensions than retired men and that women's pensions tend to be smaller). 30 Brief of Appellants at 52, Adkins v. Rumsfeld, 464 F.3d 456 (4th Cir. 2006) (No. 05-2307) [hereinafter Appellant Brief]. 31 See, e.g., Tom Philpott, Ex-Spouses Challenge Law, SuN HERALD (Biloxi, Miss.), June 12, 2004, at A8 (mentioning Adkins's as the only example of a plaintiffs experience with divorce). 32 Adkins, 464 F.3d at 470. 33 Id. at 468. 34Id. at 473. 35 Adkins v. Gates, 127 S. Ct. 2972 (2007). [Vol. 17:2 "The Toughest Job" encountered the lifestyle of three different branches of the U.S. military in different decades and in different states across the country. Yet they share the common experience of military marriage and divorce, of the complex relationship between home life and life in the armed services, and of the problem of dissolving a marriage in which one member is a veteran. This Article examines this problem, along with the equal protection argument from Adkins, through the lens of the gender assumptions behind the legislative history of the USFSPA. The Article further presents an economic analysis of the Act's policy goals. Part I critiques the enacting Congress's notions of gender, family, and work in the military. Part 1I applies equal protection jurisprudence and concludes that the court was right to dismiss the claim. Finally, Part III lays out the economic framework for deciding, constitutional claims aside, whether the USFSPA accomplishes its policy goals, both explicit and implicit. In the process, the Article will shed some light on the military marriage, a relationship society has come to view as extraordinary, and explore what military home life means for women. I. HOW THE USFSPA EXPRESSES CONGRESS'S VIEW OF THE MILITARY WIFE The USFSPA couches its language in primarily gender-neutral terms. The key provision states, "a court may treat disposable retired pay payable to a member . . . either as property solely of the member or as property of the member and his spouse in accordance with the law of the jurisdiction of such court. ' 36 "His" is "used in a generic sense or when the sex of the person is unspecified, 37 or so Congress claims; elsewhere the U.S. Code clarifies that "unless the context indicates otherwise ... words importing the masculine gender include the feminine as well. 38 Rather than "serviceman" and "wife," Congress chose "member" and "spouse." "Spouse," according to section 1408, "means the husband or wife ... of a member. 3 9 36 10 U.S.C. § 1408(c)(1) (2000). 37 MERRIAM-WEBSTER'S COLLEGIATE DICTIONARY 573 (1 1th ed. 2004) (defining "he"); id at 589 (defining "his" by reference to "he"). 31 1 U.S.C. § 1 (2000). 39 10 U.S.C. § 1408(a)(6) (2000) (emphasis added); accord S. REP. No. 97-502, at 15 (1982), reprinted in 1982 U.S.C.C.A.N. 1596, 1610 (analyzing the definition section of the Act). 20081 250 Columbia Journal of Gender and Law [Vol. 17:2 In contrast, congressional hearings, reports, and debates indicate that Congress almost universally understood traditional gender roles to factor into the statute. To Congress in 1981-82-informed by the example of the McCarty case, which Congress explicitly superseded by enacting the USFSPA 40--a member of the armed services was a man and his spouse a woman, specifically a housewife. The ideal-and, as many members believed, the reality-was the male soldier who was capable of extended tours away from home because of his homemaker wife who supported him on the domestic front. A. Hearings and the Committee Report In opening the first Senate hearing on the bill that would eventually develop into the USFSPA, Senator Roger Jepsen, Chairman of the Subcommittee on Manpower and Personnel, framed the debate around the ideal of a male soldier with a homemaker wife: One of the strongest points made to a military spouse from the time her husband chooses to join the armed services is her role as a partner in her husband's career. She is expected to fulfill an important role in the social life and welfare of the military community to care for the children and support her service member spouse as they share the rigors of military life.41 Senator Mark Hatfield, who submitted one of the bills aimed at counteracting McCarty, emphasized "the traditional family unit, composed as husband as breadwinner and wife as homemaker," which "has had a very dramatic change within a 20-year period. ' 42 To Hatfield, the traditional 40 See S. REP. No. 97-502, at 1 (1982), reprinted in 1982 U.S.C.C.A.N. 1596, 1596 ("The primary purpose of the bill is to remove the effect of the United States Supreme Court decision in McCarty v. McCarty." (citation omitted)). 41 Uniformed Services Former Spouses Protection Act: Hearings on S. 1453, S. 1648, and S. 1814 Before the Subcomm. on Manpower and Personnel of the S. Comm. on Armed Servs., 97th Cong. 12 (1982) [hereinafter USFSPA Hearings] (statement of Sen. Jepsen, Chairman, Subcomm. on Manpower and Personnel). 42 Id. at 14 (statement of Sen. Hatfield); cf id. at 31 (prepared statement of Rep. Hance) ("[The military wife's] husband's transience--often changing assignments and locations as frequently as once or twice a year-makes her a risk with potential employers, even if she has the necessary qualifications."); id. at 123 (remarks of Sen. Cohen) ("[lit is particularly difficult because you have got in effect a nomadic existence which many other wives do not have to confront ... and ... people in the service are required to move from place to place and.., the wives have to go with them .... That doesn't exactly lend stability to a potential employer."). "The Toughest Job" dynamic seemed more likely in the military because "[f]requent moves during military service prevent a military wife from establishing a career and earning accompanying retirement benefits. ' ' 3 The hearings brought out evidence of the importance of the "military wife" to the career of her husband. According to a statement from Suzanne Davis of the National Military Wives Association, officers testified to "the role their wives played in volunteer work, social functions and contributions to their husband's careers [being] routinely mentioned in their Officer Evaluation Reports. 'A4 Senator William Cohen would later cite specific examples of such reports. In one, despite the service member having "[t]echnical competence of the highest order," an evaluating officer noted that "[h]e and his wife did not try to maintain a previously active group of company officers' wives. 45 In another, an officer "criticize[d] [the member's] wife's antisocial actions which led to the disintegration of a group of wives who had been making significant contributions to the company and military community" and, "therefore, [did] not recommend him for future assignments which will be affected by her behavior. 'A 6 Representative Patricia Schroeder, sponsor of one of the bills that would ultimately form the Act, 47 brought a Navy commissary grocery bag as an example of the significance the military attributed to its members' family units as cohered by the homemaker-wife (and the paltry form of recognition with which it was appreciated): "[the bag] says, 'The Navy wife, the toughest job in the Navy, we want to keep the good families in.' 4 8 Even those who testified in opposition to the bills often used language that revealed underlying gendered ideas of family and work 41 Id. at 15. 44 Id. at 56 (prepared statement of Suzanne Davis, Nat'l Mil. Wives Ass'n); see also JAMES HOSEK ET AL., MARRIED TO THE MILITARY: THE EMPLOYMENT AND EARNINGS OF MILITARY WIVES COMPARED WITH THOSE OF CIVILIAN WIVES 3-4 (2002). 45 USFSPA Hearings, supra note 41, at 201-02 (remarks of Sen. William Cohen citing unnamed "personal evaluation reports"); cf id. at 203 (statement of Lt. Gen. Edward J. Bronars, U.S. Marine Corps, Deputy Chief of Staff for Manpower) ("[G1n frequent occasions I see fitness reports remark on the contributions made by the wife. I have never seen a fitness report that remarked on the negative aspects of perhaps her behavior. We are talking more and more in terms of the husband-wife team."). 46 Id at 201. 4' H.R. 3039, 97th Cong. (1981). 48 USFSPA Hearings, supra note 41, at 26 (statement of Rep. Schroeder). 2008] 252 Columbia Journal of Gender and Law [Vol. 17:2 dynamics. John Sheffey of the National Association for Uniformed Services expressed the view that "it is fundamentally wrong to establish a direct channel of responsibility from the armed services to a divorced spouse" because "[t]he wife of a service member does not work for his service-she works for him.", 49 While much of the language at the hearings from both members of Congress and those invited to testify referred to "spouses" and left gender aside to some extent, there are numerous instances where the use of gender- specific pronouns and the words "husband" and "wife" indicate that the agenda's underlying presumption was a relationship between a male service member and his nonmember wife. In four days and over two hundred pages of testimony, a few scant paragraphs, amounting to what could not have been more than a couple minutes on record, explicitly acknowledge the perspective of the female service member and her husband.5° 49 Id. at 76 (prepared statement of John P. Sheffey, Executive Vice President, Nat'l Ass'n for Uniformed Servs.). 50 In his testimony, Major General William L. Webb Jr., the Army's Assistant Deputy Chief of Staff for Personnel, provided an anecdote with this gender dynamic: An officer who recently spent 18 months in the area of enlisted separations and 18 months in officer retirements wrote the Department of the Army a little resume. Her views: The number of military careers ended or derailed because spouses literally ceased to support the military member were too numerous to count. I honestly believe that if Congress concludes that all spouses are entitled to a percentage of military retired pay by virtue of holding a marriage license for 10 years, the exodus from the Army prior to 10 years of service and/or 10 years of marriage could be devastating. She also points out an important perspective and it is a different one. She feels it might well be unfair to female career soldiers. This is a new aspect for all of us in the service and a growingly important one. She says if she has tried to balance a married life with a military career and it does not work out after 10 years, she would owe her ex-spouse forever for not supporting her career in lieu of his. Id. at 196 (statement of Maj. Gen. William L. Webb Jr., Assistant Deputy Chief of Staff Personnel, Dep't of the Army). "The Toughest Job" The committee report, polished for publication, uses less overtly gendered language than did the hearings, but it indirectly maintains the same assumptions of gender roles. A section on the "Military Spouse" stresses the importance and reality of the homemaker partner: A recurrent recruiting point that is made to a military couple from the time of the spouse's initial entry into the military is that the spouse is a partner in the member's career. The theme of the "military family" and its importance to military life is widespread and well publicized. Military spouses are still expected to fulfill an important role in the social life and welfare of the military community. Child care and management of the family household are many times solely the spouse's responsibility. The military spouse lends a cohesiveness to the family facing the rigors of military life, including protracted and stressful separations. The committee finds that frequent change-of-station moves and the special pressures placed on the military spouse as a homemaker make it extremely difficult to pursue a career affording economic security, job skills and pension protection.51 By way of policy justification, the report also notes, "[In divorce proceedings,] pensions are typically awarded to the worker, with an offsetting monetary award to the other spouse. . . .[S]ince men are more likely than women to hold jobs that allow them to acquire pensions, they are also more likely to be awarded those pensions at divorce. 5 2 B. The Emergence of Idealist and Realist Narratives at the Floor Debates Congressional floor debates on the USFSPA, however, resumed the more gender-infused justifications of the hearings. The justifications for the bill fell roughly into two distinct models, idealist and realist, both of which were based on traditional gender roles. The idealist justification for the USFSPA was that it would allow homemaker wives to feel secure that they could forego a career outside the home to support their service-member husbands in the traditional marriage paradigm. Under the idealist storyline, the traditional family roles of " S. REP. No. 97-502, at 6 (1982), reprinted in 1982 U.S.C.C.A.N. 1596, 1601. 52 Id. at 8, reprinted in 1982 U.S.C.C.A.N. at 1603 (quoting Lenore J. Weitzman, The Economics of Divorce: Social and Economic Consequences of Property, Alimony and Child Support Awards, 28 UCLA L. REv. 1181, 1209 (1981)). 2008] Columbia Journal of Gender and Law breadwinner husband and homemaker wife were goals worth preserving or striving for. Representative Schroeder lamented that, though society "encourage[s] women to stay home and take care of children," it lets them down by "fail[ing] to attach an economic value to the contribution of the homemaker and fail[ing] to protect her economically after the end of a long marriage." 53 One of the Act's cosponsors, Representative George Whitehurst, announced, "This is not a women's lib issue. It is an issue that strikes at the very heart of the traditional family concept . . . . ,54 Representative Don Clausen's story reported the effects of the McCarty rule on his constituent, a woman from Napa who was married for thirty-one years to a member of the army: During those 31 years, she devoted her life to her husband's career and the raising of their two children, who are now grown and reside in States other than California. She considered this her life's work, and faithfully committed her time and energies to her family, looking, in time, to the day when she and her husband would be able to settle down in retirement, with his military pension as a means of support. However, her faithfulness and devotion to family were not matched by her husband, who, in the 30th year of their marriage, began seeing another woman and started down the road toward alcoholism. Given these unhappy circumstances, she was forced to divorce him. As California is a community property State, the settlement required that all of the couple's assets, excluding his pension, be divided equally between them. As their sole asset consisted of their home, they were forced to sell it and divide the proceeds from the sale. Left only with these funds, my constituent had to shift for herself. She filed for supplemental security income (SSI) but was denied, informed that her assets had not fallen to the required level. Reaching retirement age, she has no social security benefits on 53 128 CONG. REC. 18315 (1982) (remarks of Rep. Schroeder). 541d. at 18320. 254 [Vol. 17:2 "The Toughest Job" which to rely, only a small amount of public assistance once her assets had dwindled. 55 Clausen's implication was clear: if the USFSPA became law and wives could share in military retirement pay, women like this Army wife could be secure as homemakers knowing that they would be taken care of in retirement, alongside their husbands or without them. This would alleviate some of the incentive for these women to work outside the home and, in turn, bolster the traditional breadwinner-homemaker dynamic. The realist narrative, on the other hand, approached the traditional gendered dynamic as simply a state of affairs in American society and, reserving judgment on whether it was normatively good, aimed at increasing fairness within that paradigm. Representative Donald Mitchell's statement exemplified this view: "[Divorced military spouses] play[ed] a significant role in their husbands' careers. . . . [V]ery frequently, Mr. Chairman, they have had to act as both parents when their husbands were away-had to be both father and mother to the children as they were growing up." 56 Representative Whitehurst matched his idealist justification with a realist story about a lady that came to see him: She was about 50 years old. She had been married for about 30 years. Her husband had been in the Navy a comparable length of time. He was getting out and, in leaving the Navy, he decided to make for himself a whole new life, and she was not included in it. So, he dropped her. He married a woman half his age. He bought himself a convertible, she said, with a flame down the side, and he dyed his hair. I said, "I think he has a case of male menopause." She said, "Well, what happens to me, because when I married him I wanted to go to school and improve my education." [sic] He said, "No, you stay home and take care of the children," so I have not been trained to do anything. I have no work experience, and here I am at the age of 50 and my health is not good. I cannot use the naval hospital in this district. I cannot use the commissary. I cannot use any of the facilities and I get no support from him. For 30 years I was married to this man and I took all 55 Id. at 18332 (remarks ofRep. Clausen). 56 Id. at 18328. 20081 255 Columbia Journal of Gender and Law the sacrifices of separation and trauma that go with being a Navy wife. Is this fair?"' 57 The disenchanted Navy ex-wife who visited Representative Whitehurst had wanted to develop an independent career outside the home, but she had been convinced-by her husband, by the military, or by society as a whole-that it would be better to focus on rearing children and maintaining a household. Out of simple fairness, the realist argument asserted, we must we compensate her for her sacrifice. But like the idealist argument, the realist argument was grounded in the concept of serviceman and housewife as the central relationship the Act would address. Taken as a whole, the legislative history of the USFSPA suggests that legislators understood the McCarty problem to be, largely, Patricia McCarty's problem: the displaced housewife was left with no claim to a portion of her husband's military retirement regardless of whether her state's divorce law would have otherwise provided her with one. This concept encompasses assumptions that are both gendered-the husband is the serviceman and the wife is not-and economic-the non-service- member spouse is likely to be a homemaker or otherwise have no pension of her own. The Adkins plaintiffs sought to present the factual antithesis of these assumptions. II. THE ADK!NSEQUAL PROTECTION CLAIM In Adkins v. Rumsfeld, the Fourth Circuit reviewed this legislative history only cursorily, construing it to be inconclusive regarding gender- discriminatory purposes. Accordingly, the court applied rational basis review to the equal protection claim and upheld the USFSPA. Had the court more thoroughly engaged the legislative history, it likely would have found covert gender discrimination and applied intermediate scrutiny. Though such an analysis would have been more complex, it too would likely uphold the Act. Tammy Adkins and her co-plaintiffs brought a multifaceted equal protection claim against the Department of Defense alleging three distinct classifications. Two of these classifications, between "uniformed service members" and "other groups of former federal government employees" and between "the spouses (whether men or women) of former uniformed service members" and the members themselves,58 did not implicate gender or any 57 Id. at 18319 (remarks of Rep. Whitehurst). 58 Complaint, supra note 14, at 22; accord Appellant Brief, supra note 30, at 47- [Vol. 17:2 "The Toughest Job" other suspect or quasi-suspect classification and hence were automatically subject to rational basis review. 59 The third alleged that "the former spouses of uniformed servicemembers, a group largely defined as women by Congress (in enacting the statute in 1981 [sic]), are entitled to a higher degree of income and property protection as would be accorded in relation to the former spouses of other classes of federal government employees. ' 60 The coded gender classification, according to this argument, is military spouses (mostly women) versus other government spouses (not as predominantly women). What seems to be a fourth theory of equal protection violation, also gender-based, arose in the Reply Brief: "the original congressional assumptions in legislating a statute that today puts female veterans at a disadvantage in relation to their non-servicemember spouses. '61 It is this classification-women service members versus men, either their spouses or their male counterparts in the services-that the Fourth Circuit saw as the veterans' primary claim for sex discrimination. 62 A. Where the Fourth Circuit Erred The Fourth Circuit analyzed the Adkins claim according to the holding in Personnel Administrator of Massachusetts v. Feeney.63 The Feeney plaintiffs had a claim similar to those in Adkins: A group of nonveteran women who had been denied civil-service positions because of a Massachusetts veterans'-preference statute sued the state alleging gender discrimination. 64 Since veterans tended overwhelmingly to be men, the argument went, the statute disproportionately burdened women. 65 The court articulated a test for determining whether a facially gender-neutral statute is nevertheless sex-discriminatory, hence subject to intermediate scrutiny: 59 See Adkins v. Rumsfeld, 464 F.3d 456, 468 (4th Cir. 2006). 60 Complaint, supra note 14, at 22; accord Appellant Brief, supra note 30, at 48. 61 Reply Brief of Appellants at 26, Adkins v. Rumsfeld, 464 F.3d 456 (4th Cir. 2006) (No. 05-2307), cert. denied sub nom. Adkins v. Gates, 127 S. Ct. 2972 (2007) [hereinafter Reply Brief]. 62 See Adkins v. Rumsfeld, 464 F.3d 456, 468 (4th Cir. 2006), cert. denied sub nom. Adkins v. Gates, 127 S. Ct. 2972 (2007). 63 442 U.S. 256 (1979). 64 Id. at 259. 65 Id. at 260. 2572008] Columbia Journal of Gender and Law "The first question is whether the statutory classification is indeed neutral in the sense that it is not gender-based. If the classification itself, covert or overt, is not based upon gender, the second question is whether the adverse effect reflects invidious gender-based discrimination. 66 The Adkins court, reasoning that the USFSPA, like the Massachusetts statute in Feeney, was gender-neutral on its face, followed Feeney in testing for two characteristics, each of which, had it gone in favor of the veterans, would have called for intermediate scrutiny. 67 First the court examined whether there was covert sex discrimination, determining that, despite "isolated statements in the legislative history describing the Act as concerned about service members' wives," such evidence was "inconclusive because the bulk of the congressional materials used sex- neutral terms, speaking of spouses and not wives. '68 The second theory also broke against the veterans, who "d[id] not and cannot plausibly allege that Congress passed the Act 'at least in part "because of," not merely "in spite of," its adverse effects' upon female service members." 69 The thrust of the plaintiffs' argument, according to the Fourth Circuit, was "that Congress did not consider the plight of female service members., 70 To the court, this cut against them; Congress could not have intended invidious discrimination against a group it did not consider.7 1 Since it held that there was no gender discrimination, the court applied rational basis review: Congress could have properly concluded that those sacrifices were even more intense than the ordinary sacrifices associated with marriage to civilian employees, and that former spouses of service members thus deserved additional protection not afforded to former spouses of civilian employees. While not the sole purpose, former spouse protection was a legitimate purpose animating the Act. 7 2 61 Id. at 274. 67 Adkins, 464 F.3d at 468-69. 68 Id. at 468. 69 Id. at 468-69 (quoting Feeney, 442 U.S. at 279). 70 ld. at 469. 71 id [Vol. 17:2 "The Toughest Job" After shooting down the Department's arguments that military retirement pay is more like a retainer and that equal protection demanded a remarriage cutoff as with former spouses of other government employees, the court held the USFSPA not violative of equal protection.73 The weak point in the court's reasoning was its holding that "isolated statements in the legislative history" were "inconclusive" as to whether Congress was covertly discriminating based on sex. 74 The overwhelming evidence suggests that, even as members of Congress used neutral language, they understood service members to refer almost exclusively to men and spouses to refer to women. While it is true that Congress may have been using its gendered pronouns and narratives as examples of an experience that could happen to both men and women (and, to the extent that the military was predominantly male and members' spouses predominantly female, realistic examples at that),75 the equal protection guarantee is not a question of whether lawmakers sought to be nondiscriminatory. The inquiry is whether the classifications were truly based on gender, and on that point the legislative history makes Congress's assumptions clear. With few exceptions, each time the representatives, senators, and outside authorities did use gendered language, they referred to the male-breadwinner and female-homemaker dynamic and not its reverse. It seems almost naYve for the court to deny that Congress loaded spouse with a gendered meaning. B. Were the Real Justifications for the USFSPA Constitutional? Had the court found covert discrimination based on sex, it would have applied intermediate scrutiny rather than rational basis review. Califano v. Webster guides this alternative analysis.76 Califano confronted a Social Security Act provision that computed benefits differently for women and men. 77 Unlike the USFSPA, the Califano provision overtly utilized a gender distinction.78 The Supreme Court reasoned that the provision could 71 Id. at 469-70. 74 Id at 468. 75 See infra note 96 and accompanying text. 76 Califano v. Webster, 430 U.S. 313 (1977) (per curiam). 77Id. at 314. 2592008] Columbia Journal of Gender and Law nevertheless pass muster if it were "substantially related" to an "important governmental objective[].,,7 9 The question was whether the objective was permissible. On the one hand, laws that are the "result of 'archaic and overbroad' generalizations about women or of the 'role-typing society has long imposed' upon women, such as casual assumptions that women are 'the weaker sex' or are more likely to be child-rearers or dependents" do 80not have legitimate objectives. On the other hand, "'redressing our society's longstanding disparate treatment of women"' is an important objective.81 The legislative history of the USFSPA demonstrates elements of both these justifications. Congress largely assumed that women would be the target spouses and further generalized that they would take on the traditional feminine role of homemaker and child-rearer. Yet Congress also sought to provide redress for the inequity of"fail[ing] to attach an economic value to the contribution of the homemaker."8 2 These purposes mirror the idealist and realist narratives in the floor debate: sometimes Congress sought to protect and guarantee the traditional role of women, while other times, acknowledging that dynamic as a reality, Congress aimed at equality through compensating affected women.8 3 Califano does not provide a straightforward answer for a case like Adkins, where the statute had such mixed purposes. Racial and employment discrimination cases, however, offer some guidance. In Arlington Heights v. Metropolitan Housing Development Corp., the Supreme Court considered the constitutionality of a zoning regulation that was neutral on its face but had racially discriminatory effects.84 Since legislatures rarely pass a law for one discrete purpose, the Court reasoned, the plaintiffs needed only to show that racial discrimination was a motivating factor.85 This would trigger a 79 1d. at 317 (quoting Craig v. Boren, 429 U.S. 190, 197 (1976)). 80 Id. (citations omitted) (quoting Schlesinger v. Ballard, 419 U.S. 498, 508 (1975), and Stanton v. Stanton, 421 U.S. 7, 15 (1975)). 81 Id. (quoting Califano v. Goldfarb, 430 U.S. 199, 209 n.8 (1977) (plurality opinion)). 82 128 CONG. REC. 18315 (1982) (remarks of Rep. Schroeder); see also supra text accompanying note 53 (discussing the context for this quotation). 83 See supra text accompanying notes 53-57. 84 Arlington Heights v. Metro. Housing Dev. Corp., 429 U.S. 252, 259-60 (1977). 85 Id. at 265-66. 260 [Vol. 17:2 "The Toughest Job" burden-shifting scheme borrowed from the employment discrimination context of Mt. Healthy Board of Education v. Doyle:86 the plaintiff bore the burden of establishing that discrimination was a motivating factor, but once the factor was clear, the burden shifted to the defendant to prove "by a preponderance of the evidence that it would have reached the same decision" even had it not considered the unconstitutional motivation.87 In Adkins, the veterans could point to the idealist justification as a motivating factor impermissibly furthering "a traditional way of thinking about females. 8 8 After the burden shifted to the Department of Defense, however, the realist justification could provide evidence that Congress's remedial purpose, though discriminatory, was both a proper "important governmental objective" and sufficient alone to justify the USFSPA's passage. Accordingly, the Act would survive intermediate scrutiny. Though the Fourth Circuit glossed over substantial evidence of discriminatory intent and thus applied a lower level of scrutiny than such discrimination called for, it was right to conclude that the USFSPA did not violate the Equal Protection Clause. III. THE MARITAL POLICY IMPLICATIONS OF THE USFSPA A. How the USFSPA Affects Property Outcomes upon Divorce Whether the USFSPA is good policy depends on how well it accomplishes its goals and whether those goals are proper to begin with. To evaluate the question of how well the USFSPA accomplishes its goals, it is useful to consider the outcomes in different scenarios. If we assume that the military pension is a constant value, the member's spouse sometimes will have no pension of his or her own, sometimes will have a pension smaller than the military pension, sometimes will have a larger pension, and sometimes will have a pension equal to the member's. Different divorce law regimes affect each of these scenarios differently. First, consider the effect 86 429 U.S. 274 (1977). In Mt. Healthy, a public school refused to rehire an untenured school teacher after he exercised his First Amendment rights, id at 274, but the Court held that he had only suffered a harm if he could demonstrate that he would not have been fired on other grounds, id at 285 ("[The Court should not] place an employee in a better position as a result of the exercise of constitutionally protected conduct than he would have occupied had he done nothing."). 87 Id. at 287; accord Arlington Heights, 429 U.S. at 270 n.21. s8 Califano v. Webster, 430 U.S. 313, 320 (1977) (per curiam) (quoting Califano v. Goldfarb, 430 U.S. 199, 223 (Stevens, J., concurring)). 2008] Columbia Journal of Gender and Law of the USFSPA on a fifty-fifty community property regime where, in the absence of the opportunity to divide the military pension, the state court would opt not to divide the spouse's pension (and assume that, if the court does divide the military pension, it divides the spouse's pension as well): Pension values Division upon Division upon before divorce divorce without the divorce under the USFSPA USFSPA Only the member Member = $40,000 Member = $40,000 Member = $20,000 has a pension Spouse = $0 Spouse = $0 Spouse = $20,000 Member pension Member = $40,000 Member = $40,000 Member = $30,000 > spouse pension Spouse = $20,000 Spouse = $20,000 Spouse = $30,000 Member pension Member = $40,000 Member = $40,000 Member = $40,000 = spouse pension Spouse = $40,000 Spouse = $40,000 Spouse = $40,000 Member pension Member = $40,000 Member = $40,000 Member = $50,000 < spouse pension Spouse = $60,000 Spouse = $60,000 Spouse = $50,000 Table 1: Divided Pension Values Where the Court Would Opt, or Be Compelled, Not to Divide the Spouse's Pension Unless It Divides the Military Pension as Well. Assume member pension = $40,000. Here the court would divide the spouse's pension if and only if it divided the military pension. Without the USFSPA, all parties are entitled to the pensions that accrue in their names during the marriage, but under the USFSPA, both the member and the spouse end up with equal pensions. The person with the larger pension is better off if the pensions are not divided, and the person with the smaller pension is better off under the USFSPA, where both pensions are divided. Without the USFSPA, both the member and the spouse are indifferent upon divorce as to the size of their partner's pension. Under the USFSPA, the larger the spouse's pension, the better off both parties are upon divorce. Next, let us consider the effect of the USFSPA on the same state law regime but assume that, regardless of whether the court divides the military pension, it does divide the spouse's pension: 262 [Vol. 17:2 "The Toughest Job" Pension values Division upon Division upon before divorce divorce without divorce under the the USFSPA USFSPA Only the member Member = $40,000 Member = $40,000 Member = $20,000 has a pension Spouse = $0 Spouse = $0 Spouse = $20,000 Member pension > Member = $40,000 Member = $50,000 Member = $30,000 spouse pension Spouse = $20,000 Spouse = $10,000 Spouse = $30,000 Member pension = Member = $40,000 Member = $60,000 Member = $40,000 spousepension Spouse = $40,000 Spouse = $20,000 Spouse = $40,000 Member pension < Member = $40,000 Member = $70,000 Member = $50,000 spouse pension Spouse = $60,000 Spouse = $30,000 Spouse = $50,000 Table 2: Divided Pension Values Where the Court Would Opt or Be Compelled to Divide the Spouse's Pension Whether or Not It Divides the Military Pension. Assume member pension = $40,000. Under these assumptions, the court would not consider the value of the military pension part of the marital property unless the USFSPA allowed it to do so, but would nevertheless consider the value of the spouse's pension. Without the USFSPA, the service member is always better off than the spouse regardless of the size of the spouse's pension. It is thus always the spouse who benefits from the USFSPA, and the service member who loses, if the court takes the military pension into account. In both tables, under the USFSPA, the larger the spouse's pension, the better off both parties are upon divorce. But in Table 2, unlike Table 1, this holds true even without the USFSPA. Both parties would be indifferent between the outcomes of the two tables under the USFSPA. Absent the USFSPA, the spouse would prefer the outcomes of Table 1 and the service member those of Table 2, except where the spouse has no pension, in which case both parties would be indifferent between the two sets of outcomes. Whether the outcomes end up more like Table 1 or Table 2 depends on state law. Part of the legislative history suggests that Congress considered Table 1's outcomes to be both fairer and likelier.89 Yet in other 89 In answer to a question posed by Senator Jim Exon regarding the fairness of a situation where "a pension earner divorced from a military member could have no right to any portion of the member's pension, while the military member could receive a major portion of his former spouse's pension," USFSPA Hearings, supra note 41, at 228 (statement of Sen. Exon). Lt. Gen. Andrew losue of the Air Force "[c]oncur[red] that the situation is unfair": 2008] 264 Columbia Journal of Gender and Law [Vol. 17:2 instances Congress seemed to contemplate Table 2's outcomes as a real possibility.90 Three states-California, Louisiana, and New Mexico- require courts to divide marital property equally, which compels the outcomes of Table 2.91 A handful of other states start with a presumption of fifty-fifty division based either on statute 92 or case law.93 In these states, courts would likely divide pensions according to Table 2 unless they found sufficient reason, such as mitigating factors or inequity, to do otherwise. All other states divide property upon divorce guided by principles of "equitable distribution" and would likely select among those outcomes represented in Table 1.9 4 However, as a practical matter it is unlikely that a court would award a military spouse a portion of the civilian spouse's pension when military retirement pay is not divisible. If a court made such an award, it seems likely that the court would compensate the civilian spouse with an offset of other marital property. Id. (statement of Lt. Gen. Andrew P. losue, Deputy Chief of Staff for Manpower and Personnel, U.S. Air Force). 90 See 128 CONG. REC. 18332 (1982) (remarks of Rep. Clausen) ("As California is a community property State, the settlement required that all of the couple's assets, excluding his pension, be divided equally between them." (emphasis added)). 91 See CAL. FAM. CODE ANN. § 2550 (West 2004) (requiring equal property division unless the parties agree to an unequal division or the statute otherwise provides); LA. REV. STAT. ANN. § 9:2801 (1997) (requiring equal property division unless the parties agree to an unequal division); Ruggles v. Ruggles, 860 P.2d 182, 192 (N.M. 1993). 92 See ARK. CODE ANN. § 9-12-315 (2002); IDAHO CODE ANN. § 32-712 (2006); IND. CODE ANN. § 31-15-7-5 (LexisNexis 2003); NEV. REV. STAT. ANN. § 125.150 (LexisNexis 2004); N.H. REV. STAT. ANN. § 458:16-a (2004); N.C. GEN. STAT. ANN. § 50-20 (2005); OR. REV. STAT. ANN. § 107.105 (1990); TEX. FAM. CODE ANN. §§ 3.003(a), 7.001 (Vernon 2005); W. VA. CODE ANN. §§ 48-7-101, -103 (LexisNexis 2004); WIS. STAT. ANN. § 767.61 (West 2001). 93 See Toth v. Toth, 946 P.2d 900, 903 (Ariz. 1997) (en banc); Robertson v. Robertson, 593 So. 2d 491, 493 (Fla. 1991) (per curiam); Booth v. Booth, 541 N.E.2d 1028, 1031 (Ohio 1989); see also Anderson v. Anderson, 707 S.W.2d 166, 169 (Tex. App. 1986) (holding an award of one-half of the former husband's military retirement benefits to the former wife upon divorce proper since such benefits were community property). 94 At the time of the USFSPA's passage, Congress was aware that three states remained "separate 'title' States-Mississippi, Virginia, and West Virginia." 128 CONG. REc. 18314 (remarks of Rep. Schroeder). Under a strict title theory, regardless of the USFSPA, states would distribute property according to those outcomes represented in the "Division upon divorce without the USFSPA" column of Table 1. All three states have since abandoned strict title in favor of either equitable distribution or community property. For 2008] "The Toughest Job" 265 Spouse Outcomes Upon Divorce 00 u >02 00 .. Cn Cl, $0 $20,000 $40,000 Spouse Ex-ante Pension Value -A Court Divides Neither Pension ---- Court Divides Spouse's Pension Only - Court [Divides Both Pensions $60,000 Figure 1: The Effects of Three Legal Schemes on Spouse Pension Value. Assume member pension is $40,000. The "Court Divides Neither Pension" line corresponds to the middle column in Table 1. The "Court Divides Spouse's Pension Only" line corresponds to the middle column in Table 2. The "Court Divides Both Pensions" line corresponds to the right column in both tables, where the USFSPA allows courts to consider the military pension as part of the marital property. The spouse's outcomes upon divorce, as shown along the vertical axis, are always higher under the USFSPA than where the court only divides the spouse's pension. They are also higher under the USFSPA than where the court divides neither until the point where the spouse's predivorce pension surpasses the service member's. instance, in renouncing strict title in 1994, the Supreme Court of Mississippi recognized a potential spousal interest in retirement benefits: A spouse who has made a material contribution toward the acquisition of an asset titled in the name of the other spouse may claim an equitable interest in such jointly accumulated property. Although contributions of domestic services are not made directly to a retirement fund, they are nonetheless valid material contributions which indirectly contribute to any number of marital assets, thereby making such assets jointly acquired. Ferguson v. Ferguson, 639 So. 2d 921, 934 (Miss. 1994) (en banc) (citations omitted). 266 Columbia Journal of Gender and Law [Vol. 17:2 Gender factors into these outcomes in interesting and somewhat counterintuitive ways. As Figure 1 demonstrates, the spouses of service members benefit from the USFSPA in all scenarios except where the spouse has a more valuable pension than the service member and the court divides that pension if and only if it can divide the military pension as well.95 To the extent that the spouses are women, as Congress assumed, women benefit from the USFSPA in most scenarios. Further, to the extent that those spouses who are women are likely to be homemakers, or to have jobs with smaller or no pensions, it is even more likely for military wives to fall into the top two rows of either table and the left half of Figure 1, in which case the USFSPA would benefit them at the expense of their husbands. To a large extent, the demographic information put forth by the Adkins plaintiffs pans out. In both absolute and percentage terms, there were indeed more women in the military in 2005 than there were in 1980: female enrollment has increased from 171,400 and 8 percent in 1980 to 202,400 and nearly 15 percent in 2005.96 The chance of a service member being a woman nearly doubled in those twenty-five years, but women in the military remain outnumbered by men by about six to one. Military wives today work fewer hours, earn lower wages, and are less likely to engage in market work at all than their civilian counterparts.97 In the late 1970s, however, they participated in the labor force at a rate roughly equal to that of civilian wives. 98 Married women in general are less likely than men to have a pension plan, but the rates are much closer today (46 percent to 48 percent, respectively, as of 2004) than they were in 1979 (35 percent to 60 percent, respectively). 99 Army men marry at higher rates than civilian men. 100 95 This exception corresponds to the portion of Figure 1 where the "Court Divides Neither" line surpasses the "Court Divides Both" line. See also supra p. 262 tbl.1, last row (demonstrating that spouses whose pensions surpass those of the service members lose when the court divides both pensions). 96 U.S. CENSUS BUREAU, STATISTICAL ABSTRACT OF THE UNITED STATES 2007, at 332 (2007). 97 HOSEK ET AL., supra note 44, at 34-41. 98 Allyson Sherman Grossman, The Employment Situation for Military Wives, MONTHLY LAB. REV., Feb. 1981, at 60, 60. 99 Munnell with Zhivan, supra note 29, at 16. 100 CASEY WARDYNSKI, MILITARY COMPENSATION IN THE AGE OF Two-INCOME HOUSEHOLDS: ADDING SPOUSES' EARNINGS TO THE COMPENSATION POLICY MIX 10 (2000), available at http://www.rand.org/pubs/rgs dissertations/RGSD 154. "The Toughest Job" For service members, the USFSPA is a losing proposition, except in the last row of Table 1 (for the same reasons that the last row is the only scenario in which the USFSPA is undesirable to spouses). But the Adkins veterans argue that this exception is more likely for female service members than it is for male service members: "former husbands are more likely than former wives to have sources of income other than the divided military retirement pay."' 01 Having a spouse with a large pension, or indeed any pension, is a benefit to the service member in all scenarios except where the court opts not to divide pensions (as represented in the middle column of Table 1), in which case the service member would be indifferent to the size of his or her spouse's pension. If female service members are more likely to have pensioned spouses than male service members, they will either have equal outcomes to the male service members or be better off, regardless of whether the USFSPA allows courts to divide their pensions. The USFSPA makes female service members lose in relation to their spouses the same way it makes male service members lose in relation to theirs. But the USFSPA makes no difference to the comparison between female and male service members: female service members either break even or win, with or without the Act. The Adkins plaintiffs have either wholly overlooked these outcomes or instead implied that the USFSPA will force courts to divide the military pension but not the spouse's pension-rather than allowing courts that would otherwise have divided neither pension to divide both, as in Table 1, or allowing courts that would otherwise have divided only the spouse's pension to divide both, as in Table 2. The plaintiffs present no evidence to support this theory. Further, it seems highly improbable that courts concerned with fairness and equity would divide the military pension but not the spouse's pension when there are no legal restrictions on considering nonmilitary pensions to be marital property. This outcome scheme presents three primary groups of winners, along with corollary losers, under the USFSPA. First, spouses tend to win, and service members to lose, because the Act allows for the redistribution of income to which the service member would have otherwise been wholly entitled. Second, female service members tend to win and males to lose, as noted above, to the extent women more frequently have spouses with larger pensions 0 2 and the USFSPA encourages courts to divide both the military 101 Adkins v. Rumsfeld, 464 F.3d 456, 468 (4th Cir. 2006); accord Appellant Brief, supra note 30, at 51-52. 102 See Munnell with Zhivan, supra note 29, at 11, 16. 2008] Columbia Journal of Gender and Law pension and the spouse's pension. Third, those whose spouses have larger pensions win or break even against those whose spouses have smaller pensions, regardless of whether the service member is the husband or wife, because the redistribution allows the smaller-pensioned or nonpensioned spouse to share in the income of the spouse with the larger pension. B. Incentives the USFSPA Develops There are four major decisions an individual might make that the USFSPA will weigh on: the decision to divorce a service member, the decision to marry a service member in the first place, the decision to join the military, and the decision, having married a service member, whether to engage in market work. For each, the Act creates a set of incentives that seem to align with what Congress intended, despite being somewhat counterintuitive. The absolute winners with the adoption of the USFSPA are spouses in all scenarios except where the spouse's pension is larger than the military pension and the court would not have otherwise divided them.'0 3 The relative winners are those members whose spouses have pensions, and the larger the pension, the better. From the perspective of spouses with pensions smaller than those of the members, the Act makes divorce a more attractive option than it would be otherwise: a divorcing spouse can be certain that the court will take the member's valuable pension into account. The corollary to this is that, by allowing for the safety net of a divided pension, the Act makes marriage to a service member a less risky option for those spouses who have either smaller pensions or no pension. Individuals will be more willing to divorce a service member, but, precisely because of this relative freedom to divorce, they will be more willing to marry the service member in the first place. 10 4 Assuming that such spouses provide emotional and domestic support to the service member, the Act makes becoming a service member attractive to the extent that it makes becoming a service member's spouse attractive. If more potential spouses will be more willing to marry service members under the USFSPA, individuals will be more willing to join the military because the Act assuages their fear that the military will make them 103 See supra p. 265 fig. 1. 104 See also supra note 100 and accompanying text (noting that members of the military tend to marry at higher rates than civilians). 268 [Vol. 17:2 2008] "The Toughest Job" 269 unsuitable partners. 05 When congressional idealists were concerned with providing incentives, this is one of the effects they most sought from the statute.10 6 Further, in situations where the court would not divide a spouse's pension unless it divided the military pension, the USFSPA provides incentives for service members to find spouses who have higher pensions. Spouse Outcomes Upon Divorce as a Percentage of Ex-ante Pension Values 200%• = 20- 150% -0 100% . xO. *0 uj :7 cj, - 100% 00 G 0% $20,000 $40,000 $60,000 Spouse Ex-ante Pension Value A Court Divides Neither Pension -- W- Court Divides Spouse's Pension Only -0- Court Divides Both Pensions Figure 2: The Effects of Three Legal Schemes on Postdivorce Percentage Return on Ex-ante Spouse Pension Value. Assume member pension is $40,000. The "Court Divides Neither Pension" line corresponds to the middle column in 105 It seems intuitive that members of the military will prefer having stable, supportive family situations, and that an improvement in the family life of service members will encourage more people to remain in the military, but there is a dearth of empirical research on the topic. See HOSEK ET AL., supra note 44, at 1-2. 106 See, e.g., USFSPA Hearings, supra note 41, at 15 (statement of Sen. Hatfield) ("If, indeed, we are to strengthen incentives for participation in the military, as the chairman and Senator Exon have both commented, we cannot continue such economic disincentives for the wife of the military employee."); id at 16 ("Morale, motivation, and reenlistment of our armed services depend on more than take-home pay. Long-range benefits which insure the future financial security of both partners in a military marriage will improve morale and increase reenlistment."); id. at 19 (statement of Sen. DeConcini) ("[A] spouse who is secure in the knowledge of his or her entitlement to a portion of the member's retirement benefit is likely to be more supportive of that member, encourage the member to participate in the military until retirement age and generally add to the stability of the military family."). Columbia Journal of Gender and Law Table I. The "Court Divides Spouse's Pension Only" line corresponds to the middle column in Table 2. The "Court Divides Both Pensions" line corresponds to the right column in both tables, where the USFSPA allows courts to consider the military pension as part of the marital property. The two non-USFSPA lines demonstrate constant returns on pensions for spouses regardless of the size of their own pension: in the "Court Divides Neither" case, spouses get a 100 percent return; in the "Court Divides Spouse's Only" case, spouses get a 50 percent return. In contrast, the scheme the USFSPA allows, where the court divides both pensions, presents postdivorce returns that decrease as the spouse's predivorce pension increases. The Act presents interesting incentives regarding the decision of the service member's spouse to engage in market work to collect a pension. In cases where the court does not divide the military pension, the spouse's postdivorce return on his or her pension benefits will be a constant percentage of the predivorce value: if the court divides neither pension, as in the triangle line in Figure 2, the spouse will collect 100 percent of his or her pension after a divorce, regardless of its absolute value; if the court divides the spouse's pension only, as in Figure 2's box line, the spouse will always reap 50 percent. In the scheme that USFSPA allows for, however, a court could divide both pensions, creating a scenario where the spouse's return varies according to the absolute value of his or her predivorce pension, as in Figure 2's circle line. If the spouse's predivorce pension is worth x, and the military pension is a constant M, the spouse can expect to receive, upon divorce, divided pension benefits equal to M/2 + x/2. Effectively, the rate of return on the spouse's own pension upon divorce diminishes as the spouse's pre-divorce pension value increases. The USFSPA creates a safety net for low- or no-pensioned spouses, making nonpensioned work relatively more attractive. As a result, some spouses who would otherwise have opted to work at jobs that could give them pensions might rely on the USFSPA and instead work at lower-pensioned or nonpensioned jobs-or drop out of the market entirely. The USFSPA is successful in achieving both of the incentive structures invoked by the congressional "traditional family" idealist narrative: the Act makes more attractive not only becoming a military spouse but becoming a homemaker military spouse. 1 07 107 See supra text accompanying notes 53-54. This incentive for the spouse to stay home conflicts with the incentive for the service member to find a spouse with a large pension. 270 [Vol. 17:2 "The Toughest Job" CONCLUSION The USFSPA was designed to address the fairness concerns raised by the McCarty decision's refusal to allow state courts to consider military retirement part of marital property upon divorce. But it was also designed with particular ideas about the traditional roles of the sexes in the minds of those legislators who considered it. The legislative history demonstrates that, even as Congress wrote the Act in chiefly gender-neutral terms, it understood service members to be men and their spouses to be women, predominantly housewives. This presents an equal protection problem with somewhat more depth than the court in Adkins acknowledged; Congress's implicit conception suggested evidence of covert gender discrimination. Yet this discrimination probably did not rise to the level of a violation of Fifth Amendment equal protection. Somewhat more difficult is the question of whether the USFSPA is good family policy. Its scheme of outcomes tends to benefit spouses over service members, female over male service members, and those whose spouses have larger pensions over those whose spouses have smaller pensions. This provides incentives for prospective spouses to marry service members, for all parties to seek spouses with larger pensions, and for military spouses to become homemakers. Where the positive evaluation of the USFSPA ends, the normative evaluation of its goals begins. Most would agree that creating an additional incentive to serve the nation in the armed forces is a beneficial outcome of the Act. But is incentivizing prospective spouses to marry service members laudable for making military home life easier, or is it just making the spouses take on a burden for which the service members still get most of the credit? Is incentivizing current spouses to stay home a good aim? Some scholars of family economics argue that laws that recognize the contributions of the homemaker to the family as a market-earning unit and divide assets accordingly are progressive and laudable. Ann Crittenden, for example, points to "a new generation of family law reformers" whose "ultimate goal is for all family members to emerge from a marital breakdown with roughly equal standards of living."' 0 8 Some feminists argue that what women really need is to get out of the home and into the workplace. Catherine MacKinnon advocates for women to abandon the private sphere for the public sphere, where they can have "freedom of 108 ANN CRITrENDEN, THE PRICE OF MOTHERHOOD 152-53 (2001); see also JOAN WILLIAMS, UNBENDING GENDER 129-38 (2001) (proposing a joint-property scheme for equalizing income between breadwinners and caregiver-homemakers after a divorce). 2008] Columbia Journal of Gender and Law action, resources, and access to a larger world,"'0 9 and Linda Hirshman's aptly named manifesto Get to Work urges women to find empowerment by entering the market."l 0 These reformers would undoubtedly be critical of laws that persuaded women to become homemakers. Such questions depend ultimately on one's values, and their answers are beyond the scope of this Article. But whatever one's views, understanding the incentive framework the USFSPA builds is an important step toward judging it. In the meantime, women in Jessica Miller's and Patricia McCarty's positions are faced with a legal scheme that offers both benefits and drawbacks. They may be rewarded for their contributions to the military home with a share in their husbands' retirement income,1 but it may be small consolation for, not only the hardship of military life and the inherent pain of divorce, but also the lost opportunity to seek market work or a partner even better able to bring in income. The law suggests that their contributions as military wives are valued by society, perhaps as much as the roles of their service-member husbands, but it insinuates that the former is dependent on and secondary to the latter, that the wives' contributions are valuable merely because of the help they provide their husbands. And women in the military, like Tammy Adkins, may be better off than their male service member counterparts if they have spouses with more property to redistribute, but as service members they are disadvantaged relative to their spouses under the Act. It is hard to convince someone who feels swindled out of a retirement they have worked toward that what the government has done to them is a good thing because it is fairer to the spouse they are now leaving. What ultimately makes the consequences of the USFSPA so mixed is that the Act leaves so much unresolved. It does not direct courts to divide retirement pay between a member and his or her spouse in certain proportions, require courts to divide it at all, or even specifically deem retirement pay to be community property within the marriage, leaving what to do with it open to interpretation. It simply states that "court[s] may" consider the pension marital property." 12 By the words of the act, what this actually means for individuals facing divorce will depend on the law of the 109 CATHARINE A. MACKINNON, WOMEN'S LIVES, MEN'S LAWS 237 (2005). 110 LINDA R. HIRSHMAN, GET TO WORK 4 (2006). 111 Strictly speaking, women in Jessica Miller's position have no recourse under the USFSPA; the marriage must have lasted ten or more years for the Congressional redistribution authorization to be valid. 10 U.S.C. § 1408(d)(2) (2000). 2 Id. § 1408(c)(1) (emphasis added). [Vol. 17:2 2008] "The Toughest Job" 273 state in which the proceeding is taking place and the discretion of the judge. The issues the USFSPA raises, then, are largely the issues raised by divorce law in general: issues about how easy it should be to dissolve a marriage, who should bear divorce's burdens, and to whom marital property truly belongs. Military marriages are extraordinary, and their legal and social ramifications raise a set of extraordinary questions, but underlying them are the problems marital law always faces. Before we solve those problems, we have much work to do.