Paul Forman appeals the grant of summary judgment to the Smithsonian Institution on his claims of age discrimination and retaliation under the Age Discrimination in Employment Act ("ADEA"), 29 U.S.C. § 633a (1994 & Supp. V 1999). He contends that he established a prima facie case on each of his claims and rebutted the Smithsonian's stated explanations for its actions. We affirm the judgment on his claims of discrimination with regard to his 1991 and 1995 promotions, but we reverse as to his claim of retaliation with regard to his 1995 promotion.

Paul Forman is a curator for Modern Physics at the National Museum of American History of the Smithsonian Institution. He was hired in 1972 as an associate curator, Grade 12, and received a promotion in 1975 to curator, Grade 13. He was passed over for a non-competitive promotion to Grade 14 in 1988 and 1991-92. In 1995, a decision concerning his promotion was postponed for one year. He was promoted to Grade 14 in 1996 when he was 59 years old. The relevant background to these decisions is as follows.

In May 1988, Dr. Forman requested and was granted a two-year temporary duty assignment to New York City with the primary task of preparing a draft of a book on the history of atomic clocks.1 His normal day-to-day duties as curator, relating to exhibitions and collections, were minimized. For the rating year September 1, 1988, to August 31, 1989, Dr. Forman received a performance appraisal of "fully successful" from his supervisor. In the same performance evaluation, however, his supervisor advised Dr. Forman that he expected a "concentrated and sustained effort ... during th[e] next year on the book project, now that it is underway." At the time, Dr. Forman had only drafted about one-half of a chapter, albeit a lengthy one.

In January 1990, Dr. Forman proposed that the one chapter he had drafted for his book on atomic clocks about Charles Townes and the maser2 become the basis of a different, shorter book, narrower in scope than the one he had originally committed to write about atomic clocks. His supervisor approved the shorter book. His interim performance appraisal (evaluating only his performance from September 1, 1989 to May 1990 for the evaluation period of September 1, 1989 to August 31, 1990), however, was "unacceptable" as to the "single critical element" of his assignment, namely to draft a "book-length manuscript." An accompanying letter from his supervisor, dated May 22, 1990, described Dr. Forman's lack of substantial progress on the promised book manuscript, focusing on a substantial period of "under productivity" in his central assignment, which was the principal area of his research during the last decade. In his final performance evaluation for the rating period of September 1, 1989 to August 31, 1990, Dr. Forman received a rating of "fully successful"; his supervisor noted that Dr. Forman had begun in the latter part of the performance year to produce "commendable draft chapters of the Townes and the Maser manuscript at a steady pace." Dr. Forman returned to work at the Smithsonian Institution in October 1990. In the next rating year, from September 1, 1990 to August 31, 1991, he again received a "fully successful" rating.

Robert McCormick Adams was the Secretary when Dr. Forman was considered for promotion in 1991. During this period there were six principal features of the promotion process for all curators at the National Museum of American History. The Professional Accomplishment Evaluation Committee, which is a peer evaluation committee of curators appointed by the Director of the Museum, considers curators at Grade 13 for possible non-competitive promotion every three years. The peer review committee's recommendation is advisory to the Director of the Museum. The Director also considered other factors such as annual summary performance appraisals as well as the opinions of the curator's supervisors.3 The Director made an advisory recommendation to the Secretary. The Secretary also customarily received advisory recommendations from his Assistant Secretaries before making his final decision. Thus, the Secretary had the final authority to make decisions regarding promotions.

In April 1991, the peer review committee recommended Dr. Forman for promotion to Grade 14. The Director of the Museum advised Dr. Forman in June 1991 that in light of the fact that none of his supervisors thought he was working at a Grade 14 level, and the primacy of a book in his performance plan since 1978, Dr. Forman would not be recommended for promotion. The Director nonetheless forwarded Dr. Forman's promotion package to the Assistant Secretary for Research. The two Assistant Secretaries, Robert Hoffman and Tom Freudenheim, reviewed Dr. Forman's promotion package. Hoffman recommended to the Secretary that Dr. Forman be promoted; Freudenheim recommended against promotion. In March 1992, after reviewing Dr. Forman's promotion package and discussing the matter with both Assistant Secretaries, Secretary Adams decided not to promote him, expressing concern that notwithstanding Dr. Forman's international reputation as an historian, he had failed to produce a book-length manuscript on atomic clocks "or any other work of comparable scope." Secretary Adams decided that consideration of a promotion should be postponed until Dr. Forman completed "a major scholarly work such as the manuscript on atomic clocks, or his proposed biography of Charles Townes, or some other work of his choosing."

During the Secretary's discussion of Dr. Forman's promotion with the Assistant Secretaries, comments were made regarding Dr. Forman's age, generally to the effect that he might be "beyond his years of scholarly productivity"; Secretary Adams denied making these statements. Dr. Forman filed an administrative complaint of age discrimination on May 26, 1992, and upon being denied relief, he filed an administrative appeal with the Equal Employment Opportunity Commission, which was still pending when he was considered for promotion in 1995.

When Dr. Forman was next considered for a promotion in 1995, I. Michael Heyman was the Secretary and Spencer Crew was the Director of the National Museum of American History. Secretary Heyman instituted various changes in the structure and promotion process of the Museum. Secretary Heyman abolished the positions of Assistant Secretary and created in theirplace the position of Provost. During Dr. Forman's 1995 promotion decision, Robert Hoffman served as Acting Provost. Secretary Heyman also instituted a different decision-making system for promotions, delegating responsibility for promotions of scholarly staff to the Directors but with oversight responsibility in the Provost. Generally, the Provost could consider promotions only of persons recommended for promotion by the Director of the Museum. Dr. Crew, in turn, reorganized the Museum to shift its strategic priorities from an "academic mode" toward a "customer service" mode that would be more responsive to the public. The curatorial units were reduced from twenty to five to ensure that curators would be better aware of the interrelationship between their field of expertise and others' and share their knowledge and research with the larger public.

In April 1995, the peer evaluation committee recommended, for a third time, to the Museum Director that Dr. Forman be promoted to Grade 14. Dr. Crew, however, advised Dr. Forman that he was going to postpone his final decision until he could review the results of Dr. Forman's performance plan for 1995-96. While acknowledging the importance of scholarship, Dr. Crew stated that "other factors also weigh quite heavily," most notably the relationship of one's work to the "strategic priorities of the museum" and "the priorities of [one's] supervisor." Dr. Forman had expressed strong opposition to the new strategic priorities, and Dr. Crew explained that he wanted to determine whether Dr. Forman's performance was consistent with the new priorities of the Museum and Forman's supervisors. Dr. Crew did not forward the promotion package to the Acting Provost.

Dr. Forman submitted a complaint to Acting Provost Hoffman, claiming that Hoffman had the authority to promote him unilaterally to Grade 14. In a letter dated October 6, 1995, Secretary Heyman stated that he had requested that Hoffman advise him as to how to act on Dr. Forman's complaint; the Secretary was responding to a letter expressing concern about Dr. Forman's "long overdue promotion" and the importance to the Smithsonian of indicating that it "prize[s] scholarship, originality, and independence" as demonstrated by Dr. Forman. Hoffman turned the complaint and accompanying materials over to Assistant Acting Provost Freudenheim for a recommendation; Freudenheim responded with a memorandum, dated October 27, 1995, which Hoffman interpreted as implicitly recommending Dr. Forman's promotion. In the absence of a recommendation from the Museum Director, however, Hoffman decided to ask Dr. Crew to reconsider his decision not to recommend Dr. Forman's promotion; Dr. Crew did not respond. Although Hoffman again favored Dr. Forman's promotion in light of his research accomplishments, he never "tested the system to determine" if he had "direct authority to overrule the museum director's recommendation," and he did not forward Dr. Forman's complaint to the Secretary, notwithstanding the Secretary's statement in October 1995 that "[Hoffman] expects to talk with all parties and then offer me guidance on how to proceed" regarding Dr. Forman's 1995 promotion. Hoffman explained that he did not forward the complaint because Dr. Forman had already filed an EEO complaint, in which a decision would be made concerning the legitimacy of his claim.

After exhausting his administrative remedies, see 29 C.F.R. § 1614.201(c), Dr. Forman filed a lawsuit against the Smithsonian under the ADEA, 29 U.S.C. § 633a, for age discrimination and retaliation. The district court, observing that "[i]t may very well be that [Dr. Forman] ha[d] not been treated fairly by the Smithsonian," grantedsummary judgment to the Smithsonian on Dr. Forman's age discrimination and retaliation claims. The district court found that Dr. Forman had failed to show that age was a factor in the Smithsonian's refusal to promote him in 1991-92. In so concluding, the district court found that the Smithsonian had articulated a legitimate nondiscriminatory reason for its decision not to promote him in 1991-92, namely his failure to produce a book or major publishable work as outlined in his performance plans, and that Dr. Forman had failed to show that this explanation was pretext for age discrimination. The district court made similar findings as to the 1995 promotion, referencing Dr. Crew's memorandum explaining why he was postponing Dr. Forman's promotion. The court found that Dr. Forman failed to show that Dr. Crew's stated explanation for postponing promotion, namely that Dr. Forman was not meeting the expectations of his supervisors or aligning with the priorities of the Museum, was a pretext for retaliation, and presumably age discrimination. The court also found no evidence of discriminatory retaliation by Dr. Crew or Acting Provost Hoffman when he failed to forward Dr. Forman's complaint to the Secretary. In Part II we address Dr. Forman's promotion claims. In Part III we address his retaliation claims.

On appeal, Dr. Forman contends that he presented a prima facie case of age discrimination because he was over forty years old when his promotions were denied, he was extraordinarily accomplished in his field, the peer committee recommended him in relation to both promotions, he presented statistical evidence that reflected preferential treatment of younger curators, and, as to his 1991 promotion, age-laden comments strongly suggested age bias. He also contends that he established a prima facie case of retaliation as to his 1995 promotion because he engaged in protected activity by filing an EEO complaint regarding the denial of his 1991 promotion, his supervisors knew of his EEO activity, and both Assistant Provost Freudenheim's memorandum and Acting Provost Hoffman's statement that he did not bring Dr. Forman's promotion to the Secretary for decision because Forman had filed an EEO challenge, constituted direct causal evidence between his protected activity and the denial of his promotion. Dr. Forman further contends that he presented sufficient evidence to discredit the Smithsonian's reasons for rejecting both of his promotions.

Our review of the grant of summary judgment is de novo. See Tao v. Freeh, 27 F.3d 635, 638 (D.C. Cir. 1994). Accordingly, the court must view the record in the light most favorable to the nonmoving party, according that party the benefit of all reasonable inferences. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255 (1986); see also Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133, 150 (2000). Consistent with the courts' reluctance to become involved in the micromanagement of everyday employment decisions, see Brown v. Brody, 199 F.3d 446, 451-52 (D.C. Cir. 1999) (citing Mungin v. Katten, Muchin & Zavis, 116 F.3d 1549, 1556-57 (D.C. Cir. 1997)); Fishbach v. D.C. Dep't of Corr., 86 F.3d 1180, 1183 (D.C. Cir. 1996), the question before the court is limited to whether Dr. Forman produced sufficient evidence of age discrimination, not whether he was treated fairly or otherwise entitled to promotion. The Smithsonian does not dispute that Dr. Forman is highly praised by outside scholars for both his exhibits and scholarly writing and that he was generally qualified for promotion.

Section 633a of the ADEA provides that "All personnel actions affecting employees ... in the Smithsonian Institution ... who are at least 40 years of age ... shall be made free from any discrimination based on age." 29 U.S.C. § 633a(a) (Supp. V 1999). This circuit applies to ADEA cases the scheme for allocating evidentiary burdens that has evolved in Title VII discrimination cases. See Krodel v. Young, 748 F.2d 701, 705 (D.C. Cir. 1984). Thus, as summarized in Cuddy v. Carmen, 762 F.2d 119 (D.C. Cir. 1985), the plaintiff must first establish a prima facie case of discrimination. Id. at 122. Upon so doing, the burden of production shifts to the employer to offer a legitimate nondiscriminatory reason for its action. Id. Upon the employer's meeting of this burden of production, the plaintiff, however, carries the overall burden of persuasion, which may be met either indirectly by showing the employer's reason is pretextual or directly by showing that it was more likely than not that the employer was motivated by discrimination. Id. at 123; see Reeves, 530 U.S. at 143, 146-47.

The ultimate question is whether age was a determining factor in the disputed employment decision. See Cuddy, 762 F.2d at 123. In failure to promote cases, a prima facie case is made by showing: (1) the plaintiff is at least forty years of age; (2) the plaintiff was qualified for the position in question; (3) the plaintiff was not promoted; and (4) the plaintiff was disadvantaged in favor of a younger person. See Cuddy v. Carmen, 694 F.2d 853, 856-57 (D.C. Cir. 1982).

Regarding the denial of his 1991 promotion, Dr. Forman presented evidence that he was a member of the protected class, he was generally qualified for promotion to Grade 14, and yet he was not promoted. He also presented evidence that he alone of all curators had been twice denied promotions in the face of two recommendations by the peer committee. To support the fourth element of his prima facie case, Dr. Forman presented expert statistical evidence to show that younger employees were favored for promotion. Specifically, Dr. Forman presented evidence that persons under forty-five years of age had a higher rate of promotion to Grade 14 than those over forty-five, and that there was an inverse correlation between a curator's age and the annual ratings given for research. These differences were statistically significant using either a one-tailed or two-tailed test of significance. See Palmer v. Shultz, 815 F.2d 84, 90-97 (D.C. Cir. 1987). This circuit recognizes statistical data as relevant in individual discrimination claims. See Minority Employees at NASA v. Beggs, 723 F.2d 958, 962 (D.C. Cir. 1983); see also Bell v. EPA, 232 F.3d 546, 553 (7th Cir. 2000); Adams v. Ameritech Servs., Inc., 231 F.3d 414, 423-24, 427 (7th Cir. 2000). Although the Smithsonian showed that several older curators were promoted, this is not dispositive, see O'Connor v. Consol. Coin Caterers Corp., 517 U.S. 308, 312 (1996), and Dr. Forman maintains that the Smithsonian never disputed the overall statistical trend. The Smithsonian maintains that the statistics are deficient because they rely on an overly broad data pool, but the Smithsonian does not dispute that the statistics were based on information that it supplied in response to Dr. Forman's discovery requests.

In any event, Dr. Forman introduced other evidence that age was a primary consideration in the denial of his promotion in 1991 to meet his prima facie burden, which is not onerous. See Tex. Dep't of Cmty. Affairs v. Burdine, 450U.S. 248, 253 (1981). Dr. Forman presented evidence that when Secretary Adams was reviewing Dr. Forman's 1991 promotion papers with the two Assistant Secretaries, a series of comments were made that implicitly referred to Dr. Forman's age. According to Assistant Secretary Hoffman, comments were made that Dr. Forman may be "over the hill" or in the "twilight of his career," and may have "written his last significant article." As pointed out in Hunt v. City of Markham, 219 F.3d 649 (7th Cir. 2000), when decision makers, or those who have input into the decision, express such discriminatory feelings around the relevant time in regard to the adverse employment action complained of, "then it may be possible to infer that the decision makers were influenced by those feelings in making their decisions." Id. at 653. Moreover, the employer's correlation of old age with declining productivity represents the very essence of age discrimination. See Hazen Paper Co. v. Biggins, 507 U.S. 604, 610 (1993); cf. Price Waterhouse v. Hopkins, 490 U.S. 228, 251 (1989).

With this evidence, Dr. Forman has presented a prima facie case that shifts the burden of coming forward with evidence to the Smithsonian to show that its action was not based on Dr. Forman's age. The Smithsonian has met this burden of production, presenting evidence that Dr. Forman was not promoted because of his failure to produce a booklength manuscript on atomic clocks "or any other work of comparable scope." Because Dr. Forman has no direct evidence of age discrimination, the dispositive question is whether he showed that the Smithsonian's explanation for its decision not to promote him in 1991 was a pretext for discrimination. See St. Mary's Honor Ctr. v. Hicks, 509 U.S. 502, 511 (1993). Dr. Forman disputes that he was obligated to produce a book during his New York City sabbatical, and maintains that he, in fact, produced the quantitative word target set for his sabbatical and was productively focusing on Townes (but could not complete the biography because Townes was refusing access to his papers). Dr. Forman questions why completion of yet another major scholarly work was made a condition of his promotion, observing that this supposedly critical requirement vanished in later years and that completion of a book was not generally a requirement for promotion. This is insufficient evidence to show pretext.

It is undisputed that Dr. Forman's performance plans for the relevant period called for him to produce a book or comparable body of work. Dr. Forman did not produce evidence to show fulfillment of this requirement. The fact that completion of a book dropped from later promotion decisions is insufficient to show pretext because the later promotion decisions occurred under different decision makers using different procedures, having different priorities, and considering different performance evaluations. Further, the fact that Dr. Forman may have met word targets is not the equivalent of producing a final book-length manuscript; the latter, not merely the former, was specified in his performance plans, and Dr. Forman admitted that he finished neither his atomic clock book nor his Townes manuscript. Nor does Dr. Forman's inability to finish his Townes manuscript because of factors beyond his control rebut the fact that he did not produce a critical element of his performance plan for several years. Hence, notwithstanding the agebased comments at the discussion of his promotion, the Smithsonian produced evidence of a nondiscriminatory reason fordenying him a noncompetitive promotion in 1991.

Dr. Forman's reliance on Aka v. Washington Hospital Center, 156 F.3d 1284 (D.C. Cir. 1998), is appropriate to the extent it sets forth the proper legal analysis, but Aka highlights precisely what is missing here. In Aka, the plaintiff offered evidence from which a reasonable jury could find that he was "markedly more qualified" than the person selected for the position at issue. Id. at 1299. Dr. Forman's task is more difficult to the extent he is competing against himself. That the promotion of others did not depend on completion of a book is irrelevant to Dr. Forman's particular promotion decision. Unless he could show that he had fulfilled the central purpose of his sabbatical and performance plans, he cannot show that Secretary Adams' reason for denying his promotion was pretextual. As the district court explained to Dr. Forman:

What is relevant is that they thought your work product or your output was inadequate, given the fact that you had no other significant responsibilities during that period of time, and that you were expected, during that twoyear period of time, to produce publishable-quality written material. * * * You may disagree with their evaluation of what you were doing that period of time, but that's not age discrimination. B.

Dr. Forman's 1995 promotion age-discrimination claim is supported by neither the statistical evidence nor the agebased remarks by decision makers that he presented in connection with the denial of his 1991 promotion. The statistical evidence, which examined only 1990 to 1993 and 1987 to 1992, does not address the relevant period, and there is no evidence to support the inference that the statistical trends during these periods extended to 1995. As to Dr. Forman's age, Dr. Crew, the Director who made the decision to postpone his decision on Dr. Forman's promotion until his performance during the upcoming year could be evaluated, stated that he was unaware of Dr. Forman's age when he decided to postpone Forman's promotion. Although Dr. Forman maintains this denial is evidence of age discrimination, this is speculation, which is not the same as evidence showing that age was a substantial factor in Dr. Crew's decision. See McGill v. Munoz, 203 F.3d 843, 846 (D.C. Cir. 2000); see also Fed. R. Civ. P. 56(e). Rather, the evidence showed that Dr. Crew focused on whether Dr. Forman would adapt to the Museum's new direction and his supervisor's expectations. Other evidence corroborates Dr. Crew's explanation. Dr. Forman had protested the Museum's new direction, thereby indicating that the Museum did have a new focus. Further, Dr. Forman's immediate supervisor had raised some of the same concerns expressed by Dr. Crew, urging Dr. Forman to collaborate more with others and do some new exhibition work and thus "expand the audiences with whom he is communicating." The evidence that Dr. Forman produced to suggest that Assistant Acting Provost Freudenheim questioned whether the Museum had such a new direction was later retracted by Freudenheim as being based on only information provided by Dr. Forman, and in any event, at most, suggests that Dr. Crew's stated explanation might be false, not that Dr. Crew's decision was age-based.

For these reasons, we hold that Dr. Forman failed to present a prima facie case of age discrimination in the 1995 denial of his promotion.

We conclude, however, that Dr. Forman produced sufficient evidence to establish aprima facie case of retaliation when Acting Provost Hoffman failed to forward Dr. Forman's complaint materials in response to Secretary Heyman's request for advice on Dr. Forman's 1995 promotion. In contrast, Dr. Forman did not present a prima facie case of retaliation as a result of Dr. Crew's failure to promote him. We first address a threshold jurisdictional issue, however, before turning to the merits.

For purposes of the ADEA, the Smithsonian is included in the section addressing age discrimination in federal agency employment. See 29 U.S.C. § 633a(a) (Supp. V 1999). Although the court has considered whether the Smithsonian is a federal agency under certain statutes, see Expeditions Unlimited Aquatic Enters. v. Smithsonian Inst., 566 F.2d 289, 296 (D.C. Cir. 1977); Dong v. Smithsonian Inst., 125 F.3d 877, 879 (D.C. Cir. 1997), it has yet to address whether the Smithsonian is entitled to sovereign immunity.

Several elements of the Smithsonian's congressional design would appear to suggest that it does have sovereign immunity. First, it operates under a federal charter, 20 U.S.C. § 41, and its Board of Regents is composed of or selected by federal officials, id. §§ 42-43. Second, it is authorized to receive appropriations from Congress. See id. §§ 53a, 54; General Hearings Before the Subcommittee on Library and Memorials, 91st Cong. 323 (1970), cited in Expeditions Unlimited Aquatic Enters., 566 F.2d at 296 n.4. Third, "[a]ll moneys recovered by or accruing to, the institution shall be paid into the Treasury of the United States, to the credit of the Smithsonian bequest, and separately accounted for," 20 U.S.C. § 53, and disbursements for payments of debt are submitted to the Treasury, id. § 57. Ultimately, as the Supreme Court observed in Land v. Dollar, 330 U.S. 731 (1947), whether "a suit is one against the sovereign" turns on whether "[t]he 'essential nature and effect of the proceeding' may be such as to make plain that the judgment sought would expend itself on the public treasury or domain, or interfere with the public administration." Id. at 738 (quoting Ex parte State of New York, 256 U.S. 490, 500, 502 (1921)). Thus, notwithstanding that the Smithsonian is authorized to receive gifts from private sources, see 20 U.S.C. § 55, the Smithsonian's structure and federal funding would suggest that Congress's interest in safeguarding the public fisc from money judgments is no less significant with respect to the Smithsonian than any federal agency. Cf. Story v. Snyder, 184 F.2d 454, 457 (D.C. Cir. 1950). Nonetheless, we do not decide the issue. Rather, in order to ensure a consistent reading of the scope of § 633a, we assume that the Smithsonian has sovereign immunity.

Consequently, before addressing the merits of Dr. Forman's retaliation claims, we must first determine whether Dr. Forman, as an employee of the Smithsonian, may bring a retaliation claim under § 633a of the ADEA. Although the Smithsonian, which is represented by the United States Attorney, does not question whether § 633a prohibits retaliation, the court must because "officers of the United States possess no power through their actions to waive an immunity of the United States or to confer jurisdiction on a court in the absence of some express provision of Congress." Dep't of the Navy v. Fed. Labor Relations Auth., 56 F.3d 273, 275 (D.C. Cir. 1995) (quoting United States v. N.Y. Rayon Importing Co., 329 U.S. 654, 660 (1947)); see also First Va. Bank v. Randolph, 110 F.3d 75, 77 (D.C. Cir. 1997).

"In analyzing whether Congress has waived the immunity of the United States, we must construe waivers strictly in favor of the sovereign and not enlarge the waiver beyond what the language requires." Library of Congress v. Shaw, 478 U.S. 310, 318 (1986) (citations omitted) (internal quotation marks omitted); accord United States v. Nordic Village, Inc., 503 U.S. 30, 34 (1992). Thus, waiver cannot be implied; it must be unequivocally expressed. United States v. Mitchell, 445 U.S. 535, 538 (1980); Dorsey v. U.S. Dep't of Labor, 41 F.3d 1551, 1554-55 (D.C. Cir. 1994). Accordingly, we turn to the text of § 633a.

Congress expanded the scope of the ADEA in 1974 to include state and local governments and federal employers. See Pub. L. No. 93-259, 88 Stat. 74 (1974) (codified as amended at 29 U.S.C. §§ 630(b), 633a). Unlike state and local governments, which were merely added to the definition of "employer" in the ADEA, Congress created an entirely new section of the ADEA in which it waived federal sovereign immunity. This section, codified as § 633a, provides that "[a]ll personnel actions affecting [federal agency] employees ... shall be made free from any discrimination based on age." 29 U.S.C. § 633a(a). Thus, "Congress deliberately prescribed a distinct statutory scheme applicable only to the federal sector." Lehman v. Nakshian, 453 U.S. 156, 167 n.15 (1981). Unlike § 623(d), the ADEA provision governing private, state, and local employers, however, § 633a does not by its terms expressly prohibit retaliation. Section 623(d) explicitly includes retaliation within the specified prohibited forms of discrimination under the ADEA, providing that "[i]t shall be unlawful for an employer to discriminate against any of his employees ... because such individual ... has made a charge, ... or participated in any manner in an investigation, proceeding, or litigation under [the ADEA]." 29 U.S.C. § 623(d) (1994). Moreover, the prohibition of retaliation contained in § 623(d) does not apply to federal employees not only because the ADEA defines "employer" as used in § 623(d) to exclude the federal government, see 29 U.S.C. § 630(b), but also because § 633a(f) specifically provides that § 633a should not be subject to the provisions of § 623.

These statutory differences between the federal and private sectors are not dispositive, however, as some courts have concluded in holding that § 633a does not allow a claim for retaliation, see Tomasello v. Rubin, 920 F. Supp. 4, 5-6 (D.D.C. 1996), aff'd on other grounds, 167 F.3d 612 (D.C. Cir. 1999); Koslow v. Hundt, 919 F. Supp. 18, 19-21, 21 (D.D.C. 1995), for it is the language that Congress used in § 633a(a) alone that determines the scope of that provision. Unlike § 623, which is narrowly drawn and sets forth specific prohibited forms of age discrimination in private employment, Congress used sweeping language when it subsequently extended the ADEA to cover federal agency employees. Congress required no less than that "[a]ll personnel actions affecting employees ... who are at least 40 years of age ... shall be made free from any discrimination based on age." 29 U.S.C. § 633a(a) (emphasis added). In enacting § 633a(a), Congress used unqualified language that encompasses a claim of retaliation because "analytically a reprisal for an age discrimination charge is an action in which age bias is a substantial factor." See Siegel v. Kreps, 654 F.2d 773, 782 n.43 (D.C. Cir. 1981) (Robinson, J., concurring in part and dissenting in part) (citations omitted). Congress's failure to mention "retaliation" explicitly does not undermine itsintended breadth of the provision. Cf. PGA Tour, Inc. v. Martin, 121 S. Ct. 1879, 1897, (2001); Teva Pharm., USA, Inc. v. U.S. Food & Drug Admin., 182 F.3d 1003, 1011 (D.C. Cir. 1999). It is difficult to imagine how a workplace could be "free from any discrimination based on age" if, in response to an age discrimination claim, a federal employer could fire or take other action that was adverse to an employee. To treat Congress's mandate as other than comprehensive would produce absurd results, which courts are to avoid. See Griffin v Oceanic Contractors, Inc., 458 U.S. 564, 575 (1982). Nothing in the plain language of § 633a suggests that Congress intended the federal workplace to be less free of age discrimination than the private workplace. To the contrary, Congress's actions show that it intended its mandate to reach more broadly in the federal sector than in the private sector. In amending the ADEA in 1978, Congress eliminated the upper age limit for federal employees in order to effectively end mandatory retirement in the federal sector in most instances, whereas it merely increased the coverage from 65 to 70 for private employers, limiting the protection from mandatory retirement in the private sector. See H.R. Rep. No. 95-950, at 2, 7-8, 10-11 (1978) (Conference Report); 124 Cong. Rec. 8,218 (1978) (Sen. Javits, ranking minority member of the Human Resources Committee). Moreover, the intent of Congress as expressed in the legislative history of § 633a(a) was to "remove discriminatory barriers against employment of older workers in government jobs at the Federal and local government levels as [the ADEA] has and continues to do in private employment." S. Rep. No. 93-690, at 56 (1974); see also 120 Cong. Rec. 8,768 (1974) (remarks of Sen. Bentsen, principal proponent of ADEA extension to federal employees).

This focus on the sweeping language used by Congress is the same reasoning that the court relied upon in holding that § 2000e-16, in which Congress waived sovereign immunity for claims under Title VII, includes a claim for retaliation. See Ethnic Employees of the Library of Congress v. Boorstin, 751 F.2d 1405, 1415 & n.13 (D.C. Cir. 1985) (citing Porter v. Adams, 639 F.2d 273, 277-78 (5th Cir. 1981)). In Porter, the Fifth Circuit explained that § 2000e-16 differs from §§ 2000e-3 and 2000e-4, which are narrowly drawn and prohibit only specific forms of discrimination, because § 2000e-16 is drafted broadly to prohibit "any discrimination based on race, color, religion, sex, or national origin." 42 U.S.C. § 2000e-16; see Porter, 639 F.2d at 277-78. The court reasoned that "the reasonable conclusion, therefore, is that by drafting [§ 2000e-16] to prohibit 'any discrimination,' Congress intended to bar the federal government from engaging in all those forms of discrimination identified in [§§ 2000e-3 and 2000e-4], and others as well." Porter, 639 F.2d at 278; see also White v. Gen. Servs. Admin., 652 F.2d 913, 917 (9th Cir. 1981). Sections 633a and 2000e-16 use identical language in creating a cause of action for federal employees under the ADEA and Title VII, respectively, and thus should be interpreted consistently. Indeed, the Supreme Court has noted that § 633a, as finally enacted, is "patterned directly after [§ 2000e-16] of the Civil Rights Act of 1964, which extend Title VII protections to federal employees." Lehman, 453 U.S. at 167 n.15. Notably, the statutory pattern here favors an unqualified interpretation of Congressional intent, unlike the statutory pattern that confronted the Court in Lehman. See id. at 161.

The fact that, unlike § 2000e-16 of Title VII, § 633a of the ADEA contains an exclusivity provision does not defeat our analysis. The exclusivity provision provides that federal personnel actions under § 633a "shall not be subject to, or affected by, any provision of this chapter," with one exception not relevant here, see 29 U.S.C. § 633a(f), and makes § 633a "self-contained and unaffected by other sections." Lehman, 453 U.S. at 168. Courts relying on § 633a(f) in concluding that § 633a does not allow a claim of retaliation, see Tomasello, 920 F. Supp. at 6; Koslow, 919 F. Supp. at 19-20, point to the Supreme Court's language in Lehman that § 633a(f) means that "federal personnel actions covered by [§ 633a] are not subject to any other section of the ADEA," Lehman, 453 U.S. at 168, and reason that "Congress has made clear that in interpreting section 633a, the Court may not borrow provisions from elsewhere in the ADEA." Koslow, 919 F. Supp. at 19-20; Tomasello, 920 F. Supp. at 6. The reasoning fails for two reasons.

First, nothing in the legislative history of § 633a(f), which was added to § 633a in 1978, see Pub. L. No. 95-256, 92 Stat. 191 (1978), suggests that it was intended to limit the broad coverage of § 633a that was originally intended. As noted, the 1978 amendments imposed more stringent requirements upon the federal sector than the private sector.

Second, our analysis is consistent both with § 633a(f) and Lehman's interpretation of it because we do not borrow provisions from elsewhere in the ADEA; rather, we rely on Congress's use of sweeping language in § 633a(a) itself to make unlawful "any discrimination" based on age, as age is defined in the ADEA. In Lehman, the Supreme Court considered whether a federal employee bringing suit pursuant to § 633a had a right to a jury trial. Id. at 157. The Court did not hold, as Koslow implies, that § 633a(f) precludes courts from interpreting § 633a(a) as prohibiting the same conduct prohibited in the private sector in other provisions of the ADEA; rather, Lehman began its analysis with the plain language of § 633a, asking first whether it contained an express provision of a jury trial. Working against the background principle that "[w]hen Congress has waived the sovereign immunity of the United States, it has almost always conditioned that waiver upon a plaintiff's relinquishing any claim to a jury trial," id. at 161, the Court found no Congressional intent to provide federal employees a jury trial because there was no express provision for a jury trial in § 633a, whereas Congress had expressly provided for one for private employees. Id. at 163. In further support of its conclusion, the Court pointed to § 633a(f), noting that, in light of Congress's emphasis that § 633a was self-contained, Congress would not have overlooked the need to provide federal employees a jury trial if it had so intended. Id. at 168.

In the end, then, § 633a(f) presents somewhat of a red herring. We do not incorporate the provisions of § 623(d) into § 633a in concluding that § 633a supports a retaliation claim against the federal government. Compare Ayon v. Sampson, 547 F.2d 446, 449-50 (9th Cir. 1976). To the contrary, we are relying on the plain language of § 633a(a) in holding that a work place cannot be free from any age discrimination if an employer can take an adverse employment action against its employees because the employee has brought an age discrimination claim against the employer. This is age discrimination, which § 633a(a) by its own terms alone prohibits.

For these reasons, we hold that § 633a waives sovereign immunity as to claims of retaliation. We proceed, therefore, to address the merits of Dr. Forman's claims.

The McDonnell Douglas framework is applicable to claims of retaliation. See Passer v. Am. Chem. Soc'y, 935 F.2d 322, 330 (D.C. Cir. 1991); cf. McKenna v. Weinberger, 729 F.2d 783, 790 (D.C. Cir. 1984). In order to establish a prima facie case of retaliation, a plaintiff must show that (1) he engaged in protected activity, (2) he was qualified for the promotion, (3) the employer took an adverse personnel action, and (4) a causal connection existed between the protected activity and the adverse action. See Paquin v. Fed. Nat'l Mortgage Ass'n, 119 F.3d 23, 31 (D.C. Cir. 1997); Mitchell v. Baldrige, 759 F.2d 80, 86 n.5 (D.C. Cir. 1985); McKenna, 729 F.2d at 790. The initial burden is not great, as the plaintiff need only establish facts adequate to permit an inference of retaliatory motive. See McKenna, 729 F.2d at 790.

The district court was unpersuaded that Dr. Forman had made a prima facie case of retaliation. The court construed Hoffman's statement that he did not bring Dr. Forman's complaint to the Secretary "since Dr. Forman had already filed an EEO complaint, in the course of which a decision

would be reached concerning the legitimacy of his claim" to be, "Let's leave it to the courts." Recognizing that Hoffman had previously supported Dr. Forman's promotion, the court rejected Dr. Forman's argument that Hoffman's statement was per se reprisal and that as a result of Hoffman's inaction, Dr. Forman was deprived of consideration and procedures from which he otherwise would have benefitted. The district court erred by not viewing the evidence most favorably to Forman in granting summary judgment. Dr. Forman met his burden by presenting direct evidence of retaliatory motive. Despite Secretary Heyman's request for advice about how to proceed on Dr. Forman's 1995 promotion, Acting Provost Hoffman never forwarded Dr. Forman's complaint materials to the Secretary because, according to Hoffman himself, Dr. Forman had filed an EEO complaint about his 1991 promotion. Hoffman's explanation for not doing so was that the EEO proceeding would determine whether Dr. Forman was entitled to his promotion. While these and other evidentiary issues will remain open on remand, for purposes of summary judgment, Dr. Forman is entitled to the benefit of all reasonable inferences from the evidence before the district court.

It is true that Hoffman supported Dr. Forman's promotion. And it may be true that his failure to forward the complaint to the Secretary was in good faith. But motive, in the sense of malice is not required for liability under the ADEA. Malicious or reckless motive is only pertinent to the issue of liquidated or double damages, which Congress intended to be punitive in nature and are not relevant here. See 29 U.S.C. § 626(b); id. § 216(b); Trans World Airlines, Inc. v. Thurston, 469 U.S. 111, 125 (1985); Smith v. Office of Personnel Mgmt., 778 F.2d 258, 261 (5th Cir. 1985); see also 42 U.S. C. l981a(b)(1). "[A]n employer may offer a legitimate nondiscriminatory reason for taking an adverse action against an employee who has engaged in protected activity.... However, the employer may not proffer a good faith reason for taking retaliatory action." EEOC v. Bd. of Governors of State Colls. & Univs., 957 F.2d 424, 427-28 (7th Cir. 1992); see also Hazen Paper, 507 U.S. at 616; Trans World Airlines, 469 U.S. at 126 & n.19. Unlawful motive, not malicious motive, is all that Dr. Forman had to show.

Consequently, even if Hoffman acted in good faith in failing to forward Dr. Forman's complaint to the Secretary, he nonetheless would violate the ADEA if his reason for doing so was retaliatory, i.e., in response to Dr. Forman's 1991 EEO complaint. Dr. Forman offered evidence, sufficient to defeat summary judgment, of such a retaliatory and hence unlawful motive through Hoffman's own explanation of his inaction. Hoffman's statement is direct evidence that his failure to take Dr. Forman's complaint to the Secretary was substantially motivated by Dr. Forman's prior EEO complaint and hence was retaliatory. Moreover, giving Dr. Forman the benefit of all reasonable inferences, it is unclear whether Hoffman could reasonably have thought that the pending EEO proceeding involving Dr. Forman's 1991 promotion, which turned on his failure to produce a book or book-length manuscript, would resolve Dr. Forman's complaint about the denial of his 1995 promotion, which turned on other factors, such as concern whether he would adapt to the Museum's new focus. Although Hoffman may have thought that administrative resolution of the 1991 promotion in Dr. Forman's favor would have resolved the question of promotion in 1995, an adverse resolution would have left unresolved Dr. Forman's claim of discrimination in 1995. Without a connection between the two, Hoffman's conduct could reasonably be interpreted as involving more than "leaving it to the courts."

Dr. Forman also provided evidence of the remaining elements of a prima facie case of retaliation. First, Dr. Forman's filing of an administrative complaint regarding the denial of his 1991 promotion, as well as his appeal to the Equal Employment Opportunity Commission, were protected activities. See 29 U.S.C. § 623; Holbrook v. Reno, 196 F.3d 255, 263 (D.C. Cir. 1999). Second, he was generally qualified for the promotion, and the Smithsonian does not dispute this. Third, Hoffman's failure to take Dr. Forman's complaint to the Secretary constituted an adverse employment action, viewing the record most favorably to Dr. Forman. The record before the district court showed that Secretary Heyman had delegated promotion responsibility to the Directors, but Dr. Forman introduced evidence that the Secretary had made an exception in Dr. Forman's case, expressly stating that Hoffman was to advise him on Dr. Forman's promotion. Also, the Secretary retained the authority to unilaterally promote curators. Notwithstanding the Secretary's request, Hoffman, although aware of the Secretary's statement that he was awaiting Hoffman's recommendation, never advised the Secretary how to proceed on Dr. Forman's promotion. Hoffman's reason for not acting was that Dr. Forman had filed an EEO complaint. As a result of Hoffman's inaction, a reasonable fact finder could find that Dr. Forman was denied the opportunity for promotion that the Secretary had afforded him. The Smithsonian did not contend that Dr. Forman would not have been promoted had Hoffman taken Dr. Forman's complaint to the Secretary for action. Thus, Dr. Forman's evidence that Hoffman's inaction was an adverse employment action is sufficient to defeat summary judgment for failure to establish a prima facie case of retaliation. See Cones v. Shalala, 199 F.3d 512, 521 (D.C. Cir. 2000); see also Ruggles v. Cal. Polytechnic State Univ., 797 F.2d 782, 785-86 (9th Cir. 1986). Further, the Smithsonian's characterization of its action as a postponement rather than a denial is to no avail; for purposes of summary judgment, Hoffman's inaction was the equivalent of non-promotion. See Price Waterhouse, 490 U.S. at 233 n.1.

Dr. Forman's claim that Dr. Crew retaliated against him as a result of his EEO complaint fails, however, because Dr. Forman does not allege sufficient facts to show causation. Although Dr. Crew knew of the EEO complaint, his decision not to promote Dr. Forman in 1995 occurred three years after Dr. Forman filed his EEO complaint, which challenged his non-promotion in 1991 under a different Smithsonian administration, and after changes had been made in the Museum's curatorial staff. Because of the time lapse, Dr. Forman cannot rely solely on the timing of Dr. Crew's decision not to promote him to show causation. See Holbrook v. Reno, 196 F.3d 255, 263 (D.C. Cir. 1999); Mitchell v. Baldridge, 759 F.2d 80, 86 (D.C. Cir. 1985). Assistant Acting Provost Freudenheim's evaluation of Dr. Crew's denial of promotion also cannot support a causal connection between Dr. Crew's failure to promote Dr. Forman and Dr. Forman's protected activity. After discussing the 1995 promotion decision with Dr. Forman, Freudenheim indicated that he thought that "it looks like [Dr. Forman] is being handled prejudicially (either because he previously sued, or because he's not part of some vague team concept), and I suspect that [the Smithsonian] will not win this one if it goes into formal legal processes." As noted, Freudenheim later qualified this statement, explaining that it was based solely on information Dr. Forman had provided and that he did not have "all of the pertinent information at [his] disposal at the time [he] expressed view." Although Dr. Forman is entitled on summary judgment to have Freudenheim's initial letter credited, see Reeves, 530 U.S. at 150, it is insufficient to establish causation because it is not based on personal knowledge, but rather is mere speculation. See McGill, 203 F.3d at 846; see also Fed. R. Civ. P. 56(e).

Accordingly, because Dr. Forman has established a prima facie case of retaliation, and the Smithsonian has failed to meet its burden of production to set forth a legitimate, nonretaliatory reason for Hoffman's failure to act on Dr. Forman's complaint, we reverse the grant of summary judgment on Forman's 1995 retaliation claim; we otherwise affirm. Dr. Forman's only other contention, that the district court abused its discretion in denying him further discovery is unpersuasive in view of the wide scope of discretion accorded to the district court. See, e.g., United States v. Microsoft Corp., 253 F.3d 34, 100-01 (D.C. Cir. 2001) (per curiam), cert. denied on other grounds, 70 U.S.L.W. 3107 (U.S. Oct. 9, 2001); Carey Canada, Inc. v. Columbia Cas. Co., 940 F.2d 1548, 1559 (D.C. Cir. 1991).