GINSBURG, Circuit Judge:

The Business Roundtable and the Chamber of Commerce of the United States, each of which has corporate members that issue publicly traded securities, petition for review of Exchange Act Rule 14a-11. The rule requires public companies to provide shareholders with information about, and their ability to vote for, shareholder-nominated candidates for the board of directors. The petitioners argue the Securities and Exchange Commission promulgated the rule in violation of the Administrative Procedure Act, 5 U.S.C. § 551 et seq., because, among other reasons, the Commission failed adequately to consider the rule's effect upon efficiency, competition, and capital formation, as required by Section 3(f) of the Exchange Act and Section 2(c) of the Investment Company Act of 1940, codified at 15 U.S.C. §§ 78c(f) and 80a-2(c), respectively. For these reasons and more, we grant the petition for review and vacate the rule.

I. Background

The proxy process is the principal means by which shareholders of a publicly traded corporation elect the company's board of directors. Typically, incumbent directors nominate a candidate for each vacancy prior to the election, which is held at the company's annual meeting. Before the meeting the company puts information about each nominee in the set of "proxy materials"—usually comprising a proxy voting card and a proxy statement—it distributes to all shareholders. The proxy statement concerns voting procedures and background information about the board's nominee(s); the proxy card enables shareholders to vote for or against the nominee(s) without attending the meeting. A shareholder who wishes to nominate a different candidate may separately file his own proxy statement and solicit votes from shareholders, thereby initiating a "proxy contest."

Rule 14a-11 provides shareholders an alternative path for nominating and electing directors. Concerned the current process impedes the expression of shareholders' right under state corporation laws to nominate and elect directors, the Commission proposed the rule, see Facilitating Shareholder Director Nominations, 74 Fed.Reg. 29,024, 29,025-26 (2009) (hereinafter Proposing Release), and adopted it with the goal of ensuring "the proxy process functions, as nearly as possible, as a replacement for an actual in-person meeting of shareholders," 75 Fed.Reg. 56,668, 56,670 (2010) (hereinafter Adopting Release). After responding to public comments, the Commission amended the proposed rule and, by a vote of three to two, adopted Rule 14a-11. Id. at 56,677. The rule requires a company subject to the Exchange Act proxy rules, including an investment company (such as a mutual fund) registered under the Investment Company Act of 1940(ICA), to include in its proxy materials "the name of a person or persons nominated by a [qualifying] shareholder or group of shareholders for election to the board of directors." Id. at 56,682-83, 56,782/3.

To use Rule 14a-11, a shareholder or group of shareholders must have continuously held "at least 3% of the voting power of the company's securities entitled to be voted" for at least three years prior to the date the nominating shareholder or group submits notice of its intent to use the rule, and must continue to own those securities through the date of the annual meeting. Id. at 56,674-75. The nominating shareholder or group must submit the notice, which may include a statement of up to 500 words in support of each of its nominees, to the Commission and to the company. Id. at 56,675-76. A company that receives notice from an eligible shareholder or group must include the proffered information about the shareholder(s) and his nominee(s) in its proxy statement and include the nominee(s) on the proxy voting card. Id. at 56,676/1.

The Commission did place certain limitations upon the application of Rule 14a-11. The rule does not apply if applicable state law or a company's governing documents "prohibit shareholders from nominating a candidate for election as a director." Id. at 56,674/3. Nor may a shareholder use Rule 14a-11 if he is holding the company's securities with the intent of effecting a change of control of the company. Id. at 56,675/1. The company is not required to include in its proxy materials more than one shareholder nominee or the number of nominees, if more than one, equal to 25 percent of the number of directors on the board. Id. at 56,675/2.*

The Commission concluded that Rule 14a-11 could create "potential benefits of improved board and company performance and shareholder value" sufficient to "justify [its] potential costs." Id. at 56,761/1. The agency rejected proposals to let each company's board or a majority of its shareholders decide whether to incorporate Rule 14a-11 in its bylaws, saying that "exclusive reliance on private ordering under State law would not be as effective and efficient" in facilitating shareholders' right to nominate and elect directors. Id. at 56,759-60. The Commission also rejected the suggestion it exclude investment companies from Rule 14a-11. Id. at 56,684/1. The two Commissioners voting against the rule faulted the Commission on both theoretical and empirical grounds. See Commissioner Troy A. Paredes, Statement at Open Meeting to Adopt the Final Rule Regarding "Proxy Access" (Aug. 25, 2010), available at http://www.sec.gov/news/ speech/2010/spch082510tap.htm; Commissioner Kathleen L. Casey, Statement at Open Meeting to Adopt Amendments Regarding "Proxy Access" (Aug. 25, 2010), available at http://www.sec.gov/news/ speech/2010/spch082510klc.htm (faulting Commission for failing to act "on the basis of empirical data and sound analysis").

The petitioners sought review in this court in September 2010. The Commission then stayed the final rule, which was to have been effective on November 15, pending the outcome of this case.

II. Analysis

Under the APA, we will set aside agency action that is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A). We must assure ourselves the agency has "examine[d] the relevant data and articulate[d] a satisfactory explanation for its action including a rational connection between the facts found and the choices made." Motor Vehicle Mfrs. Ass'n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983) (internal quotation marks omitted). The Commission also has a "statutory obligation to determine as best it can the economic implications of the rule." Chamber of Commerce v. SEC, 412 F.3d 133, 143 (D.C.Cir.2005).

Indeed, the Commission has a unique obligation to consider the effect of a new rule upon "efficiency, competition, and capital formation," 15 U.S.C. §§ 78c(f), 78w(a)(2), 80a-2(c), and its failure to "apprise itself—and hence the public and the Congress—of the economic consequences of a proposed regulation" makes promulgation of the rule arbitrary and capricious and not in accordance with law. Chamber of Commerce, 412 F.3d at 144; Pub. Citizen v. Fed. Motor Carrier Safety Admin., 374 F.3d 1209, 1216 (D.C.Cir.2004) (rule was arbitrary and capricious because agency failed to consider a factor required by statute).

The petitioners argue the Commission acted arbitrarily and capriciously here because it neglected its statutory responsibility to determine the likely economic consequences of Rule 14a-11 and to connect those consequences to efficiency, competition, and capital formation. They also maintain the Commission's decision to apply Rule 14a-11 to investment companies is arbitrary and capricious.

We agree with the petitioners and hold the Commission acted arbitrarily and capriciously for having failed once again—as it did most recently in American Equity Investment Life Insurance Company v. SEC, 613 F.3d 166, 167-68 (D.C.Cir.2010), and before that in Chamber of Commerce, 412 F.3d at 136—adequately to assess the economic effects of a new rule. Here the Commission inconsistently and opportunis-tically framed the costs and benefits of the rule; failed adequately to quantify the certain costs or to explain why those costs could not be quantified; neglected to support its predictive judgments; contradicted itself; and failed to respond to substantial problems raised by commenters. For these and other reasons, its decision to apply the rule to investment companies was also arbitrary. Because we conclude the Commission failed to justify Rule 14a-11, we need not address the petitioners' additional argument the Commission arbitrarily rejected proposed alternatives that would have allowed shareholders of each company to decide for that company whether to adopt a mechanism for shareholders' nominees to get access to proxy materials.

A. Consideration of Economic Consequences

In the Adopting Release, the Commission predicted Rule 14a-11 would lead to "[d]irect cost savings" for shareholders in part due to "reduced printing and postage costs" and reduced expenditures for advertising compared to those of a "traditional" proxy contest. 75 Fed.Reg. at 56,756/2. The Commission also identified some intangible, or at least less readily quantifiable, benefits, principally that the rule "will mitigate collective action and free-rider concerns," which can discourage a shareholder from exercising his right to nominate a director in a traditional proxy contest, id., and "has the potential of creating the benefit of improved board performance and enhanced shareholder value," id. at 56,761/1. The Commission anticipated the rule would also impose costs upon companies and shareholders related to "the preparation of required disclosure, printing and mailing . . ., and [to] additional solicitations," id. at 56,768/3, and could have "adverse effects on company and board performance," id. at 56,764/3, for example, by distracting management, id. at 56,765/1. The Commission nonetheless concluded the rule would promote the "efficiency of the economy on the whole," and the benefits of the rule would "justify the costs" of the rule. Id. at 56,771/3.

The petitioners contend the Commission neglected both to quantify the costs companies would incur opposing shareholder nominees and to substantiate the rule's predicted benefits. They also argue the Commission failed to consider the consequences of union and state pension funds using the rule and failed properly to evaluate the frequency with which shareholders would initiate election contests.

1. Consideration of Costs and Benefits

In the Adopting Release, the Commission recognized "company boards may be motivated by the issues at stake to expend significant resources to challenge shareholder director nominees." 75 Fed.Reg. at 56,770/2. Nonetheless, the Commission believed a company's solicitation and campaign costs "may be limited by two factors": first, "to the extent that the directors' fiduciary duties prevent them from using corporate funds to resist shareholder director nominations for no good-faith corporate purpose," they may decide "simply [to] include the shareholder director nominees. . . in the company's proxy materials"; and second, the "requisite ownership threshold and holding period" would "limit the number of shareholder director nominations that a board may receive, consider, and possibly contest." Id. at 56,770/2-3.

The petitioners object that the Commission failed to appreciate the intensity with which issuers would oppose nominees and arbitrarily dismissed the probability that directors would conclude their fiduciary duties required them to support their own nominees. The petitioners also argue it was arbitrary for the Commission not to estimate the costs of solicitation and campaigning that companies would incur to oppose candidates nominated by shareholders, which costs commenters expected to be quite large. The Chamber of Commerce submitted a comment predicting boards would incur substantial expenditures opposing shareholder nominees through "significant media and public relations efforts, advertising . . ., mass mailings, and other communication efforts, as well as the hiring of outside advisors and the expenditure of significant time and effort by the company's employees." Id. at 56,770/1. It pointed out that in recent proxy contests at larger companies costs "ranged from $14 million to $4 million" and at smaller companies "from $3 million to $800,000." Id. In its brief the Commission maintains it did consider the commenters' estimates of the costs, but reasonably explained why those costs "may prove less than these estimates."

We agree with the petitioners that the Commission's prediction directors might choose not to oppose shareholder nominees had no basis beyond mere speculation. Although it is possible that a board, consistent with its fiduciary duties, might forgo expending resources to oppose a shareholder nominee—for example, if it believes the cost of opposition would exceed the cost to the company of the board's preferred candidate losing the election, discounted by the probability of that happening—the Commission has presented no evidence that such forbearance is ever seen in practice. To the contrary, the American Bar Association Committee on Federal Regulation of Securities commented:

as boards now do in traditional proxy contests. Letter from Jeffrey W. Rubin, Chair, Comm. on Fed. Regulation of Secs., Am. Bar Ass'n, to SEC 35 (August 31, 2009), available at http://www.sec.gov/ comments/s7-10-09/s71009-456.pdf.

The Commission's second point, that the required minimum amount and duration of share ownership will limit the number of directors nominated under the new rule, is a reason to expect election contests to be infrequent; it says nothing about the amount a company will spend on solicitation and campaign costs when there is a contested election. Although the Commission acknowledged that companies may expend resources to oppose shareholder nominees, see 75 Fed.Reg. at 56,770/2, it did nothing to estimate and quantify the costs it expected companies to incur; nor did it claim estimating those costs was not possible, for empirical evidence about expenditures in traditional proxy contests was readily available. Because the agency failed to "make tough choices about which of the competing estimates is most plausible, [or] to hazard a guess as to which is correct," Pub. Citizen, 374 F.3d at 1221, we believe it neglected its statutory obligation to assess the economic consequences of its rule, see Chamber of Commerce, 412 F.3d at 143.

The petitioners also maintain, and we agree, the Commission relied upon insufficient empirical data when it concluded that Rule 14a-11 will improve board performance and increase shareholder value by facilitating the election of dissident shareholder nominees. See 75 Fed.Reg. at 56,761-62. The Commission acknowledged the numerous studies submitted by commenters that reached the opposite result. Id. at 56,762/2 & n. 924. One commenter, for example, submitted an empirical study showing that "when dissident directors win board seats, those firms underperform peers by 19 to 40% over the two years following the proxy contest." Elaine Buckberg, NERA Econ. Consulting, & Jonathan Macey, Yale Law School, Report on Effects of Proposed SEC Rule 14a-11 on Efficiency, Competitiveness and Capital Formation 9 (2009), available at www. nera.com/upload/Buckberg_Macey_ Report_FINAL.pdf. The Commission completely discounted those studies "because of questions raised by subsequent studies, limitations acknowledged by the studies' authors, or [its] own concerns about the studies' methodology or scope." 75 Fed. Reg. at 56,762-63 & n. 926-28.

The Commission instead relied exclusively and heavily upon two relatively unpersuasive studies, one concerning the effect of "hybrid boards" (which include some dissident directors) and the other concerning the effect of proxy contests in general, upon shareholder value. Id. at 56,762 & n. 921 (citing Chris Cernich et al., IRRC Inst. for Corporate Responsibility, Effectiveness of Hybrid Boards (May 2009), available at www.irrcinstitute.org/ pdf/IRRC_05_09_EffectiveHybridBoards. pdf, and J. Harold Mulherin & Annette B. Poulsen, Proxy Contests & Corporate Change: Implications for Shareholder Wealth, 47 J. Fin. Econ. 279 (1998)). Indeed, the Commission "recognize[d] the limitations of the Cernich (2009) study," and noted "its long-term findings on shareholder value creation are difficult to interpret." Id. at 56,760/3 n. 911. In view of the admittedly (and at best) "mixed" empirical evidence, id. at 56,761/1, we think the Commission has not sufficiently supported its conclusion that increasing the potential for election of directors nominated by shareholders will result in improved board and company performance and shareholder value, id. at 56,761/1; see id. at 56,761/3.

Moreover, as petitioners point out, the Commission discounted the costs of Rule 14a-11—but not the benefits—as a mere artifact of the state law right of shareholders to elect directors. For example, with reference to the potential costs of Rule 14a-11, such as management distraction and reduction in the time a board spends "on strategic and long-term thinking," the Commission thought it "important to note that these costs are associated with the traditional State law right to nominate and elect directors, and are not costs incurred for including shareholder nominees for director in the company's proxy materials." Id. at 56,765/1-2. As we have said before, this type of reasoning, which fails to view a cost at the margin, is illogical and, in an economic analysis, unacceptable. See Chamber of Commerce, 412 F.3d at 143 (rejecting Commission's argument that rule would not create "costs associated with the hiring of staff because boards typically have this authority under state law," and assuming that "whether a board is authorized by law to hire additional staff in no way bears upon" the question whether the rule would "in fact cause the fund to incur additional staffing costs").

2. Shareholders with Special Interests

The petitioners next argue the Commission acted arbitrarily and capriciously by "entirely fail[ing] to consider an important aspect of the problem," Motor Vehicle Mfrs. Ass'n, 463 U.S. at 43, 103 S.Ct. 2856, to wit, how union and state pension funds might use Rule 14a-11. Commenters expressed concern that these employee benefit funds would impose costs upon companies by using Rule 14a-11 as leverage to gain concessions, such as additional benefits for unionized employees, unrelated to shareholder value. The Commission insists it did consider this problem, albeit not in haec verba, along the way to its conclusion that "the totality of the evidence and economic theory" both indicate the rule "has the potential of creating the benefit of improved board performance and enhanced shareholder value." 75 Fed. Reg. at 56,761/1. Specifically, the Commission recognized "companies could be negatively affected if shareholders use the new rules to promote their narrow interests at the expense of other shareholders," id. at 56,772/3, but reasoned these potential costs "may be limited" because the ownership and holding requirements would "allow the use of the rule by only holders who demonstrated a significant, long-term commitment to the company," id. at 56,766/3, and who would therefore be less likely to act in a way that would diminish shareholder value. The Commission also noted costs may be limited because other shareholders may be alerted, through the disclosure requirements, "to the narrow interests of the nominating shareholder." Id.

The petitioners also contend the Commission failed to respond to the costs companies would incur even when a shareholder nominee is not ultimately elected. These costs may be incurred either by a board succumbing to the demands, unrelated to increasing value, of a special interest shareholder threatening to nominate a director, or by opposing and defeating such nominee(s). The Commission did not completely ignore these potential costs, but neither did it adequately address them.

Notwithstanding the ownership and holding requirements, there is good reason to believe institutional investors with special interests will be able to use the rule and, as more than one commenter noted, "public and union pension funds" are the institutional investors "most likely to make use of proxy access." Letter from Jonathan D. Urick, Analyst, Council of Institutional Investors, to SEC 2 (January 14, 2010), available at http://www.cii.org/User Files/file/resource% 20center/correspondence/2010/1-14-10% 20Proxy% 20Access% 20Comment% 20Letter.pdf. Nonetheless, the Commission failed to respond to comments arguing that investors with a special interest, such as unions and state and local governments whose interests in jobs may well be greater than their interest in share value, can be expected to pursue self-interested objectives rather than the goal of maximizing shareholder value, and will likely cause companies to incur costs even when their nominee is unlikely to be elected. See, e.g., Detailed Comments of Business Roundtable on the Proposed Election Contest Rules and the Proposed Amendment to the Shareholder Proposal Rules 102 (August 17, 2009), available at http://businessroundtable.org/ uploads/hearings-letters/downloads/BRT_ Comment_Letter_to_SEC_on_File_No_S 7-10-09.pdf ("`state governments and labor unions . . . often appear to be driven by concerns other than a desire to increase the economic performance of the companies in which they invest'" (quoting Leo E. Strine, Jr., Toward a True Corporate Republic: A Traditionalist Response to Bebchuk's Solution for Improving Corporate America, 119 Harv. L.Rev. 1759, 1765 (2006))). By ducking serious evaluation of the costs that could be imposed upon companies from use of the rule by shareholders representing special interests, particularly union and government pension funds, we think the Commission acted arbitrarily.

3. Frequency of Election Contests

In the Proposing Release, the Commission estimated 269 companies per year, comprising 208 companies reporting under the Exchange Act and 61 registered investment companies, would receive nominations pursuant to Rule 14a-11. 74 Fed. Reg. at 29,064/1. In the Adopting Release, however, the Commission reduced that estimate to 51, comprising only 45 reporting companies and 6 investment companies, in view of "the additional eligibility requirements" the Commission adopted in the final version of Rule 14a-11. 75 Fed.Reg. at 56,743/3-56,744/1. (As originally proposed, Rule 14a-11 would have required a nominating shareholder to have held the securities for only one year rather than the three years required in the final rule. See id. at 56,755/1.) In revising its estimate, the Commission also newly relied upon "[t]he number of contested elections and board-related shareholder proposals" in a recent year, which it believed was "a better indicator of how many shareholders might submit a nomination" than were the data upon which it had based its estimate in the Proposing Release. Id. at 56,743/3.

The petitioners argue the Commission's revised estimate unreasonably departs from the estimate used in the Proposing Release, conflicts with its assertion the rule facilitates elections contests, and undermines its reliance upon frequent use of Rule 14a-11 to estimate the amount by which shareholders will benefit from "direct printing and mailing cost savings," id. at 56,756 & n. 872. The petitioners also contend the estimate is inconsistent with the Commission's prediction shareholders will initiate 147 proposals per year under Rule 14a-8, a rule not challenged here.** See id. at 56,677/2.

The Commission was not unreasonable in predicting investors will use Rule 14a-11 less frequently than traditional proxy contests have been used in the past. As Commission counsel pointed out at oral argument, there would still be some traditional proxy contests; the total number of efforts by shareholders to nominate and elect directors will surely be greater when shareholders have two paths rather than one open to them. In any event, the final estimated frequency (51) with which shareholders will use Rule 14a-11 does not clearly conflict with the higher estimate in the Proposing Release (269), or the estimate of proposals under Rule 14a-8 (147), both of which were based upon looser eligibility standards.

In weighing the rule's costs and benefits, however, the Commission arbitrarily ignored the effect of the final rule upon the total number of election contests. That is, the Adopting Release does not address whether and to what extent Rule 14a-11 will take the place of traditional proxy contests. Cf. 75 Fed.Reg. at 56,772/2. Without this crucial datum, the Commission has no way of knowing whether the rule will facilitate enough election contests to be of net benefit. See id. at 56,761/1 (anticipating "beneficial effects" because rule will "mak[e] election contests a more plausible avenue for shareholders to participate in the governance of their company").

We also agree with the petitioners that the Commission's discussion of the estimated frequency of nominations under Rule 14a-11 is internally inconsistent and therefore arbitrary. In discussing its benefits, the Commission predicted nominating shareholders would realize "[d]irect cost savings" from not having to print or mail their own proxy materials. Id. at 56,756/2. These savings would "remove a disincentive for shareholders to submit their own director nominations" and otherwise facilitate election contests. Id. The Commission then cited comment letters predicting the number of elections contested under Rule 14a-11 would be quite high. See id. at 56,756/3 n. 872. One of the comments reported, based upon the proposed rule and a survey of directors, that approximately 15 percent of all companies with shares listed on exchanges, that is, "hundreds" of public companies, expected a shareholder or group of shareholders to nominate a director using the new rule. Letter from Kenneth L. Altman, President, The Altman Group, Inc., to SEC 3 (January 19, 2010), available at http:// www.sec.gov/comments/s7-10-09/s71009-605.pdf. Thus, the Commission anticipated frequent use of Rule 14a-11 when estimating benefits, but assumed infrequent use when estimating costs. See, e.g., supra at 1150 (SEC asserted solicitation and campaign costs would be minimized because of limited use of the rule).

B. Application of the Rule to Investment Companies

Because the rule is arbitrary and capricious on its face, it is assuredly invalid as applied specifically to investment companies. Lest the Commission on remand apply to investment companies a newly justified version of the rule, however, only to be met in court again by valid objections, we think it prudent to take up the more serious of the concerns posed by investment companies but left unaddressed by the Commission.

Investment companies, such as mutual funds, pool investors' assets to purchase securities and other financial instruments. They are subject to different requirements, providing protections for shareholders not applicable to publicly traded stock companies. See 75 Fed.Reg. at 56,684/2. For example, the ICA requires shareholders' approval of certain key decisions. See 15 U.S.C. § 80a-13(a) (majority vote needed to change fund's "subclassification," i.e., among open-end, closed-end, or diversified).

One "investment adviser" typically manages a family of mutual funds, known as a "complex." The boards of the funds in a complex are generally organized in one of two ways: Either there is a "unitary board," comprising one group of directors who sit as the board of every fund in the complex, or there are "cluster boards," comprising two or more groups of directors, with each group overseeing a different set of funds within the complex. A recent survey showed 81 percent of responding complexes have a unitary board and 15 percent a cluster structure. In either case, boards typically address the business of multiple funds in a single meeting.

We agree with the petitioners and amici curiae, the Investment Company Institute and Independent Directors Council, that the Commission failed adequately to address whether the regulatory requirements of the ICA reduce the need for, and hence the benefit to be had from, proxy access for shareholders of investment companies, and whether the rule would impose greater costs upon investment companies by disrupting the structure of their governance. Although the Commission acknowledged the significant degree of "regulatory protection" provided by the ICA, it did almost nothing to explain why the rule would nonetheless yield the same benefits for shareholders of investment companies as it would for shareholders of operating companies. For example, the Commission justified applying Rule 14a-11 to investment companies in part on the ground that "investment company boards . . . have significant responsibilities in protecting shareholder interests, such as the approval of advisory contracts," 75 Fed.Reg. at 56,684/1-2, but did not consider that the ICA already requires shareholder approval of advisory contracts. See 15 U.S.C. § 80a-15(a). Cf. Am. Equity, 613 F.3d at 178-79 (SEC's analysis was "incomplete because it fail[ed] to determine whether, under the existing regime, sufficient protections existed" to advance the stated benefits of the rule and to promote efficiency).

The Commission also failed to deal with the concern that Rule 14a-11 will impose greater costs upon investment companies by disrupting the unitary and cluster board structures with the introduction of shareholder-nominated directors who sit on the board of a single fund, thereby requiring multiple, separate board meetings and making governance less efficient. See, e.g., Letter from Jeffrey W. Rubin, Chair, Comm. on Fed. Regulation of Secs., Am. Bar Ass'n, to SEC 61-62 (August 31, 2009), available at http://www.sec.gov/ comments/s7-10-09/s71009-456.pdf (predicting application of rule to investment companies will "eliminat[e] any benefit to the `cluster board' structure," which structure "creates[s] many efficiencies, such as concurrent meetings among several or many different investment companies that have similar interests, issues and economies of scale that result from being part of a family of funds"). The Commission acknowledged "the election of a shareholder director nominee may . . . increase costs and potentially decrease the efficiency of the boards." 75 Fed.Reg. at 56,684/3. Nonetheless, it did not consider these as incremental costs of the rule because it erroneously attributed them to "the State law right to nominate and elect directors," perhaps a necessary but not a sufficient cause, and dismissed them with the conclusory assertion that the "policy goals and the benefits of the rule justify these costs." Id.

The Commission did acknowledge that it believed costs would be lower for investment companies because their shareholders are mostly retail investors; would be less likely to meet the three-year holding requirement; and would have fewer opportunities to use the rule because some investment companies may under state law elect not to hold annual meetings. Id. at 56,685/1. It also determined disruptions to unitary and cluster boards could be mitigated through the use of confidentiality agreements "in order to preserve the status of confidential information regarding the fund complex." Id.

These observations do not adequately address the probability the rule will be of no net benefit as applied to investment companies. First, the Commission failed to consider that less frequent use of the rule by shareholders of investment companies also reduces the expected benefits of the rule. Second, the Commission's assertion that confidentiality agreements could meaningfully reduce costs is an ipse dixit, without any evidentiary support and unresponsive to the contrary claim of investment companies that confidentiality agreements would be no solution because the shareholder-nominated director would have no fiduciary duty to other funds in the complex and, in any event, could not be "legally obliged" to enter into a confidentiality agreement.

Finally, the Commission observed that "any increased costs and decreased efficiency of an investment company's board as a result of the fund complex no longer having a unitary or cluster board would occur, if at all, only in the event that investment company shareholders elect the shareholder nominee." Id. at 56,684/3. The Commission's point was that shareholders might benefit from getting proxy materials "making [them] aware of the company's views on the perceived benefits of a unitary or cluster board and the potential for increased costs and decreased efficiency if the shareholder nominees are elected." Id. at 56,684-85. And so they might, but this rationale is tantamount to saying the saving grace of the rule is that it will not entail costs if it is not used, or at least not used successfully to elect a director. That is an unutterably mindless reason for applying the rule to investment companies.

III. Conclusion

For the foregoing reasons, we hold the Commission was arbitrary and capricious in promulgating Rule 14a-11. Accordingly, we have no occasion to address the petitioners' First Amendment challenge to the rule. The petition is granted and the rule is hereby

Vacated.