Skip to main content

Dissenting Statement at an Open Meeting to Adopt the “Pay Ratio” Rule

Commissioner Daniel M. Gallagher

Aug. 5, 2015

Thank you, Chair White. Before getting started, I would like to recognize the staff who toiled at the grindstone of this rule. In particular, I note that Felicia Kung and John Fieldsend were the core of the Conflict Minerals team, and picked up this rule to carry the adopting release over the finish line. Sorry that we keep throwing you these softballs, maybe we can get you something a bit more complex and controversial to work on next. I would also like to recognize the other contributors from CorpFin, as well as those from the Office of the General Counsel for their work on this rule. Finally, I would like to recognize Simona Mola-Yost, Anzhela Knyazeva, Vlad Ivanov, and former staffer Peter Iliev from DERA, along with the rest of the DERA team.

Despite the staff’s hard efforts, it should come as no surprise that I cannot support the rule today. The release does an impressive job of creating out of whole cloth a rationale for this rule: that it could help inform investors in their oversight of executive compensation, including say-on-pay votes.[1] But, to steal a line from Justice Scalia, this is pure applesauce.[2] The purpose of this rule is not to inform a reasonable investor’s voting or investment decision.[3] The AFL-CIO, which lobbied for the rule’s inclusion in Dodd-Frank, has explained for us its true purpose: “Disclosing this pay ratio will shame companies into lowering C.E.O. pay.”[4] And, “They will be embarrassed, and that’s the whole point.”[5] But addressing perceived income inequality is not the province of the securities laws or the Commission. And yet here we are, on the cusp of adopting a nakedly political rule that hijacks the SEC’s disclosure regime to once again effect social change desired by ideologues and special interest groups.

* * *

As an initial matter, we did not have to take up this rule today. Section 953(b) of Dodd-Frank does not have a statutory deadline. There are other, more pressing matters for the Commission, including policymaking with respect to matters actually involving the SEC’s core mission. On the occasion of Dodd-Frank’s fifth anniversary, many have spoken about the law as a fait accompli. But that is far from the truth at the SEC.[6] For example, we have not completed our implementation of Title VII,[7] and we are way past the statutory deadline for adopting mandated rules on stock lending transparency. This rulemaking may well be the most useless of all our Dodd-Frank mandates — that really says something — and it warranted the “caboose” treatment.

Moreover, as I have consistently noted, Dodd-Frank did not repeal our duty to faithfully execute the 75 years of federal securities laws that predated it. Yet progress on much-needed reforms to aspects of the SEC’s core mission, our “day job,” is scant. Two vital projects — equity market structure and corporate disclosure reviews — have yet to be advanced in a meaningful way. Much-needed reform of our transfer agent rules — a rulemaking project for which four Commissioners have expressed support — has yet to be brought forward. Make no mistake, progress on those vital rules is the opportunity cost of prioritizing a constant stream of Dodd-Frank special interest rules that hog the precious bandwidth of the Commission.

* * *

With all that said, I recognize that Section 953(b) of Dodd-Frank is the law of the land. And so, when it was unfortunately prioritized back in 2013, I engaged the issue. I made clear my position that I could potentially support a pay ratio rule proposal that included one critical feature: a limitation on the population of “employees” covered by the rule to full-time employees in the U.S.[8] We had ample authority to do this, and every indication that it would be a good idea. Unfortunately, a majority of the Commission decided not to, and so a fatally-flawed rule was proposed in 2013. Weighing the comments we received on that misguided proposal, it is clear to me that a U.S., full-time definition of employee is not only one option, but the only option that could have made sense of this misbegotten mandate. Our final rule today ignores these comments, with the Commission majority preferring instead to adopt the rule largely as proposed, with only some deck chairs rearranged.

I. The Commission Puts Itself in a Box

The Commission majority goes to extensive lengths to try to rationalize its approach to today’s rule by asserting that it is only carrying out Congress’s express intent, and therefore we had no choice but to adopt the rule before us today. First, the release asserts that the point of the rule is not to name and shame executives, but rather to enhance executive compensation disclosure.[9] Second, the release asserts that, by referring to “all employees,” Congress must have intended it to apply to all employees of all consolidated subsidiaries of the issuer located anywhere in the world, whether full-time or part-time.[10] Third, the release asserts that, by adopting pay ratio as it did, Congress must have believed that there would be benefits to the rule.[11] But neither the law nor the legislative history at any point identifies the purpose of the rule, what it means by “all employees,” or what benefits the rule would achieve. It is totally silent — which the release, to its credit, concedes.[12]

So why make these assertions? The Commission majority needs them, in order to box itself into adopting a pay ratio that includes all employees of the parent or its consolidated subsidiaries located anywhere around the globe, whether full-time, part-time, or seasonal. From there, the release sets forth some different approaches that may be able to lessen the cost burden somewhat, but without fundamentally altering that supposedly inviolable pay ratio calculation.[13] The goal is to convince the public that the Commission is doing all it can reasonably do to help issuers reduce costs — that is, that the Commission is pushing at the sides of the box, while staying within its confines. But the Commission here is a mime; the box is imaginary.[14]

II. What Could the Commission Have Done Instead?

So what could the Commission have done instead? Simple: as I have advocated, in addition to making this the last SEC Dodd-Frank rulemaking (in the year 2020 or so), we could have limited the scope of the rule to full—time, U.S. employees. The Commission has ample definitional authority, interpretive authority, and exemptive authority to do so.

The statute does not include any relevant definition of the term “employee.”[15] So the SEC is free to define “employee” in the way that it chooses. The Commission commonly uses its definitional authority to make sense of legislative language that perhaps didn’t quite hit the mark.[16] Similarly, where a statute is ambiguous, the Commission has the authority to interpret the ambiguous language.[17] Here, the term “employee” is ambiguous.[18] The statutory language “all employees” simply specifies that all members of the class of “employees” must be included; it does not tell us anything about how that class is defined.[19] Thus, the Commission has broad latitude to define or interpret “employee” to mean “persons who are employed by the issuer on a full-time basis within the United States.”[20] The pay ratio would then need to be the ratio of the pay of the median of all of such persons to the pay of the PEO. Or, the Commission could simply have exempted non-U.S., non-full time employees from the scope of the rule; the release concedes that it would be within the scope of the Commission’s Section 36 exemptive authority to do so.[21]

III. What Should the Commission Have Done Instead?

So given that we clearly could have limited the scope of the rule to full-time U.S. employees, the next question is whether we should have done so. Not surprisingly, I believe the answer is an overwhelming yes.

Let us assume for a moment that Congress’s intent in adopting Section 953(b) really was to inform shareholder voting decisions with regard to executive compensation — despite the manifest lack of evidence to support that assertion. The release’s flawed reliance on Congress’s putative intent, as well as its reliance on Congress’s identification of the benefits, means that the Commission has not done the work necessary to justify its policy choices. This seems to be the new trend for the Commission: gesturing in Congress’s vague direction when it comes to a mandated rulemaking’s benefits, reciting what commenters identified as costs, and then claiming that the Commission meaningfully “considered” the benefits and costs of our rule. Here, we have done no such thing. So let’s go through that exercise together, now.

  • First, what benefits are gained through the rule?[22] I see no credible evidence in the record that a reasonable investor would find the pay ratio to be useful. We already provide a wealth of good, comparable data to investors about executive pay, which investors have been successfully using to make their say-on-pay voting decisions.[23] Pay ratio data, by contrast, is low-quality, non-comparable[24] data of use only to certain investors who have idiosyncratic reasons for wanting it. So the benefits of the pay ratio rule adopted today are negative because the additional immaterial information obscures material information about PEO compensation, and because the pay ratio disclosures are highly likely to be misused.
  • Second, would the alternative approach of limiting the rule to U.S., full-time employees have achieved greater benefits? Excluding employees from other countries and part-time or seasonal employees would eliminate a substantial amount of noise in the quality of the median employee data.[25] I’m not willing to concede that even that resulting disclosure would be useful, so let’s just say it would be marginally less useless.[26]
  • Third, do the costs justify the benefits? The initial cost of compliance with the rule before us today is estimated at an astronomical $1.3 billion dollars.[27] The ongoing costs will be $526 million.[28] Excluding non-U.S. and non-full time employees could achieve an astonishing savings of $788 million dollars.[29] The release does no work to explain why its approach achieves $788 million worth of benefits over the alternative.[30]

Given that a majority of the Commission has opted for a hugely expensive rule over a much less expensive rule, with no demonstration that the benefits to be achieved by the more expensive rule justify those additional costs, I can only conclude that there is no reasoned basis for the Commission’s action.

I am also concerned that today’s rule improperly compels corporate speech. We know that the intended purpose of Section 953(b) was to name and shame registrants into reducing CEO pay.[31] And the pay ratio being adopted today will produce so few (if any) benefits for regular shareholders that the information seems likely to be useless for anything but naming-and-shaming. Thus, the rule is not intended to, and does not, produce information in furtherance of a legitimate government purpose. We’ve seen from our Conflict Minerals rule that naming-and-shaming rules can fall afoul of the First Amendment, and so the question is raised in my mind whether pay ratio disclosures are constitutional. Perhaps a narrowly-tailored SEC rule could have cured this latent statutory flaw, but needless to say, that is a path not taken by today’s rule.

* * *

At the end of the day, each of us as Commissioners should hope that we are able to leave the Commission a little better than it was when we first arrived. Thanks to the decisions that have been made about which rules to take up and how to pursue them over my 4-year tenure — decisions unfortunately outside my control — I regretfully can’t say with certainty whether that will be true, for me, with respect to the agency’s policymaking record. With this vote, I will have cast 16 “no” votes on major Commission rulemakings, which I think is a Commission record. Yet this is the hollowest of achievements for me. Those votes represent 16 proposals or final rules that I could not in good conscience approve, because they do not stay true to our mission — rules that fail to protect investors, that degrade our markets, or that inhibit capital formation.[32] This rule highlights the sad fact that, over five years after Dodd-Frank, the Commission is still wandering through the wilderness, and that the voice of one or two minority Commissioners crying out in that wilderness can do little to put us back on the right path.

I obviously cannot support the recommendation before us today, and I have no questions.

[1] Release at 9, 11.

[2] King v. Burwell, No. 14-114, slip op. at 10 (June 25, 2015) (Scalia, J., dissenting).

[3] See TSC Indus., Inc. v. Northway, Inc., 426 U.S. 438 (1976); Basic Inc. v. Levinson, 485 U.S. 224 (1988).

[4] Andrew Ross Sorkin, S.E.C. Has Yet to Set Rule on Tricky Ratio of C.E.O.’s Pay to Workers’, N.Y. Times Dealbook (Jan. 26, 2015) at; Christopher Matthews, The Government Regulation Corporate America Hates Most, Time (Sept. 20, 2013), available at

[5] Peter Schroeder, Disputed rule intended to shame CEOs, The Hill (Feb. 2, 2012), available at

[6] See Chair Mary Jo White, Statement on the Anniversary of the Dodd-Frank Act (July 16, 2015), available at

[7] While I am not particularly fond of Title VII’s over-engineered “solution” to a lack of transparency in the swaps market, at least it does have a nexus to issues that were present in the financial crisis.

[8] I also let it be known that I would have accepted a rule that permitted issuers to elect an alternative “safe harbor” by disclosing a pay ratio of, say, 1000-to-1. This would have given the madding crowd the red meat they desired to name and shame the CEO, while imposing essentially no direct costs on the issuers (and equivalent indirect costs).

[9] Supra note 1.

[10] Release at 45, 50—51.

[11] Release at 176.

[12] E.g., Release at 9. With respect to each of the three assertions:

1. To compensate for the lack of a specific statement of the statutory purpose, the release divines Congress’s intended purpose from the fact that pay ratio was somehow dropped in and among other executive compensation-related provisions in the Dodd-Frank Act. Release at 10. This is a thin reed indeed — just because Dodd-Frank didn’t put pay ratio on Dodd-Frank’s Island of Misfit Toys, otherwise known as Title XV, doesn’t mean that its true purpose is to inform investors.

2. The release even admits as much: “Congress did not expressly state the specific objectives or intended benefits of Section 953(b), and the legislative history of the Dodd-Frank Act also does not expressly state the Congressional purpose underlying Section 953(b).” Release at 9. The release attempts to infer Congress’s intent through the general idea that other securities disclosure requirements are not usually limited to the United States. Release at 68. But there is no judicial presumption in favor of extraterritoriality in securities regulation that supersedes the judicial presumption against extraterritoriality everywhere else in the laws. Rather, the territorial scope of the securities laws is properly dependent upon the usefulness of the information that the disclosures generate for investors. For example, Item 101(d) of Regulation S-X calls for disclosure of financial information broken down by geographic areas — one for the registrant’s home country, and one for all foreign countries in the aggregate, with individual countries included only if material. Mine Safety Disclosures in Item 104 of S-X apply only to the United States (although that intent is apparent in the structure of the statute, which refers to MSHA determinations). Often, disclosures that are not bound to any one particular geographic location are the most useful for investors, and so many of the SEC’s disclosures are geography-neutral. Here, information limited to the United States may actually be more useful for investors. See Section III. It is of course ironic (or perhaps revelatory) that the release does not use investor usefulness the barometer of congressional intent.

3. The mere fact that Congress passed a law does not tell you what the expected benefits are, or whether they are for a proper purpose. Without a legislative statement of the benefits, duly enacted into law — see infra note 22 for an example — we cannot impute any particular understanding to Congress.

[13] Of course, there have been additional restrictions and catches added to these alternative approaches that will make them less useful for the intended purpose of lessening the cost burden.

[14] To illustrate just how illusory the box is, there was at least one significant change made to the definition of employee between the initial draft circulated to the Commission and the rule being adopted today. Yet the release does not cite what source of revelation appeared to the Commission in the weeks leading up to today’s meeting to correct this apparent initial misunderstanding of Congress’s “intent.” One can only assume that the Commission majority is in fact cloaking its policy choices in the mantle of statutory interpretation to avoid scrutiny.

[15] Nor do the extant securities laws provide a useful definition that Congress could be understood to have incorporated.

[16] Ironically, the SEC even does that in this very rule. The language of the statute requires disclosure of the “ratio of the amount described in subparagraph (A) [median pay] to the amount described in paragraph (B) [PEO pay].” Literally, for a median worker who received $50,000 and a PEO who received $2,500,000, this formula would result in a pay ratio of “0.02:1” So the Commission treats “ratio” as ambiguous, making it a “factor rather than a fraction,” to produce a much more intelligible ratio of 50:1. Even more ironically, in the proposed release regarding disclosures of hedging of company stock by employees and directors, the Commission defined “employee” in the statute as including “officers” — despite Exchange Act Rule 12b-2, which, while declining to define “employee,” states that “the term ‘employee’ does not include a director, trustee, or officer.” But not including “officer” within the definition of “employee” didn’t make sense given the statute’s objectives, and so the Commission’s proposed rule redefines “employee” to include “officers.” It’s not clear what changed between Section 953(b) and Section 955 of the Dodd-Frank Act that made the Commission feel so hesitant to define “employee” in one location, and so aggressive about defining it in the other.

[17] The Commission’s interpretation of ambiguity in one of the statutes it administers typically receives Chevron deference, assuming the Chevron factors are met. See generally Chevron U.S.A., Inc. v. National Resources Defense Council, 476 U.S. 837 (1984). For example, just recently, the D.C. Circuit validated the Commission’s determination to interpret the 180-day Dodd-Frank “deadline” for bringing enforcement actions following a Wells Notice as not a deadline at all, but rather a non-binding internal goal. Somewhere in the language stating that the “Commission shall either file an action . . . or provide notice of its intent not to file an action,” the D.C. Circuit found there to be ambiguity, and that any “permissible construction” of the statute by the agency was therefore entitled to deference. See Montford & Co., Inc. v. SEC, No. 14-1126, slip op. (July 10, 2015).

[18] Specifically, as noted, the term has no fixed definition in either the statute or the existing securities laws. The fact that there are so many different types of employees discussed in the release — full-time, part-time, seasonal, domestic, foreign, etc. — is itself evidence of the ambiguity. Even within the U.S., “employee” may have different (although similar) meanings — those who are employed at common law, employees under IRS guidance, employees under Department of Labor guidance (e.g., for Fair Labor Standards Act purposes). Relationships defined as “employment” in other countries around the world may also vary from common U.S. understandings.

[19] Even if one assumed employee to approximate the meaning it usually takes in the United States — common-law employees — the Commission exceeds the scope of that understanding by including within the scope of the class of “employee” individuals who are, in point of fact, not employees, but rather independent contractors — so long as the company “determines” their compensation. Release at 53—54. Despite its protestations that it is only relying on Congress, the Commission majority is actually exercising its interpretive authority to redefine the word employee to suit the ends of the majority. If the Commission had owned up to this decision and adequately justified it, perhaps the rule would be on a better footing. But it does not.

[20] Alternatively, we could have defined all employees “of the issuer” to mean all employees “of the corporate parent and its U.S.-based consolidated subsidiaries” if we wanted to address solely the U.S. issue. This approach could have eliminated the bulk of the expense (i.e., addressed the non-U.S. employee issue, while not addressing the full-time employee issue), while not engaging on the “all employees” debate.

[21] Release at 69. Yet the Commission majority declines to do so, resting on the unpersuasive argument that it wouldn’t be consistent with Congress’s “intent” to do so. I respect that the Commission should not use its exemptive authority to directly contradict Congress. But given that we cannot adequately divine Congress’s intent, we in no way would be “second-guessing” Congress by using our exemptive authority here.

[22] Congress was silent as to the benefits sought to be achieved by the rule. We must therefore “show our own work” — justify for ourselves the benefits. The analysis here is thus not the same as it was in the Conflict Minerals case, where the D.C. Circuit determined that Congress’s finding of benefits relieved the SEC of the duty to independently find them on its own. See NAM v. SEC, No. 13-5252 (Apr. 14, 2014) (SEC entitled to rely on Congress’s determination that the rule’s extensive costs are necessary and appropriate in furthering the goals of peace and security in the Congo). The federal register page therein cited by the D.C. Circuit in turn cites to the congressional findings baked into the law. Section 1502(a) contains a Sense of Congress — a concrete recitation of the benefits to be achieved by conflict minerals disclosure. See Section 1502(a): “SENSE OF CONGRESS ON EXPLOITATION AND TRADE OF CONFLICT MINERALS ORIGINATING IN THE DEMOCRATIC REPUBLIC OF THE CONGO.—It is the sense of Congress that the exploitation and trade of conflict minerals originating in the Democratic Republic of the Congo is helping to finance conflict characterized by extreme levels of violence in the eastern Democratic Republic of the Congo, particularly sexual- and gender-based violence, and contributing to an emergency humanitarian situation therein, warranting the provisions of section 13(p) of the Securities Exchange Act of 1934, as added by subsection (b).” There is nothing comparable here.

[23] This information is quite robust — indeed, some have complained that it is already overwhelming. It also bears on a legitimate question: is the company getting sufficient value for the CEO’s compensation?

[24] Just how useless this pay ratio information will be for investors is highlighted by the release’s insistence that the pay ratio disclosure is only valid for discussions about a particular company. Per the release: “we do not believe that precise conformity or comparability of the pay ratio across companies is necessarily achievable given the variety of factors that could cause the ratio to differ.” Release at 12. Indeed, the approach taken by a majority of the Commission all but ensures that pay ratios of similarly-situated companies will not be comparable with one another. Unfortunately, investors value comparability. The majority of the Commission voting today knows this: this same majority only recently approved a proposed release on pay-for-performance that was incredibly prescriptive, mandating rigid formulas and XBRL tagging where individualization would have permitted companies to tell their pay story. It is unclear to me why comparability is a virtue in pay—for-performance, but can be blithely discarded in this rule. And, despite the warnings in the release, I believe that some will inadvertently misuse or intentionally abuse pay ratio information by comparing companies against one another.

[25] What is included as pay in other countries can be significantly different from the U.S. definition, which the release notes. See Release at 64 (citing comments that noted that companies “in other countries include as ‘compensation’ transportation, food, housing, wedding, birth, education, and phone expenses, as well as profit-sharing arrangements and government provided benefits”). And some countries will be excluded from the pay ratio merely because they have data privacy laws, even though this will alter the calculation of the pay ratio. I completely agree with this determination, but it shows that the pay ratio “intended by Congress” is not inviolate, but rather can be altered where the Commission finds the reason therefor to be compelling. Finally, the release recognizes the need for COLA adjustments for non-U.S. persons because, without adjusting a wage for its value in the country where it is paid, the pay ratio is meaningless — and yet, to prevent undue expense, COLA adjustments are simply optional. All these factors introduce substantial noise into the calculation of the median employee — noise that could be eliminated by limiting ourselves to U.S.-only workers. Similarly, lumping together employees making part-time wages and employees making full-time wages without differentiating the two (i.e., a difference that would be obvious in hourly wage rate versus total compensation) will create noise in the median. This could be eliminated by limiting ourselves to full-time employees.

[26] My preferred approach might admittedly be less reflective of the issuer’s global workforce, whatever that means, but the gains in data quality might at least produce offsetting benefits.

[27] Release at 202-03.

[28] Release at 205.

[29] See Release at 224 ($525 million for non-U.S.) & 217 (up to $263 million for non-full-time). Yet, when we think about it, it’s entirely unsurprising: commenters told us very clearly that large complex multinational organizations generally do not have integrated employee payroll management systems. I heard from one oil and gas producer — which actually did manage, after 5 long years and millions of dollars of IT expenses, to produce a globally integrated system to track royalty payments to employees — that despite this massive expenditure of effort, its systems still wouldn’t be able to produce the data demanded by the SEC’s proposed rule. So the calculation of the median employee must be performed manually, bringing together disparate data sets from across the globe to find the median employee. By contrast, pulling data on employees located within the U.S. would be relatively simple.

[30] Rather, it simply asserts again that the U.S., full-time only approach would too greatly alter Congress’s “intended” approach. E.g., Release at 218. Not to beat the dead horse here, but that’s just not good enough.

[31] It will, of course, be terribly ironic when CEOs demand additional pay to compensate for risk of character assassination that this rule poses, driving CEO pay even higher.

[32] Most frustratingly, many of these rules are rules that could have been made acceptable with a few targeted changes, but nearly invariably my comments were written off.

Return to Top