Citizens United and the shareholder protection gambit

Larry Ribstein —  25 October 2010

Last January in Citizens United the Supreme Court delivered a blow to the opponents of corporate speech by enabling corporations to spend directly on political campaigns rather than relying on PACs and lobbying. A majority of the Court concluded that public debate could be best promoted by protecting all speech, regardless of speaker. 

A sizable chunk of America thinks the Supreme Court has unleashed a corporate monster that will drown out the rest of the populace.  For example, a hysterical MarketWatch column (HT Bainbridge) asserted that “[u]nder this system, the game is over. Our democracy is dead.”  This pundit figured that corporations could use their spare cash to buy any number of elections, apparently not understanding that firms need this cash for other things.  He notes that money is flooding into “right-wing groups” but neglects to count up the union cash flowing in the opposite direction.

In any event, what can CU’s opponents do about that pesky Supreme Court opinion?  Maybe try an end-run:  limit corporate speech in the name of “shareholder protection.” And so we have the Shareholder Protection Act (SPA) currently pending in Congress.  But what I’ll call the “shareholder protection gambit” (hereafter, SPG) is unlikely to work. 

To begin with, let’s be clear that it’s not really about shareholder protection.  The “purpose and summary” supporting the SPA makes this transparent in its opening sentences: 

The [Citizens United] ruling invalidated longstanding provisions in U.S. election laws and raised fresh concerns about corporate influence in our political process. To address those concerns, the Shareholder Protection Act gives shareholders of public companies the right to vote on the company’s annual budget for political expenditures.

In other words, the bill is concerned about “corporate influence.”  Shareholders’ rights are intended “to address those concerns.”  But Citizens United explicitly rejected this “corporate influence,” or anti-distortion, rationale for corporate speech limitations.  

The shareholder protection rationale not only differs from, but is directly contrary to, the concern that activates the opposition to Citizens UnitedCU opponents worry that corporate political activity will empower corporations by flooding the political marketplace with pro-corporate money.  Yet the SPA is supposedly intended to ensure that lobbying serves corporate interests.

Whether or not shareholders’ protectors are seeking to restrict corporate speech, that’s what they’re likely to accomplish.  The shareholder rights the SPA would create cut at the core of corporate efficiency.  The main point of the corporate form is to “lock in” corporate resources under strong managerial control.  Enabling shareholders to second-guess specific managerial decisions, and introducing political debate into shareholder and director meetings, is an effective way to hobble corporations. 

Whatever its intent, the SPG, including its manifestation in the Shareholder Protection Act, is dubious policy. Here it’s useful to examine recent work by prominent and bona fide defenders of the shareholder protection gambit, John Coates and Lucian Bebchuk and Robert Jackson.  Neither has a coherent agency cost theory of corporate managers’ investments in corporate political activity — that is, why they would choose to abuse their power in this particular way.  Perhaps they can’t take it all out in pay, but surely there are more satisfying perks.  Nor do these writers show how managers’ use of corporate money for political expenditures presents a more serious agency cost problem than many other exercises of management power.  The most that Bebchuk-Jackson can come up with is that corporate political speech has “expressive significance.” But the idea that shareholders generally, many of whom now invest through pension and mutual funds, care deeply about the political expenditures of individual firms in their portfolios needs far more support than B-J’s brief hypothesizing.

Coates at least has some data.  He lists a number of hypotheses regarding corporate political activity, including that it helps the corporation and constitutes efficient compensation.  However, his his tests are not strong enough to reject non-agency-cost explanations.Coates shows that the existence and amount of firms’ corporate political activity negatively correlates with (1) shareholder democracy variables and (2) corporate value as measured by Tobin’s Q.  But (1) could equally support proponents of strong management control who assert that too much shareholder democracy is inefficient.  The correlation with corporate value could mean, among other things, that firms that are hurt most by regulation need to spend more to get the government off their backs.    

There are several other problems with the shareholder protection gambit.  Its advocates can’t explain why it is so important to protect shareholders from managers running amuck with campaign expenses but not to be concerned about corporate payments to lobbyists or managers funding PACs, both of which are exempted from the Act. Perhaps Congress does not want to completely turn off the corporate spigot.  Maybe politicians would prefer to insulate themselves from corporate-funded opposition when they reneg on their promises.  But it is not clear what these distinctions have to do with shareholder protection.

And it’s odd that after all of the literature and commentary on the problems of shareholder voting, including the fundamental free-rider problem, the distinct interests of institutional investors and activist hedge funds and concerns with “empty” voting, as well as skepticism about the value of ever-increasing disclosure to already-inundated shareholders, that informed shareholder voting should be seen as the means of liberating shareholders to express themselves.

Finally, the SPG faces serious constitutional problems following CU. The Court in Citizens United rejected the “shareholder protection” argument for banning corporate speech, noting among other things that “the procedures of corporate democracy” can protect against any abuse and that “the remedy is not to restrict speech but to consider and explore other regulatory mechanisms. The regulatory mechanism here, based on speech, contravenes the First Amendment.” This language suggests any shareholder protection defense of corporate speech regulation faces a steep uphill climb. 

The SPA is unlikely to survive this climb.  Consider the restrictions on speech it seeks to impose:

  • Majority shareholder authorization of “specific” expenditures a year in advance, thereby shackling effective corporate responses to constantly shifting legislative activity.
  • Unprecedented (for the securities laws) treble damage “fiduciary duty” liability for unauthorized expenditures.  This is likely to have a significant deterrent effect on pro-shareholder political activity, since risks to managers would dwarf any potential gains from speech that serves corporate interests.
  • Extensive quarterly and annual disclosures.
  • Significant federal regulation of formerly state-controlled corporate governance issues.
  • The SPA would apply only to corporations, with no equivalent constraints on unions.

Although the Supreme Court might tolerate some shareholder-protection regulation of speech, its reasoning in CU suggests it will not be receptive to provisions like this.  CU stressed the social value of speech.  This doesn’t mean the shareholders have to subsidize anything corporate managers want to say, but it does mean that the Court will want more than just a shareholder-protection fig leaf to justify a massive increase in the burdens of corporate speech.

The SPA will not strengthen corporate governance, but rather cut corporate political activities back to the lobbying and PACs permitted under pre-Citizens United law.  In other words, it’s designed to reverse CU.  The Court is not likely to miss this fact.

Larry Ribstein


Professor of Law, University of Illinois College of Law

3 responses to Citizens United and the shareholder protection gambit


    Thanks to Larry for noting my paper. I think, though, his analysis of its implications is wrong.

    While I do not purport to claim to have proof of caution in the paper — bad corporate governance (CG) causing corporate political activity (CPA) — I do think that’s the most natural interpretation of the results. Larry’s suggested alternatives are in fact rejected by the data. In multivariate regressions, I control for industry and firm size, as well as the degree to which a firm is dependent on government procurement and imports. With those (and other) controls, the correlation between CG and CPA persist (and in fact strengthen). Those controls take out of the regressions most of the ways in which regulation (actual or threatened) might correlate with CPA. The persisting correlation thus must be attributable to something else. The most natural default would be the simple agency cost story: managers at firms with less shareholder pressure (weaker CG) spend more on politics because they like doing so even if it’s bad for shareholders.

    For many (on all sides of the political spectrum) the idea that CPA might not be “worth it” seems counterintuitive — surely CPA can benefit a firm (regardless of whether it’s good for the public). But the test of shareholder-orientation is not simply whether spending some money on CPA will produce a benefit. Rather, it is whether that is the best use of shareholder money. If Apple has a billion in cash, it can return it to shareholders (dividend, buyback), spend it on a new gadget (which may or may not be positive NPV), increase CEO pay, buy a jet, give it to charity, or spend it on CPA. The benchmark is what is the best use — not whether it might produce *some* benefit, but more of a benefit than the other uses. Thus, CPA as a sign of agency costs is perfectly consistent with findings that CPA can produce beneficial (de)regulation.

    Larry argues that no one has distinguished CPA from lobbying — but in fact, I do, in my Congressional testimony on the DISCLOSE ACT (a different bill from the SPA), available here:

    The constitutional arguments would require much more time and space, I commend readers to Bebchuk and Jackson as a starting point — they present in detail reasons why shareholder governance of CPA would be constitutional. Larry doesn’t reply to their analysis.

    Finally, unions are already subject to a more stringent form of regulation than even the SPA: they must permit workers to individually opt out of any use of dues for political purposes — the equivalent of letting individual shareholders choose to take a dividend in lieu of political expenditures.


      Thanks John. I understand that there was an industry control for the correlation with CG but I didn’t see one for firm value. That could be significant if, as I suggest in the post, lobbying is caused by regulation that reduces the value of firms in a particular industry. Also, my point about lobbying was directed at the SPA rather than your paper. Finally, I agree that CPA for the firm may be suboptimal, but surely that is true of many other management acts. I don’t suggest there are no agency costs regarding CPA — only that from a corporate governance standpoint it cannot be distinguished from other management acts. With respect to Bebchuk & Jackson, although I don’t specifically respond to their arguments regarding constitutionality, my response is inherent in the point I do make in the post that the Court would not tolerate the Act’s restrictions on speech.


    A more potent approach for opponents of Citizens United would be to try to “mute” corporations under state, not federal law. If corporations really do have their attributes defined by state law–CTS and Dartmouth College–it would seem that states could delete the attribute of the powere to make political contributions under state corporation statutes. That is unlikely politically and would raise an interesting con law question in its own right: once created, must corporations be endowed with certain traits as a matter of federal con law or do states still determine corporate attributes?