Article body

Introduction

For the last several years, financial market elites have generally held two opinions that they have assumed, without much inspection, do not conflict. The first is that the voting results in favour of Brexit and Donald Trump, both enormously unpopular in London and New York respectively, may not reflect careful, informed processes undertaken by the electorate.[1] These electoral outcomes may instead be the result of voters who have few incentives to engage in the hard work of policy analysis and information gathering. They may make decisions in the voting booth that provide them with the private pleasures of mood affiliation, tribalism, resentment, and xenophobia.[2] (New York and London may be wrong, but this view has been widely held.) The second opinion is that the voting behaviour of shareholders is of a completely different kind.

Just three decades or so ago, shareholders had very limited voting rights. For nearly all of the twentieth century, managers were largely independent of shareholder voting power, except for the most basic fact that they might lose their offices if their failures became notorious.[3] As an American court noted in 1988, shareholder voting was understood to be “a vestige or ritual of little practical importance.”[4] But not long thereafter, Canada, the United States, and the United Kingdom began to give shareholders increasing voting power over areas that used to fall wholly within board discretion.[5] This has produced an enormous body of academic legal journal articles debating the merits of the shareholder franchise.[6]

The debates among corporate law scholars are based on the assumption that when shareholders come to their voting decisions, it is by way of processes that are rationally calculated to promote certain corporate outcomes. For the most part, this means we assume shareholders vote in ways designed to improve the financial performance of their investments, but it could also mean that some investors vote in ways designed to improve environmental and social outcomes. This assumption of voter rationality is not taken for granted by political scientists. Their research on voting in civic elections shows that votes are often cast for expressive reasons unrelated to voters’ self-interest or desired outcomes. The almost non-existent marginal value of a single vote means that voters feel free to collect and process information in ways that make themselves feel good. This article argues that the empirical literature around shareholder voting shows the same thing: shareholders give their voting rights almost no value, vote in ways that do not reflect the economic performance of the company, do not vote for directors as if the individual in question matters, vote in ways that contradict their economic views (measured by looking at their trading decisions), and their voting decisions are driven by empirically questionable and often deliberately ineffective corporate governance practices.

The first Part of this article will discuss the way that all parties to current academic debates about the shareholder franchise assume shareholders vote rationally. The second Part examines the political science literature on voter ignorance. The third Part examines the political science literature on voter irrationality. The fourth Part looks at whether the well-established political science research is applicable to shareholder voting by examining the empirical evidence on the following topics: shareholders’ valuation of their voting rights, shareholder behaviour in uncontested director elections, shareholder behaviour in majority voting situations, shareholder voting in contested director elections, and shareholder voting on corporate governance matters. In all of these areas, the empirical literature strongly suggests that shareholder voting behaviour resembles the predictions generated by the political science literature. The fifth Part of the article revisits the academic debate to see what remains in light of the evidence on how shareholders process information and vote irrationally. The final section proposes a direction (but only a direction) for reform.

I. The Assumption of Rational Self-Interest in Academic Discussions of Shareholder Voting

Current academic discussions about shareholder voting have a settled format. First, arguments begin by pointing out that, in Easterbrook and Fischel’s now classic formulation, the shareholders’ residual interest in the corporation gives them “the appropriate incentives ... to make discretionary decisions ... The shareholders receive most of the marginal gains and incur most of the marginal costs. They therefore have the right incentives to exercise discretion.”[7]

This formulation of shareholder incentives gives rise to a large and heterogeneous debate over whether the economic interests of the shareholders are actually aligned with those of the corporation. The range of possible conflicts is broad. Do some important firm constituencies’ interests (such as customers, employees, and suppliers) conflict with those of the shareholders?[8] Are shareholders’ economic incentives too short-term?[9] Are there conflicts in which long-term shareholders take advantage of short-term shareholders, or sophisticated shareholders take advantage of unsophisticated shareholders?[10] Do public pension fund managers advance corporate governance agendas designed primarily to appeal to their political masters?[11] Do union pension funds use their power tactically to advance their position at the bargaining table?[12] Do mutual funds reflexively support management to avoid alienating the individuals who decide what fund options will be provided to employees?[13] What about the conflicts between shareholders and debtholders (who, after all, are now the actual suppliers of capital to America’s largest companies)?[14] The important thing to note in this vast literature is the universally shared assumption that the shareholders’ economic interests will drive their voting behaviour and thus corporate outcomes.

The second point made in virtually all discussions around shareholder voting is the collective action problem. This is usually couched in terms that again recall Easterbrook and Fischel’s point that “[w]hen many are entitled to vote, none of the voters expects his votes to decide the contest. Consequently none of the voters has the appropriate incentive at the margin to study the firm’s affairs and vote intelligently.”[15] Evidence in favour of this proposition is adduced, usually in the form of the transparent reluctance of shareholders to vote[16] or engage in related activities.[17] Counter-arguments consist of pointing out that it must make economic sense for some categories of investors (particularly large institutions) to vote if doing so improves the corporate governance, and thus the economic outcomes, of portfolio companies.[18] Alternatively, some authors argue that market institutions such as proxy advisors and mandated disclosure reduce the cost of investors informing themselves.[19] Imposing legal obligations on institutions to vote their shares has also been assumed to render this issue moot (at least by the regulators), as institutions now vote as a matter of course.[20] Again, this literature is underpinned by the belief that shareholder voting, if properly informed, is (or could be) a valuable method of advancing shareholders’ economic interests and improving operational firm outcomes.

The final locus of discussion on voting concerns institutional investors’ economic incentives, which are evaluated and generally found to be wanting. As Professor Coffee observes, the “expected gains from most such governance issues are small, deferred, and received by investors, while the costs are potentially large, immediate, and borne by money managers.”[21] A closely related argument is that institutional money managers worry primarily about their portfolio’s relative performance against other funds or index benchmarks. As gains from shareholder voting are received by their competitors equally, they will choose to devote their resources instead on the activities—such as picking stocks and executing trading strategies—that will allow them to differentiate themselves and attract investment.[22] Counter-arguments involve pointing to classes of investors, such as activist hedge funds, that create economic incentives for themselves to intelligently make use of their (and other shareholders’) voting power.[23] Proxy advisors have similarly created a business model that arguably incentivizes them to give well-informed voting advice.[24] As with the other loci of debate around voting, both sides assume that if the shareholders did devote the resources—or could follow those, like activists, who do devote the resources to inform themselves—then they would vote in value-maximizing ways.

This assumption that voting behaviour is rationally related to economic incentives is the engine that keeps the entire debate running. Both sides take it as an article of faith. As one recent Canadian Securities Administrators (CSA) Staff Notice puts it, “shareholder voting is ... fundamental to, and enhances the quality and integrity of, our public capital markets.”[25] Another Notice puts it this way: “Institutional investors are increasingly engaged in advancing good corporate governance in companies, and one of the ways by which they do so is the exercise of their voting rights.”[26]

For their part, market participants have a firm (but largely unexamined) conviction that voting and economic self-interest are linked. According to a recent survey, eighty per cent of investors “believe that proxy voting increases shareholder value.”[27] This assumption of economically rational voting seems intuitive because so many other investor behaviours are clearly economically rational.[28] It would be impossible to understand (or justify) financial markets if they were not characterized by economically rational behaviour.[29] Why should voting be any different?

II. Ignorance in Political Voting

The empirical literature around political voting in democracies starts in the same place as the literature around shareholder voting, but it goes in an unexpected direction. It finds that the average voter, though rationally self-interested in their personal life, is irrational in the way they vote. This irrationality is actually a function of the self-interest that lies at the heart of economic explanations for human behaviour. Political voters are, in the words of economist Bryan Caplan, “rationally irrational.”[30]

Like shareholders in widely held companies, the voters in a democracy individually have little chance of affecting the outcome of an election. As a result, they perform the same cost-benefit calculation around voting as shareholders. This calculation suggests the marginal value of their vote is insignificant, so they rationally choose not to expend the resources required to properly inform themselves prior to voting.[31] The empirical evidence of this ignorance is both vast and shocking to the uninitiated.[32] In the words of one author of a survey of the literature: “The reality that most voters are often ignorant of even very basic political information is one of the better-established findings of social science. Decades of accumulated evidence reinforce this conclusion.”[33]

Canada has little cause for celebration. Over half of our citizenry believes we elect the prime minister directly.[34] Indeed, a 2016 Ipsos poll found that Canadians were factually wrong on virtually every major hot-button social and economic issue from health spending, to wealth distribution, to the current Muslim population in this country.[35] A recent academic survey of Canadian political research observed that “Canadian voters are no different from voters south of the border. Scholars have repeatedly noted that they are not very informed.”[36]

This political ignorance should not come as a surprise. We are consistently ignorant on matters where our opinions will have little impact and, therefore, we have no incentive to inform ourselves. Over twenty per cent of the residents of the United States do not know that the earth revolves around the sun rather than the reverse.[37] Less than forty per cent of Americans believe in the theory of evolution (the rest either disbelieve it or have no opinion).[38] Over one third of Europeans and Americans believe genetically unmodified foods do not contain genes.[39] A quarter of Europeans believe that eating a genetically modified fruit can result in their bodies’ genes being modified.[40]

Several aspects of this literature on rational ignorance must be made clear. The first is that this ignorance is a function of the individual’s lack of influence over the outcome of an election; it is not a function of the importance of the ultimate outcome of the election. Obviously, it matters to the average citizen what their government does, in the same way that it matters to the average shareholder who occupies the seats in the boardroom. What drives voter ignorance is the insignificant marginal value of that individual’s vote.

The second notable factor is that the motivation of the voter doesn’t matter. It doesn’t matter, for example, whether the voter is strongly self-interested or altruistic. The strongly self-interested will conclude that they have better things to do than invest a great deal of resources in gathering the information needed to vote wisely; the altruistic will conclude that resources spent informing themselves as a voter would be better devoted to activities with a much higher pay-off to the people they are trying to help.[41]

An additional fact that emerges from the literature on political voting is that voter ignorance has not improved as levels of education and the availability of information have increased:

[T]he level of political knowledge in the American electorate has increased only modestly, if at all, since the beginning of mass survey research in the late 1930s. A relatively stable level of ignorance has persisted even in the face of massive increases in educational attainment and an unprecedented expansion in the quantity and quality of information available to the general public at little cost.[42]

This is obviously discouraging news for proponents of the view that better shareholder voting only requires more, and less expensive to consume, information.

III. Irrationality in Political Voting

Thus far, the research into political voting resembles the usual collective action arguments found in academic discussions of shareholder voting. But it is here that the political science data goes in an unexpected direction. Given high and persistent levels of rational ignorance about political matters, how do citizens in a democracy decide to cast their vote? George Akerlof summarizes the choice facing individuals and the way that choice plays out:

[I]nformation is interpreted in a biased way which weights [sic] two ... goals: agents’ desire to feel good about themselves, their activities, and the society they live in, on the one hand, and the need for an accurate view of the world for correct decision making, on the other hand ... [B]ecause any individual’s influence on the public choice outcome is close to zero, each individual has an incentive to choose a model of the world which maximizes his private happiness without any consideration of the consequences for social policy.[43]

This formulation goes beyond the “rational ignorance” of public choice theory. “[R]ational ignorance assumes that people tire of the search for truth, while rational irrationality says that people actively avoid the truth.”[44] What do they pursue instead? They pursue self-expression.[45] They vote in ways that make them feel better about themselves, that confirm and reflect their prejudices, and that help their political “team” score points.[46] They will blame their troubles on harmless scapegoats, punish bearers of bad news for the sin of telling the truth, vote for policies that appear to make their country look “tough,” vote for politicians who are like themselves or who tell them the solutions to problems are simple, refuse to believe news that casts a negative light on “their” politicians, choose news sources that confirm their prejudices, justify a politician’s bad behaviour by investing new importance to other aspects of their personality or actions, and adopt absurd conspiracy theories that make the other side look bad or explain away uncomfortable facts.[47] In short, their political lives are a form of “mood affiliation.”[48]

Professor Ilya Somin calls this behaviour enjoying “the psychic benefits of being a political ‘fan.’”[49] Sports fans invest time gathering information and following their team, not because of any expectation that by doing so they are affecting the outcome of the season, but rather because they find it interesting and enjoy rooting for “their” team. Political fans similarly derive enjoyment from supporting their preferred candidates, parties, or ideologies, and from denigrating the other side.[50] They also benefit from the pleasure of having their pre-existing views validated and from associating with like-minded people with the same objectives.

This view of voting explains why, for example, opposition to immigration is not a function of actual exposure to immigration or labour market competition from immigrants, but rather general xenophobic attitudes toward immigrants.[51] These attitudes seem to produce, rather than derive from, beliefs about the costs and benefits of immigration and even the proportion of immigrants in a country or region.[52]

This view of voting also explains why studies repeatedly find that the most knowledgeable voters tend to be the most biased in their interpretation of new information.[53] Bias in evaluating information increases with higher cognitive ability and stronger ideological views.[54] No matter where voters are on the spectrum, they prefer to talk politics with people who have similar opinions and receive news from sources that align with those views.[55] These are not the actions of people seeking truth, they are the actions of fans rationally pursuing their own peace of mind and sense of vindication. The entire dynamic is underwritten by the fact that “the market has a ‘user fee’ for irrationality, and democracy does not.”[56]

This last statement may seem controversial because every voter obviously has a stake in the quality of government. Indeed, one of the most common assumptions about voting patterns is that voters often cast their votes to advance their own self-interest.[57] “They vote their pocketbook” is a phrase at least as old as the time when pocketbooks existedand were referred to as such. For example, there is a popular belief that rich people vote in favour of lower taxes while poor people vote in favour of more generous social programmes.

Professor Bryan Caplan reviews an extensive political science literature on this hypothesis that voting is characterized by self-interest and concludes: “[P]olitical scientists have subjected the SIVH [self-interested voter hypothesis] to extensive and diverse empirical tests. Their results are impressively uniform: The SIVH fails.”[58] He provides numerous examples from the literature. For example, research has found there is only a slight connection between a person’s income and their ideology or political party affiliation.[59] Elderly Americans are not more likely to be supporters of Medicare than the young.[60] Males vulnerable to the draft support it at the same rates as other segments of the population.[61]

None of this should come as a surprise to even a moderately well-informed observer of American politics. Poorer parts of the country, particularly the rust belt, South-Eastern states, and parts of the Mid-West vote Republican.[62] One of the most famous books about this phenomenon of voting against your economic interests is the plaintively titled, What’s the Matter with Kansas?[63]

The answer, of course, is that nothing is uniquely wrong with Kansas. Voters don’t actually automatically vote in their economic self-interest, and we already know why. The chance that their vote will actually have an impact on their economic interests is miniscule. However, they enjoy all the psychic benefits they will receive from voting in a way that flatters their self-image, reflects strongly held prejudices, advances their social standing, and causes the minimum intellectual discomfort. Their votes are entirely rational because the marginal impact of their vote is insignificant. There is literally nothing to be gained from sacrificing these private benefits. To repeat Caplan’s apt phrase, voting does not exact a user fee for irrationality.[64]

The way voters process information is irrational even when their aims are entirely altruistic. For example, voters concerned about improving the employment prospects of blue-collar workers support tariffs despite the fact that tariffs create few jobs relative to their enormous costs.[65] In fact, they can produce net job losses.[66] Similarly, voters concerned about the environment often support high profile campaigns against horizontal energy transmission projects that interfere with the adoption of renewable energy and force energy companies to adopt dirtier methods of moving their products.[67] The adherents of the QAnon conspiracy came to their 2020 voting decisions in a way that can scarcely be called rational, but they were partially motivated by sincere concern for other people (especially children).[68]

It may be objected at this point that we are failing to give adequate weight to the rational self-interest of shareholders. This is a particularly salient objection in an article about the corporate franchise, because political voters become citizens mostly as a result of an accident of birth, and even if they immigrate consciously, they usually do so for considerations unrelated to their enthusiasm for the franchise.[69] Shareholders, in contrast, choose to become shareholders in order to make money and thus their relationship to a corporation is suffused with self-interest in a way not true for the average citizen of a democracy.

To determine whether the literature around political voting is relevant to corporate law, it will be necessary to examine the empirical evidence around shareholder voting.

IV. The Empirical Literature around Shareholder Voting

A. The Value Investors Place on Their Vote

There is plenty of evidence readily available that shareholders do not put much value on the voting rights attached to their shares. Even after all of the institutional, market, and normative changes around shareholder voting that have occurred over the past three decades, shareholders are remarkably passive in their voting behaviour. The overwhelming fact of shareholder voting is that it mostly leaves managerial and board arrangements intact. In a typical year, for example, only eight out of 31,000 American directors failed to receive a majority of votes cast by shareholders.[70] Reviewing the data, one academic suggests, “while shareholders may be willing to withhold votes when such an action is merely symbolic, such willingness may wane when the action actually has an impact on a directors’ position.”[71]

The well-known history of the rise of shareholder voting power since the 1970s supports this picture of disengagement. At every point, the expansion of the franchise was primarily driven by stock exchanges,[72] regulators of investment funds,[73] securities commissions,[74] proxy advisors[75] and academics.[76] Shareholders, themselves, have mostly been bystanders. It took strong action on the part of U.S. regulators to get institutional investors to take their voting power seriously in the first place, and, even then, most of them immediately delegated a great deal of the work around voting to proxy advisory firms.[77]

These common-sense observations are supported by the research on how voting rights are valued in the market. Looking at companies with dual-class shares, Luigi Zingales finds that the premiums for high-voting stock in America are low and often indistinguishable from zero, except in cases where control of the company is up for grabs.[78] He notes that the “value of a vote is determined by the expected additional payments vote holders will receive if there is a control contest ... [T]he size of this differential payment is a function of the private benefits obtainable from controlling a company.”[79] So, control over a company in circumstances where you can extract rents is valuable, but anything less than that is valued by the market as effectively worthless.

Another way of looking at the question of the value given to voting rights by public company investors is to examine the stock lending market. This market generally serves short sellers, but it also allows an investor to borrow stock for the purpose of utilizing the voting rights attached to it. A team of researchers found that the volume of lending activity in a company’s shares increases to a level about twenty-five per cent above normal on the record date for annual shareholder meetings, quickly returning to its usual levels afterwards.[80] There is no change in the costs associated with borrowing votes in this way. In fact, the costs of doing so are almost trivially low and do not increase on the record date when voting rights can be exercised.[81]

Recently, two different articles have attempted to determine the market value of voting rights by using bonds and options to replicate the cash flows associated with owning a share in a company.[82] This “contingent claims” approach uses this method to separate out a share’s economic value from its voting value. The two articles find voting rights form very little of the value of a share. Their estimates range from 1.23 to 1.64 per cent of a share’s value.[83]

These various lines of research explore how shareholders value their voting rights. What the research suggests is that shareholders value their vote about as much as citizens in a democracy: not much. This opens the possibility (but only the possibility at this point) that shareholders behave like ordinary political voters, remaining rationally ignorant and exclusively concerned with receiving certain private psychic and social benefits from their voting behaviour. To evaluate whether we see the same kind of irrationality visible in popular elections, we will have to look at how shareholders actually exercise their franchise.

B. Ordinary Uncontested Director Elections

The “just vote no” campaigns that generate abnormal numbers of “withhold” votes in an uncontested election only weakly reflect a corporation’s economic performance.[84] Poor economic performance predicts fewer votes in favour of a director, but a standard deviation in EBITDA-to-Assets ratio relative to industry peers results in an insignificant 0.37 per cent decrease in support.[85] Several studies find a similar result.[86] The evidence is even mixed about whether disappointing stock market returns produce withhold campaigns.[87] As one study of the literature notes, “[c]ompany performance has only a limited impact on the outcome of a director election, with results ranging from a statistically but not economically significant relationship to no relationship at all.”[88] There is thus little support for the proposition (often assumed in discussions about shareholder voting) that voting decisions are driven by bottom-line corporate economic performance.

What drives voting behaviour in uncontested elections? To state it simply: corporate governance. This is not “corporate governance” in the older and everyday sense of effectively leading the company to commercial success: it is the modern conception of “corporate governance” as adherence to a list of “best practices.”[89] The role of these corporate governance best practices in shareholder voting will be discussed in detail later in the article, as these practices arise repeatedly in research around voting. For now, it is only necessary to introduce the idea that the empirical literature examining these corporate governance best practices overwhelmingly finds that they have either no, or a negative, impact on corporate performance.[90]

There is a strong association between an Institutional Shareholder Services (ISS) withhold recommendation and the percentage of shares voted in favour of withhold.[91] Some of this is undoubtedly causation, some may merely be correlation.[92] For our purposes it suggests that, at least, the rationale given by ISS for a withhold recommendation likely reflects the voting intentions of other institutional shareholders. Given the rarity of withhold votes, it seems unlikely that in a particular year, a body of shareholders engages in the exceptional process of dissenting from a management proxy for completely unrelated reasons. This allows us to explore the motivation behind institutional shareholder voting decisions.

The kind of corporate governance best practices that seem to predict voting behaviour are familiar to anyone associated with corporate boardrooms over the past two decades.[93] One study looking at the votes received by S&P 500 companies over the period 2003–2010 found more than two-thirds of the withhold votes targeted against an individual director arose from concerns about their independence.[94] The remaining third reflected concerns over the director’s “busyness” and meeting attendance record.[95] Where an entire board committee was targeted, it was usually a function of concerns with executive pay.[96] When the board as a whole received an abnormal number of withhold votes, it was due to a lack of responsiveness to shareholder proposals receiving a majority vote (such as declassifying the board) or the board’s decision to adopt a poison pill.[97]

It is telling what does not appear to drive voting decisions in relation to directors: competence, experience, contributions to the board, and the underlying economic performance of the business. As one team of researchers observes, shareholders vote for directors as if financial performance, director performance, and firm governance matter, but the impact of these factors on actual votes is trivial.[98]

Research suggests that boards are responsive to the underlying concerns of an abnormal withhold vote. For example, the chance that a board will declassify itself increases from 4.9 to 36.9 per cent subsequent to a withhold recommendation from proxy advisors where this was a stated rationale.[99] In harmony with the vast literature about the irrelevance of these best practices for corporate performance, researchers looking at the S&P 500 companies between 2003–2010 concluded:

[W]e compare the subsequent performance of responsive and unresponsive firms, but find no evidence of differences, even in the most severe cases. One explanation is that the items on which proxy advisors and voting shareholders focus have little effect on firm value, consistent with the claim that activists misdirect their efforts towards ‘symbolic’ governance issues.[100]

In other words, shareholder voting in uncontested director elections is irrational.

This irrationality presents in other ways. For example, audit committee members are generally held responsible for accounting restatements, but not for weaknesses in the firm’s internal controls.[101] In contrast, representatives of management on the board are held responsible for the latter, but not the former. This demarcation of responsibility is mysterious, as audit committee charters usually contemplate responsibility for both matters.[102] As I have argued elsewhere, it is unlikely that the outsiders that now constitute audit committees could discover problems of either type given their dependence on management and auditors for the information needed to do their jobs.[103] That is what empirical evidence about outside directors on audit committees suggests in any event.[104] It seems arbitrary to hold audit committee members accountable for one failure, but not the other. However, the arbitrary rule that management is responsible for control failures while the audit committee is responsible for restatements is useful in giving shareholders the impression that responsibility and punishment are discriminatorily allocated and administered. The truth about actual competence and failure is irrelevant.

Directors of companies caught up in the last decade’s option backdating scandal were significantly more likely to be subjected to withhold votes, even though the actual backdating activities had occurred ten years before, and the directors had joined the board after the backdating had occurred.[105] It is true that directors who had been on the board at the time of the backdating had higher withhold votes cast against them, but the difference was a relatively insignificant additional 3.8 per cent of withheld votes.[106] It is hard to see how penalizing individuals for something done years before they joined the board is rational. It does, however, allow a shareholder to express anger, demonstrate virtue, align themselves with a community, and, in short, behave exactly in the way researchers have come to expect of the average political voter. Indeed, the absence of real shareholder concern with the substance of the backdating scandalthe failure of directors to act with integrityis clear given that directors who had overseen backdating at one firm did not receive statistically significantly more withhold votes in relation to the board positions they held at other firms.[107]

C. Majority Voting

According to the Ontario Securities Commission, majority voting policies were introduced because they would “improve corporate governance standards in Canada by providing a meaningful way for security holders to hold individual directors accountable.”[108] The way majority voting does this is by simultaneously making the shareholder vote more powerful (as directors can more easily be voted out of office) and less expensive (as a rival slate of directors and accompanying arguments in their favour is no longer required). It may be possible that this combination of greater power and lower costs changes the incentives which lead to the uninformed and irrational shareholder voting that exists under plurality voting regimes.

As usual, most of the research on this topic uses American data. Majority voting has been introduced into the United States by way of shareholder pressure. In 2005, less than ten per cent of the S&P 100 had majority voting policies; by 2014 almost ninety per cent of the S&P 500 had some sort of majority voting.[109] At the same time less than twenty per cent of small-cap companies had adopted majority voting.[110] In keeping with the picture of shareholder disengagement on voting discussed thus far, labour union pension funds (which hold less than 0.01 per cent of America’s companies’ shares) sponsored over eighty per cent of the majority voting proposals.[111] Given what we know about the incentives of managers of labour union pension funds, it seems very possible that the primary motive of the fund managers in this area was not to improve firms’ economic outcomes.[112] Indeed, the early adopters of majority voting policies cannot be distinguished from their peers in terms of economic performance.[113]

Even more telling is that the companies initially targeted by majority voting proposals were the “most shareholder responsive” measured by their previous withhold vote totals.[114] Some scholars have suggested this might have been the result of a careful strategy by shareholder activists to start with the most responsive companies as a way of putting pressure on the more recalcitrant ones.[115] But it is also possible that activists are motivated by the private benefits that victories of this sort afford, so they choose the weakest opponents.

The facts which support this latter explanation are: (1) there is considerable evidence that labour fund managers primarily engage in activist campaigns to please the union officials who hire them;[116] (2) there has been a notable lack of pressure on smaller, less visible companies to adopt majority voting, which makes sense if the goal is high-profile victories, but does not make sense if the goal is a more robust shareholder franchise;[117] (3) there is little evidence that shareholders have applied pressure on companies going public to adopt majority voting (IPOs being a time when investors are considered to possess considerable power over the corporate contract);[118] (4) as discussed, there is no evidence that either the early or late companies targeted for adoption had corporate governance issues; and (5) majority voting is actually more likely to be adopted by companies if they are incorporated in jurisdictions that do not require a general vote of the shareholders to do so.[119]

The most compelling evidence that activism around majority voting is more about self-expression than improving corporate performance is that directors of companies with majority voting are significantly less likely to be voted against. The difference is huge: the likelihood that a director of a company with plurality voting fails to receive a majority “for” vote is nineteen times higher than if he or she is subject to a majority voting policy.[120] So, at the very moment when voting against a director ceases to be symbolic and becomes effective, the shareholders stop voting against directors. The gulf between voting behaviour between the two regimes persists even when one looks at the “non-shareholder responsive” late adopters of majority voting.[121] Even the percentage of shares cast in director elections declines slightly after adopting majority voting, a result that seems incompatible with assumptions that shareholders are motivated to substantively impact corporate performance.[122]

For their part, the people with the deepest knowledge about individual directors and the value that those directors contribute to board activities, treat shareholder withhold votes with extreme skepticism. Because majority voting policies usually permit the board to refuse to accept a director’s resignation when she receives a majority of withhold votes, we can evaluate the board’s view of the quality of these votes from its behaviour. Where they have the power, boards tend to reject the director’s resignation and, in many cases, the director is approved by the shareholders the next year.[123]

The failure of boards to respond to the shareholders’ vote is usually seen as evidence of unaccountable self-interest and a scandal.[124] It is just as likely that it reflects superior knowledge about the targeted director’s contributions to board processes. As we have seen, directors are targeted largely due to decisions of the board as a whole, such as those relating to executive compensation, or the adoption of corporate governance best practices unlikely to improve corporate performance. Almost the only area where directors are targeted for their own behaviour is failing to attend a certain percentage of board meetings.[125] In an era of significant informal and year-round communication amongst directors and managers, formal meeting attendance is a very crude measurement of engagement and value creation. Thus, the most common reasons for a majority withhold vote are likely arbitrary and the votes themselves cast in ignorance of the actual role played by the director in question. Similarly to plurality voting, the empirical evidence about voting patterns under a system of majority voting suggests that “rather than a channel to remove specific directors, director elections [under majority voting] are viewed by shareholders as a means to obtain specific governance changes.”[126]

It seems unlikely that the same boards which voluntarily adopted majority voting policies in the first place,[127] de-staggered themselves,[128] created super-majorities of independent directors,[129] and began to compensate their executives in the ways promoted by shareholder advocates,[130] suddenly decided to engage in dismissive self-dealing. It seems at least as likely that boards retaining directors who failed to obtain majority support are just trying to do their best to advance the interests of the company, notwithstanding a shareholder vote they regard as a mistake.

The economic effects following the adoption of majority voting are not well studied. There are three studies that look at the stock price reaction to the announcement that firms were adopting majority voting policies. One found a positive abnormal price return,[131] and two found no statistically significant price movement.[132] Event studies are not particularly helpful when evaluating corporate governance events.[133] This is because news about most governance events is often anticipated by the market before it is officially announced. Moreover, it is rare that the adoption of some governance structure is announced at a time when no other announcements or conflating events occur (this is particularly the case when the announcement is about the results of a shareholder meeting). Similarly, it is rare that a major governance change is not part of some larger corporate transition, and many announcements (such as 13D announcements) are invariably associated with some kind of market reaction as speculators acquire shares in the hope something further will occur.[134]

The most useful empirical evidence illustrates the longer-term effects on corporate performance. There is only one study that looks past the initial announcement and it finds that majority voting is associated with worse firm performance relative to matched firms over the year following its adoption.[135] This underperformance can be seen in several areas including return on assets and market adjusted stock returns.[136] It is hard to know what to make of these results, as one year seems too short to see the effects of a change in corporate governance structures. So, perhaps the only conclusion that can be drawn is that the empirical research on the subject (published in 2013) has not noticeably impacted shareholder enthusiasm for majority voting, and this might be the most telling fact of all.

Indeed, there is some evidence that shareholders themselves understand that majority voting provides little economic benefit. Approximately ninety per cent of companies going public in America have plurality voting for uncontested director elections.[137] An IPO is a time when companies are under pressure to maximize their value, both because it marks the moment powerful private investors exit, and because the shares sold often make it an unusually dilutive transaction for those shareholders who remain. Companies going public are not part of any index, do not have an established track record, and are frequently not yet profitable.[138] They must work unusually hard to attract investors. There are a lot of reasons why we might expect companies and investors to push for majority voting if it was accretive. Instead, we find the opposite. A study discussing majority voting structures observes, “the portfolio managers who buy shares in the IPO are less concerned with the hot-button governance issues ... than are their colleagues who later have responsibility for voting those shares.”[139] This gulf between the asset managers who make economic decisions and those that make voting decisions is suggestive of divergent objectives.[140]

D. Voting in Contested Director Elections

Contested director elections are usually fought on the basis of the quality of corporate governance at the targeted firm. As Carl Icahn described the reason for his activism:

Too many companies in this country are terribly run and there’s no system in place to hold the chief executives and boards of these inadequately managed companies accountable ... Our current system of corporate governance protects mediocre chief executives and boards that are mismanaging companies and this must be changed.[141]

The majority of communications in a proxy campaign explicitly reference the quality of corporate governance, even when there is also a clear disagreement on economic strategy at the heart of the campaign.

In a typical proxy campaign, most of the claims made by the insurgents concern allegations of poor corporate performance, bad governance practices, conflicts of interest, insider trading, overly-generous executive compensation, and problems in the quality and experience of individual directors.[142] The arguments from the incumbents are similarly focused on repudiating the allegations, describing governance failures on the other side (such as “golden leash” payments), and criticizing the quality, track record, and independence of the candidates making up the dissident slate.[143] It is this focus on corporate governance, the most visible aspect of proxy contests, that caused The Economist to famously refer to activist shareholders as “Capitalism’s Unlikely Heroes.”[144] Similarly, academics tend to have a favourable view toward activism because of its role in generating “superior corporate governance.”[145] In a much-cited recent journal article, professors Ronald Gilson and Jeffrey Gordon argue, “[a]s governance intermediaries or governance arbitrageurs, activist shareholders can, in the right circumstances, serve to reduce the market’s undervaluation of governance rights to the advantage of all shareholders.”[146]

The surprising thing, then, about contested director elections is that when researchers study them, the elections do not appear to have anything at all to do with corporate governance.[147] First, the companies that experience contested elections actually appear to be generally well-run. There are very few contested elections in the United States. Out of approximately 4,000 public companies, until 2013 there was only an average of thirty-five firms per year where the directors faced competition.[148] This total has increased in the years since to 187 companies in 2019.[149] Targeted companies tend have “low market value relative to book value ... with sound operating cash flows and return on assets.”[150] Indeed, most studies have found that the targets of proxy fights are more profitable than control samples.[151] Stated simply, the common target of a contested election is an unusually profitable company, but with recent stock price returns that are lower than its peers.[152]

Second, if we take activist shareholders as the most significant source of contested elections, we know what drives their economic returns, and it is not changes to corporate governance. The real goal of activist shareholders is usually one of a limited range of measures designed to increase the short-term financial returns to the shareholders.[153] These include restructuring the company (spinning off a non-core asset or blocking an acquisition), changing a payout policy (increasing or implementing a share buyback programme or increasing dividend payments), or selling the company.[154] Board or management changes are generally a prelude to enacting one of these strategies.[155] Virtually all the returns experienced from shareholder activism are attributable to those companies which are sold following their interventions.[156] This explains why unusually profitable companies are the most frequent targets: all of these financial maneuvers depend on fundamentally sound businesses with strong cash flows.[157]

Third, there is no evidence of measurable improvement in a firm’s corporate governance following a successful proxy campaign. Operational metrics (such as growth in sales, asset size, profit margin, the spread in borrowing costs, return on assets, return on equity, and profitability) are all unaffected.[158] Indeed, for companies that experience a board change and that are not sold, the best reading of the available evidence is that they lag behind their peers over the long term.[159] Even when an activist only succeeds in placing a few of its directors on the incumbent board, the companies tend to experience significant underperformance in the following years.[160] All of this is problematic for proponents of the theory that shareholder activists improve corporate governance.

It is possible that corporate governance is, in a very weak sense, improved if we take its most narrow definition as referring solely to board independence. It is hard to imagine a more independent board than one imposed on the company by the victors of a contested election. Is there evidence that companies that experience a change of directors do a better job of constraining executive compensation and other illegitimate diversions of the firm’s free cash flow? Unfortunately, this does not appear to be the case. As two scholars note, after surveying the empirical literature about what successful challengers actually do following an electoral victory, “the majority [of studies] do not report evidence of changes in real variables consistent with this free cash flow hypothesis.”[161]

If the companies were not badly run before they were targeted for a contested election, and if it is clear that activist shareholders actually generate their returns in ways that have nothing to do with improving corporate governance, and if there is no improvement (or even a decline) in firm performance following the replacement of incumbent directors, then what is going on? Contested corporate elections appear to be fought on the basis of one thingcorporate governancethat doesn’t actually seem to be the point. There are several possible explanations for what we see in contested elections.

One possible explanation is that the war of words around corporate governance has the effect of misleading shareholders in ways calculated to influence their voting. It is not that shareholders in a contested election vote against their economic interests, it is that they can be misinformed about where those interests lie. This is probably the most common explanation for critics of shareholder activism. It is probably not true, however. When we look at the actual economic decisions of shareholders, we find that they show a perfect understanding of the reality of contested board elections. In general, the announcement of a management victory in a contested election does not result in negative abnormal returns as we would expect if the market really believed a negligent or compromised board had succeeded in retaining its authority over the company.[162] Investment bank analysts (who have only an economic interest in predicting the actual results from a contested shareholder election) do not expect post-activism improvements in corporate earnings as shown by their earnings per share forecasts.[163] As we have seen, they are not wrong.

Even more impressively, in a recent article, several scholars show that the market accurately prices the impact of the increase in information leakage that follows a settlement agreement placing hedge fund employees on a corporate board.[164] The accuracy of the market’s assessment extends to distinguishing between employees of the hedge fund and independent directors proposed by the hedge fund (information leakage only increases when the former goes on the board), and whether the settlement agreement contains confidentiality provisions (information leakage only increases in the absence of these provisions).[165] When it comes to buying and selling shares, it is clear that the market understands exactly what the effects of a change of directors is going to be and prices it surprisingly accurately.

This brings us to the second possible explanation for the prominence given to corporate governance claims in contested elections: the mistake is not on the part of the shareholders, but on the part of activist shareholders and incumbent boards. These latter groups fight over corporate governance in the mistaken view that it matters, or in the cynical (but also mistaken) view that they can fool the shareholders.

This also seems an unlikely explanation. For one thing, it depends on a fairly fundamental mistake being made by quite sophisticated and well informed parties, both with a deep knowledge of the market. If activists and boards hold the mistaken view that the corporate governance issues matter, then we have to believe that the two parties don’t understand the actual economic drivers of the returns expected by the activist shareholder or, in the case of the board, the actual aspects of the activists’ proposed strategy that will impact the long-term prospects of the company.[166] If the mistake is that activists and incumbent boards incorrectly believe shareholders can be fooled, then we are again left with the question of how long this mistake could realistically last? There is near constant communication with shareholders in a contested election. Is it really plausible that a delusion of this sort could last through even a single contested election, much less the collective market experience of hundreds of such elections?

The argument being made in this article provides a third possible explanation: the corporate governance aspects of a contested election actually make a difference to how the shareholders vote, but this voting has nothing to do with the shareholders’ clear-eyed assessment of the economic consequences of the corporate governance dispute. The merit of this explanation is that it does not require any of the three sophisticated parties to labour under a persistent misapprehension. The activists and board are right to emphasize corporate governance matters because these, in fact, drive shareholder voting behaviour. The shareholders are aware, in their buying and selling activities, that the corporate governance stuff won’t make a difference, but, in their voting, they are taking a virtually costless opportunity to express their values, show solidarity with their tribe, and reaffirm strongly held beliefs about the necessity of shareholder oversight that reflect well on themselves.[167] The outcome may be irrational, but each party in the contest is behaving perfectly rationally.[168]

E. Voting on Corporate Governance Matters

As we have seen, corporate governance best practices drive most shareholder voting in director elections, to the near exclusion of factors such as the firm’s economic performance.[169] When we look at voting in other areas of the shareholder franchise, we find a similar focus on governance practices. In a recent representative year, for example, there were 315 shareholder proposals, of which 182 related to the adoption of corporate governance structures (including forty-seven on executive compensation practices).[170] Of the thirty-three proposals that won majority support, all were related to corporate governance (two of them related to executive compensation).[171]

Corporate governance proposals are primarily made by a few public pension and labour funds.[172] For-profit investment managers are unlikely to initiate these types of proposals (in a particular year, less than one per cent originate from this source) but they often vote in favour of corporate governance proposals.[173] Indeed, one study finds that the average governance proposal attracts the support of sixty-five per cent of mutual funds and sixty-nine per cent of pension funds.[174] The support for other sorts of non-governance proposals is much lower.[175]

Viewed from the perspective of economic outcomes, corporate governance best practices form an irrational basis for shareholder voting. The most optimistic reading of the empirical literature is that corporate governance best practices make no difference to firm performance.[176] There is plenty of evidence that some of those best practices, for some companies, lead to worse outcomes.[177] Among scholars familiar with the literature, this is neither controversial, nor breaking news. As two finance professors noted just over a decade ago after conducting a review of the literature around board independence, the cornerstone of modern corporate governance, “we are not aware of a body of literature in corporate governance—or elsewhere—where null results present with such consistency.”[178] This was written more than a decade after the first meta-analysis on independence found no relationship between board composition and corporate performance.[179] It is not just independence that lacks support in the empirical literature (though independence is the foundation for most governance best practices) but rather the entire gamut of governance mechanisms that have been proposed and adopted over the past two decades.[180] The problematic outcomes of modern corporate governance practices extend to the one-size-fits-all executive compensation practices promoted as part of the corporate governance activities of institutional shareholders and proxy advisors.[181] As a result of this vast and, in most cases, quite established empirical literature, the mere fact that these practices appear to form the primary motivation for shareholder voting decisions is evidence of irrationality in the corporate franchise.

This irrationality persists even when empirical research has discredited the activities of specific shareholders. For example, CalPERS continues its activism around corporate governance even though a study published back in 1996 found that its interventions had only a very minor impact on share price and no impact at all on operating performance.[182] TIAA-CREF similarly continues its corporate governance activities notwithstanding a 1998 study that found these initiatives produced no significant changes in accounting measures of performance and even, in some cases, caused a decline in share prices.[183] More recent studies about specific funds active in corporate governance matters have found the same lack of effect, but this has not led to noticeable changes in those funds’ behaviour.[184]

In contrast, the average institutional shareholder clearly understands that the corporate governance issues driving its voting decisions don’t make much difference to the economic performance of its portfolios. As we have seen, most institutions—in particular those that compete for money on the basis of fund performance—don’t actually engage in corporate governance activism, aside from casting votes.[185] When scholars look at the companies acquired by institutional shareholders, they “find little evidence of an association between total institutional investor ownership and corporate governance.”[186] In fact, only about ten per cent of institutional investors appear to invest in ways that are at all sensitive to firms’ corporate governance arrangements.[187] We saw a similar pattern in behaviour around IPOs, where the decision to add a new company to an institution’s portfolio appears to have little to do with the company’s adherence to governance best practices.[188]

The clearest sign that shareholders generally understand the economic consequences of their voting behaviour can be seen in the kinds of governance initiatives they support. Professors Kahan and Rock describe their behaviour as “symbolic corporate governance politics.”[189] They observe repeated instances where shareholders choose to invest their energy in purely inconsequential reforms of corporate governance, staying away from changes that would actually impact operations. For example, shareholders invest considerable energy in pressuring companies to remove poison pills, but their efforts do nothing to prevent a board from introducing a poison pill unilaterally and instantly in the event of a hostile takeover bid.[190] The shareholders could constrain the board’s ability to adopt a pill through a charter amendment, but shareholder proposals do not ask for this. Instead, shareholders ask only for the purely symbolic removal of existing pills.

There are other circumstances which display similar patterns of shareholders denouncing certain governance structures but refraining from making proposals that would actually force a change on the company. These circumstances can be observed in relation to: proxy access (where shareholders choose not to force proxy access through bylaw amendments after Delaware made this possible), majority voting (where, as we have seen, directors are not much more likely to leave boards than under plurality voting), proposals to remove supermajority requirements in bylaws (which commonly apply only to matters that are either benign or practically irrelevant), and the scarcity of mandatory (as opposed to precatory) shareholder proposals.[191] Shareholders and their proxy advisors cannot be ignorant of the inconsequential aspects of their engagement with corporate governance, or of the stronger alternatives available to them. Yet, they often remain content with purely expressive activity in this field.

V. Revisiting the Academic Debate

Expressive voting provides a reason to rethink the grounds on which academic arguments about shareholder voting are conducted. To take an easy example, the extensive debate around whether the economic interests of shareholders correspond with the long-term interests of the corporation becomes, in the context of voting behaviour at least, irrelevant. As we have seen, over and over again, the reasons for shareholder voting patterns have nothing at all to do with those shareholders’ economic interests.[192] Indeed, there are frequent cases when shareholders’ trading decisions, which reflect their economic analysis, are entirely at odds with their voting behaviour.[193] Markets extract a user fee for irrationality; voting does not.[194]

The collective action problem, with its focus on the costs of obtaining the information to properly inform voting decisions, also seems largely beside the point. There isn’t a problem with gathering information; it is the way that information is gathered, processed, and used for voting decisions that is irrational. Anti-vaxxers have access to all the information and experts available to the rest of the population, they just choose to get most of their information from other sources and to engage in highly motivated reasoning about the rest. This is why, for example, proxy advisors—whose business is to flatter the prejudices and self-regard of their clients, the institutional shareholders—continue to recommend governance best practices that have been extensively discredited in the empirical literature.[195] And it is why institutional investors continue to turn to proxy advisors for voting recommendations. It is why a peer-reviewed study can be published about the adverse consequences of an institutional shareholders’ activism and its findings ignored by that very institution.[196]

Finally, if we look at the debate around the economic incentives that apply to professional fund managers, we find that this also looks less relevant than it did before. The reality is that the voting decisions of fund managers are a function of neither an economic evaluation of the costs of informed voting, nor the benefits received by that manager (through fund performance relative to benchmarks). It doesn’t matter if the cost of informed voting was reduced to zero. Fund managers appear to vote their shares to express themselves, and fund performance, relative or otherwise, appears to have little to do with it. Indeed, all that has been accomplished by majority voting, the rise of activist shareholders, and the coordinating function of proxy advisors, has been a growing phenomenon of voting on purely expressive or symbolic matters.[197]

It should be noted that by focusing on the marginal value of the vote itself, we can avoid lengthy evaluations of the varying incentives and institutional features of different types of shareholders.[198] There is no “right” kind of shareholder, because all of them face exactly the same incentives to reap the psychic benefits of expressive voting. Hedge funds, index funds, pension funds, mutual funds, and retail investors all face different kinds of incentives and pressures, but the low-to-non-existent marginal value of their vote remains the same. There is no evidence in the literature, for example, that the largest institutional shareholders, or the institutional investors with the strongest interest in firm-specific governance (such as hedge funds), come to their voting decisions in materially different ways. They consume the same advice, support the same proposals and, as we have seen, they vote in broadly similar patterns.[199]

In 2013, Jana Partners launched a proxy battle against Agrium, a well-known Canadian company.[200] At stake was a fundamental difference of opinion about corporate strategy. Jana advocated for the sale of certain business divisions (and the distribution of the resulting proceeds to the shareholders), while the board was adamant that the business lines in question were essential to the long-term success of the entire company.[201] As usual, the quality of corporate governance at Agrium became an issue, with Jana Partners harshly criticizing the board, as well as specific directors.[202] Agrium’s board returned the favour with criticisms of Jana’s board nominees and the “golden leash” contracts these nominees accepted from Jana Partners.[203] There was even a fight over the correct valuation metric to use when evaluating corporate performance.[204] The lengthy war of words quickly became ugly. (The press characterized it as “dirty” and “vicious,” noting a Jana executive’s “jaw dropping rant” at the shareholder meeting.)[205]

At the shareholder meeting, a sizeable minority of Agrium’s shareholders voted in favour of Jana Partner’s proposal to create a “hybrid” board consisting of some of the existing Agrium directors plus several of Jana’s proposed directors.[206] It is hard to understand this vote as rational. It is not a recipe for success to saddle a body that normally works on consensus, trust, and a supportive attitude toward management (on whose candour and integrity boards depend), with two groups which are at odds with each other. This is especially the case when the groups not only possess fundamental differences of opinion on strategy and how to value the business, but also a history of bad blood and publicly aired criticisms of one another. While the defects of a hybrid board are obvious to anyone who has spent much time in a boardroom, you don’t have to have served as a director to understand how little chance such an arrangement has of succeeding. Has any kind of team been effective when divided by intractable disagreements and personal animosity?

As it happens, shareholders didn’t need any personal experience to cast their votes wisely. By the time of the conflict, several studies had been performed looking at the performance of hybrid boards, finding their performance lagged their peers over long periods of time.[207] The underperformance effects were not small, ranging from nineteen to forty per cent over the two years following the hybrid board’s installation.[208] These were not obscure studies, either. They were referenced only two years earlier in one of the most famous corporate law cases of the era, which used the Buckberg study in rejecting the Securities and Exchange Commission’s proxy access rules.[209]

Nevertheless, a range of shareholders decided to vote in favour of a hybrid board for Agrium. There was no lack of information. The dispute was covered widely in Canada’s business press. Both sides to the dispute were sophisticated and wielded significant resources to communicate their views. What motivated the voting in favour of a hybrid board? One possibility is that some shareholders simply followed ISS’ recommendation to vote for the hybrid board, even though other proxy firms, including Glass-Lewis and Egan Jones, made contradictory recommendations.[210] Giving no independent thought to a voting decision of considerable importance to the company in the face of contradictory voting advice is not rational.

What about those shareholders who actually considered the issue? Again, we see in ISS’ written voting recommendation an indication of the kind of expressive logic that could lead to an irrational voting decision. ISS’ recommendations make a number of assumptions including: (1) that managers are often mistaken about strategy, with a strong preference for self-interested empire-building; (2) that boards are often ineffective or compromised so they can’t correct management; (3) that shareholders are thus needed to identify the firm’s best interests; (4) that shareholders are often better able, by reason of skill or a lack of conflicts of interest, to see what ought to be done; (5) that shareholders are entitled to have their wishes effected in the firm; (6) that monitoring management is the primary role of the board; and (7) that independence is the most important quality in a director (and what could be more independent than a director representing a hostile shareholder?).[211] These beliefs all reflect a flattering image of shareholders as the indispensable element in good corporate governance: its centre, object, and tribune. The contrast with unreliable managers and directors could not be sharper. ISS’ analysts apparently believe this picture; isn’t it likely many of those shareholders voting for the hybrid board believed it as well?

Conclusion: The Scope of Corporate Democracy

Political voting behaviour in democracies is not, of course, always irrational. In circumstances where the cost of a miscast vote is very high, citizens behave rationally. The empirical literature about policy outcomes in democracies has some very good news. No mass famine has ever occurred in a modern democracy, no matter how poor the country.[212] Unlike dictatorships, democratic governments almost never engage in mass murder against their citizens.[213] Systems where politicians are accountable to the public tend to do better at limiting the disruptions and harms caused by natural disasters.[214] In these sorts of cases, voters have enormous incentives to vote rationally. The actions of politicians in these areas have huge significance, are easily attributable, and are highly visible.[215] In this context, the evidence is that voters invest the resources necessary to become informed and vote in their country’s best interests. Politicians in democracies know this and conduct themselves accordingly; this is why democracies do better.

It is hard to avoid the fact that traditional corporate law reserved the shareholder franchise precisely for the corporate equivalent of these major and highly-public events: mergers, wind-ups, and amendments to the charter documents. Even voting for directors served mainly as a method of ensuring accountability that only really functioned when there was a major failure. But over the past several decades, we have attempted to drive shareholder voting into areas such as corporate governance, executive compensation, business strategy, and even fine-grained assessments of the quality of individual directors. This goes well beyond the highly visible, hugely significant, and easily attributable matters that traditional corporate law reserved for shareholders.

When scholars of political voting look at the quality of votes cast in relation to matters that are complex, long-term, or require specialized knowledge, they have found that a little less democracy produces better results. Professor Garret Jones discusses this in his aptly titled book, 10% Less Democracy.[216] His examples include the extensive empirical evidence that the outcomes produced by central banks are improved when the central bankers are made independent of the democratic process.[217] For similar reasons, when the performance of U.S. city treasurers who are appointed by their city council is compared against those treasurers who are directly elected, scholars find that the appointed treasurers are able to get significantly lower interest on the city debt than their elected counterparts.[218] In particular, there are noticeable improvements when a city transitions from elected to appointed treasurers.[219]

There is also literature that compares democratically elected American judges with their appointed counterparts. It finds that compared to appointed judges, elected judges give significantly bigger awards in tort cases when the defendant is from out-of-state, a reasonable measure of bias.[220] Evaluated by citations by other courts, elected judges write lower quality opinions than appointed judges.[221] When a U.S. state changes its method of selecting judges from an election to some sort of appointment process, “compared to judges selected by voters, there is consistent evidence that judges selected by a merit commission are better at their jobs.”[222]

In light of the voting patterns described in this article, there is no reason to think that the conclusions of political scientists should not apply with equal force to corporate law. The attempt to give shareholders increasing amounts of authority over increasingly fine-grained corporate decisions, such as the merits of individual directors or compensation decisions, is bound to run aground on the implacable fact that the marginal value of a vote is indistinguishable from zero and so shareholders will vote on that basis. Mostly this means that they will process information and vote in ways that are expressive, not effective.