Early corporate law scholarship argued both that anti-takeover devices are inefficient (they reduce the value of the firm) and that firms adopt efficient governance terms before they make their initial public offering. Some of this scholarship asserted that firms go public without anti-takeover devices and adopt them later when agency costs are higher. However, subsequent research revealed that most firms adopt anti-takeover devices before completing their initial public offerings. For example, over eighty-six percent of firms that have gone public in 2012 have a staggered board of directors, and both Google and Facebook chose dual-class capital structures that allow the founders to retain voting control disproportionate to their economic stake.
The literature offers a number of explanations for this apparent puzzle. Capital market imperfections may prevent initial public offering prices from reflecting differences in corporate governance terms. Firms may choose inefficient terms due to bad legal advice or because of frictions in the market for financing prior to the initial public offering. Anti-takeover protections could be efficient after all, at least for some firms, because they correct for myopic investors or some other problem. Finally, managers may choose anti-takeover provisions to signal something about their firms. In an essay forthcoming in the Journal of Corporation Law I offer a very different explanation, one based on the tax code.
My argument begins with a variant of one of the existing explanations for anti-takeover protections. The heart of the argument is that managers are not driven solely by a desire for material gain but derive some happiness or utility from the control they exercise over their firm. To the extent that managers derive happiness from control, they may not choose governance terms that maximize the dollar value of the firm. However, unless there is some contracting failure, they will still choose efficient terms — terms that maximize the total value of the firm (the dollar value plus the control value).
Once we introduce taxation, the manager no longer chooses efficient terms because the government taxes the pecuniary returns of firm ownership (corporate income, share appreciation, etc.) more effectively than it taxes non-pecuniary returns. Even if the manager can increase the total value of the firm by ceding some control, she may fail to do so as she must share some of the increase in the monetary value of the firm with the public treasury. The government could eliminate this distortion by taxing the potential value of a firm (its value with optimal governance terms) or by taxing the benefits of control. The government could also mitigate the problem by taxing devices that allow the manager to retain control such as dual-class common stock, staggered boards and poison pills or by simply prohibiting devices that it finds unreasonable.
Unfortunately, it may be difficult or impossible to implement the corrective taxes described above, and these taxes could have perverse effects. For example, if we assess taxes on anti-takeover devices used by a publicly traded firm (or if we prohibit these terms), we distort the manager’s decision to sell shares to the public. In a world of limited information, there are no easy choices. Whether the distortions that would be created by corrective measures exceed those created by the existing tax code is a difficult empirical question that is left to future research. However, even if we choose not to adopt corrective measures we should still recognize the distortion of corporate control as an additional cost of taxation.
The full paper is available for download here.