Print
E-MailMarcel Kahan is the George T. Lowy Professor of Law at the New York University School of Law. This post is based on a paper co-authored by Professor Kahan and Emiliano M. Catan at the New York University School of Law.
Over the last 15 years, numerous economics articles, many published in top finance journals, have examined the effect of takeover law on performance, leverage, managerial stock ownership, worker wages, patenting, acquisitions, and other firm actions. These studies have concluded, among other things, that anti-takeover laws are associated with a decline in managerial stock ownership, and increase in wages, and a decline in dividend payout ratios.
From a legal perspective, however, the varying methods that financial economists use to measure the takeover protection afforded by state law make little sense. Economists generally look either at whether (and when) a state adopted a business combination statute; at when a state adopted the first of a set of statutes (typically, business combination statutes, control share acquisition statutes, and fair price statutes); or at how many different types of statutes a state has adopted.
These methods all result in a gross mischaracterization of Delaware—a state that typically accounts for about half of the firm observations in the studies—as either having changed from a pro- to an anti-takeover state when it adopted its 1988 business combination statute or as being largely pro-takeover because it has only a single anti-takeover statute. This characterization ignores the centrality of case law on poison pills in Delaware and the fact that pills, which were validated in Delaware in 1985, mostly moot its business combination statute. The most important development in 1988 was thus not the adoption of the business combination statute, but the Chancery Court decisions in City Capital Assocs. v. Interco and Grand Metropolitan v. Pillsbury, which—temporarily—imposed severe constraints on the use of the poison pill.
More generally, most empirical studies do not start from a valid theory on how anti-takeover statutes affect the target’s marginal ability to defend itself. Thus, for example, the studies usually do not take account of the fact that targets in states where pills are valid have a high ability to defend themselves against takeovers even if the state has not adopted any anti-takeover statute or that, even ignoring the pill, business combination statutes render control share statutes largely irrelevant. For fair price and control share statutes, the studies ignore whether companies had adopted fair price charter provisions, which offer protection against coercive bids similar to that afforded by those statutes. Finally, many studies ignore the high degree of uncertainty over the validity of state anti-takeover statutes prior to the United States Supreme Court decision in CTS v. Dynamics and almost all fail to account for the decline in uncertainty over the validity of poison pills outside Delaware.
So why do empirical studies performed by economists yield results? To answer this question, we review three empirical studies published in leading finance journals. For each study, we show that the results are affected by omitted variables, large scale coding errors, or improper specifications. When corrected for these problems, the association between anti-takeover statutes and the hypothesized effect disappeared.
Our analysis has implications for several important debates. Most basically, our results are consistent with the view that anti-takeover statutes do not matter after all and suggest that the results of other studies reporting a statistically significant and economically meaningful relationship between firm behavior and anti-takeover statutes may also be due to other factors or caused by shortcomings similar to the ones in the studies we analyzed.
More importantly, since the adoption of anti-takeover statutes is the “natural experiment” most frequently used by scholars trying to estimate the effects of takeover threat, our analysis calls into doubt most of the empirical knowledge about the real economic effects of takeover threat. Starting in the 1980s, theorists have taken different positions on what these effects are. One set of scholars has argued that the threat of a takeover acts as a beneficial disciplining device that induces managers to act in the interest of shareholders. Another set has argued that the threat of a takeover induces an excessive short-term focus on management and thereby lowers long-term shareholder value. Yet others have suggested that the takeover threat leads management to take actions that benefit shareholders, but harm other constituents, and may therefore not enhance overall social value.
The evidence generated by the empirical studies we criticize, which use anti-takeover statutes in various configurations as a measure of the takeover threat, has tended to favor the first set. Thus, studies have found that the adoption of these statutes is associated with firms reducing their use of debt, letting the wages they pay rise above competitive levels, and experiencing a decline in productivity and operating performance. These results have been interpreted by their authors and others as showing that a reduction of the takeover threat leads to an increase in managerial slack. Such studies have also found that the adoption of these statutes is associated with a decline in the creation of new plants and patents—results seemingly at odds with the notion that the takeover threat induces short-termism.
But if the passage of anti-takeover statutes did not change the takeover threat facing firms—as we argue—these studies have no bearing on how takeover threat affects managerial slack or focus. The association between anti-takeover statutes and the effects studied may thus be spurious, due either to methodological flaws, omitted variables, or coincidence, but not present a causal relationship.
The full paper is available for download here.
