Einde inhoudsopgave
The Decoupling of Voting and Economic Ownership (IVOR nr. 88) 2012/3.3.2.3
3.3.2.3 Attribution Errors
mr. M.C. Schouten, datum 01-06-2012
- Datum
01-06-2012
- Auteur
mr. M.C. Schouten
- JCDI
JCDI:ADS598269:1
- Vakgebied(en)
Ondernemingsrecht / Rechtspersonenrecht
Voetnoten
Voetnoten
Justin J. Wolfers, Are Poters Rational? Evidence from Gubernatorial Elections, 1 (Stan. Graduate Sch. of Bus., Working Paper No. 1730, 2006), available at http://ssm.com/abstract=305740.
Id. at 17 (fmding also that voters in procyclical states 'are systematically fooled into reelecting incumbents during national booms, only to dump them during national recessions').
Id.
Raymond J. Fisman, Rakesh Khurana & Matthew Rhodes-Kropf, Governance and CEO Turnover: Do Something or Do the Right Thing? 4, 20 (Jan. 2005) (EFA 2005 Moscow Meetings Paper), available at http://ssm.com/abstract=656085 (examining 139 CEO firings between 1980 and 1996 and obtaining results suggesting that 'in circumstances where the CEO might be expected to be dismissed based on previous performance, but was retained, subsequent performance is stronger in firms with entrenched boards that may have been able to resist shareholder pressure to dismiss the CEO'). Additional evidence suggesting that attribution errors are made in the corporate context is presented by a study of the behavior of directors charged with rewarding CEO's, which concludes that the average Forbes 500 firm rewards its CEO 'as much for luck as it does for a general movement in performance.' Marianne Bertrand & Sendhil Mullainathan, Are CEOS Rewarded for Luck? The Ones Without Principals Are, 116 J. Econ. 901, 907-08, 914 (2001) (studying pay and performance for the 51 largest American oil companies between 1977 and 1994 and finding that 'pay changes and oil price changes correlate quite well.. [which] is suggestive of pay for luck,' and examining a broader dataset to determine whether the observations can be generalized and finding that they can).
The three examples of bounded rationality suggest it is unlikely that each shareholder votes for the correct answer with a probability exceeding 0.5. Some might argue that this need not undermine the usefulness of the Jury Theorem as a theoretical foundation for corporate voting since the Theorem also holds if the average competence level exceeds 0.5. See Thompson & Edelman, supra note 17, at 132 n.5. While this may be true as a statistical matter, applied to corporate voting it confronts us with the problem that we really have no reliable way of measuring investors' competence levels, and thus to determine whether the average competence level exceeds 0.5. To illustrate this point, imagine a company with three shareholders: two biased retail shareholders and one rational institutional shareholder. How do we assign probabilities of voting for the correct answer? If we deem retail investors to vote for the correct answer with probability 0.4 and the institutional investor to vote for the correct answer with probability 0.8, the average competence level exceeds 0.5. But if we deem the retail investors to vote for the correct answer with probability 0.3 and the institutional investor to vote for the correct answer with probability 0.7, the average competence level is below 0.5. In the absence of suitable criteria, the task of assigning probabilities remains somewhat arbitrary in nature. And although it might be objected that the example doesn't correspond to reality because the majority of the shares in public firms are not held by retail investors but by institutional investors, this only shifts our attention to a second problem, which is that institutional investors too may be irrational voters. See Shiller, supra note 35, at 33 (noting that '[p]rofessional investors . . . are not immune from the effects of popular investing culture that we observe in individual investors, and many [cultural and psychological factors] no doubt influence their thinking as well').
As a final example of how bounded rationality may affect voting behavior, consider the following evidence from a study of gubernatorial elections. The study explores the hypothesis that in deciding whether or not to vote for an incumbent governor who is up for re-election, rational voters will reward good economic outcomes that reflect the governor's actions, but filter from their assessment economic events that reflect influences outside the politician's locus of contro1.1 The findings suggest the opposite: voters in oil-producing states tend to re-elect incumbent governors during oil price rises and vote them out of office when the oil price drops.2 This suggests that voters make "systematic attribution errors" and are best characterized as "quasi-rational."3
Do shareholders make the same attribution errors, voting to re-elect, for example, directors of oil companies during oil price rises (when profits tend to rise) and vote them out of office when the oil price drops (and profits tend to decline)? More generally, do shareholders overly attribute a firm's results to management's performance and insufficiently to industry-specific or macroeconomie trends? It certainly isn't hard to imagine. In fact, one empirical study of CEO-firings presents tentative evidence that shareholders tend to "shoot the messenger" by dismissing managers for failures that cannot be attributed to them.4
To conclude, rational voting is an important mechanism of voting efficiency. In terras of the Jury Theorem, if individual shareholders do not vote rationally, the probability that they will vote for the correct option may drop below 0.5 in which case the Theorem predicts that a majority vote will fail to identify the correct option.5 Meanwhile, behavioral finance suggests that investors have bounded rationality and political science suggests that voters have bounded rationality. This section has merely offered a glimpse into how bounded rationality may affect corporate voting; future research will hopefully deepen our understanding.