PERHAPS the most hackneyed and intractable debate in all of business law concerns the question of whether Delaware has incentives to provide an optimal corporate law, whatever that is. The world seems divided into the race-to-the-topers and the race-to-the-bottomers, with increasing amounts of scholarship piling up on both sides, none of which seems to be convincing the other side or moving policy forward in a meaningful way.1 When asked to respond to the latest salvo in this battle, I feared more of the same. But after reading Professor Michal Barzuza’s thought-provoking article, Delaware’s Compensation,2 I am convinced that there are still interesting things to be said about the optimality of the state-as-competitor-for-charters model of modern American corporate governance. I do not find Professor Barzuza’s proposal for making the franchise tax proportional to firm value convincing or necessarily desirable, but, because of the natural check provided by state competition, it is unlikely to do much harm.
Professor Barzuza articulates her thesis as follows:
If Delaware’s [franchise] tax were more sensitive to firm value, or if Delaware increased its tax to reflect changes in the quality of its law, the state would have better incentives to invest in quality, even in the absence of competition, because Delaware would be rewarded for such changes with higher tax collections.3
The idea is that the current system of franchise taxation—basically a flat fee of $165,000 for large firms—does not provide legislators with sufficient incentives to overcome the ability of managers to lobby for management-friendly legislation. Professor Barzuza claims that managers dominate the process of incorporation (both at the IPO and reincorporation stages), and that legislators rationally favor them over shareholders, in part because the benefits from favoring managers are real and sizable, while increasing shareholder value does little to attract or keep firms and does not increase the state’s $165,000 take.
Professor Barzuza’s paper makes an important point: taxes are not only a form of regulation, but also can be an incentive for efficient regulation.4 To see this, compare two worlds in which there is a single legislator who writes the rules for firm governance. In the first, much like modern-day Delaware, the legislator receives a flat fee from each firm to spend on public goods; and in the second, much like Professor Barzuza’s imagined Delaware, the legislator receives a large percentage of firm profits (say 50%) that the legislator can dole out to constituents. All else being equal, it is obvious that the legislator in the second world has a stronger incentive to increase firm value, as the benefits flow directly to the legislator.
It is difficult to object to this claim at a theoretical level, since it may improve incentives for legislators on the margin, and if the legislators set the rate too high or enact changes that actually destroy shareholder value, companies will simply move to Maryland or lobby Delaware to change back. My guess is, however, that the impact of such a dramatic change in tax law is likely to be trivial (and potentially harmful for Delaware) for two reasons.5
First, Professor Barzuza’s proposal omits any analysis of the legislative process that a proportional tax is designed to influence. There are two parts to this, roughly corresponding to the supply and demand of legislation. Legislators supply legislation, but Professor Barzuza offers no account of their motives to explain why the increase in state revenues that would come from a proportional tax would benefit the marginal Delaware legislator. On the other side of the coin, managers are part of the demand for corporate legislation, but they too are missing from Professor Barzuza’s calculus. I consider each in turn.
Professor Barzuza notes that Delaware will be rewarded with higher taxes and this will encourage it to enact optimal (or more optimal) legislation. The problem is that Delaware can act only through its legislators, and these individual legislators are missing from Professor Barzuza’s account. Without a coherent claim about how legislators respond to the various incentives created by the parties in the legislative process, the argument she makes is less persuasive. The point here is simply that once we move from the single-legislator example above to a multi-member body, the calculus of weighing the benefits from a change in corporate governance is more complex. For example, do legislators only care about the size of the public fisc? Perhaps for some legislators sitting on key committees, the ability to pass benefits along to constituents may help them get reelected, but for others the opposite may be true. Legislators face constituencies with heterogeneous preferences, not all of which will view increased state revenues as an unmitigated positive. Assuming that the utility function of the marginal legislator rises with increased tax revenues seems, at best, overly simplistic. Is it not just as likely that legislators in multi-member bodies might care about maximizing other things, such as their chance at reelection or their personal influence or prestige? Passing shareholder-wealth-maximizing legislation may have little to no effect on such considerations.
Legislators who could not (or would not want to) take credit for increased state revenues would not be influenced by a proportional tax. One might argue that the state could use the increased cash from corporate taxes to reduce taxes on other entities, such as individuals. But this argument needs a theory of why increased taxation of firms relative to individuals is more efficient. Optimizing the mix of tax funding sources is a difficult calculation, considering the relative ability of firms and individuals to evade taxes (say, through structuring, compliance, or leaving the jurisdiction entirely), the impact on incentives to produce (that is, the choice between work and consuming leisure), the impact on other tax burdens (such as federal taxes and sales taxes), and so on. Substituting corporate taxes for individual taxes might seem desirable, but it could create unintended consequences or dry up the tax base in ways that might make it difficult to replace because of the political stickiness of tax rates for both individuals and corporations.
Another problem on the supply side involves uncertainty about whether legislators really have an incentive to prefer higher tax revenues over a higher number of incorporated firms. There may in fact be an inverse correlation between revenue and the number of charters, and the marginal legislator might sensibly prefer to have more companies chartered in Delaware than to maximize the treasury (or minimize other tax burdens). Maximizing the number of firms may mean more work for lawyers, judges, and other service providers in Delaware, and thus increase campaign contributions to and the prestige of legislators responsible for attracting firms to the state. The public choice calculations about what legislators maximize is far from clear, and not obviously pointed in the direction of "better" firm governance, even in a world of increased monetary incentives for the state as a whole.
Professor Barzuza’s account also leaves managers out of the legislative process, and thus overestimates the potential impact that increased revenues will have on overall legislative incentives. Professor Barzuza notes that managers are powerful players in the current legislative process (in fact, strong enough to distort it in perverse ways),6 but then underestimates the role they will have in objecting to any legislation that, while designed to increase shareholder value, may destroy manager value. This is especially odd, since the argument for the tax change is premised on how powerful managers are. Why would this power to influence legislators wane under a new tax regime? Presumably the answer is because the legislators now have a larger incentive (because of the increased tax revenues) to resist the managers. But one must compare the relative impact of the new tax revenues and the power of managers in the new world, and it is not at all clear that the new incentives will be anywhere near strong enough to make a difference. If the ability of managers to resist improvements in the law that would increase shareholder value is as significant as Professor Barzuza believes, taxes are unlikely to do anything to change the situation. To see this, consider a simple example.
Imagine that the Delaware legislature is considering a bill that would require firms to destagger their boards. Professor Barzuza cites evidence suggesting that this would increase shareholder value by $40 billion. We must consider the gains to legislators from both passage and defeat of particular legislation. If the bill is passed, Delaware’s treasury will receive $40 billion multiplied by some tax rate, T. Legislators who vote for the bill will benefit derivatively from this, receiving some benefit, B, for increasing state revenues. B is, by definition, less than $40 billion times T, because the gains are divided up among many legislators, there exists some question about which legislators get "credit" for bills, and the money is flowing to the state (and the people) instead of directly to the legislators themselves.
Managers will try to influence legislators too, by delivering benefits—call them B*—to individual legislators. We can measure the upper bound of this influence by estimating the value managers would have from maintaining the status quo. Professor Barzuza claims that firms, acting through managers, do not have incentives to destagger their own boards, because managers prefer the private benefits of control, which would be diminished if the board were destaggered. To put a dollar amount on these private benefits one need only estimate the dollar gains managers would share with shareholders if the board were destaggered. Assuming managers own, on average, 5% of firm shares, the managers would gain about $2 billion from the change ($40 billion x 5%), and thus the private benefits of control must exceed that amount. This means managers would be willing to pay over $2 billion to avoid the legal change; this is the upper limit of B*.
To determine whether legislators will have incentives to pass shareholder-friendly legislation, we simply compare B* to B. B* is likely to be much greater than B for an individual legislator, if for no reason other than the fact that B* is a direct benefit whereas B is, in most cases, an indirect one. In fact, the direct nature of the benefit for legislators may be one reason why managers are currently able to exert a disproportionate influence on legislators compared with diffuse (and generally disinterested) shareholders.
This approach also allows us to estimate the tax rate necessary for B to exceed B*. As noted above, the managers would be willing to "pay" $2 billion to legislators to avoid this law to preserve their private benefits of control.7 In the extreme, this means that the tax rate, T, would have to exceed 5% for legislators to favor the bill. This would be an absurdly large percentage of firm value to demand in taxes, and one unlikely to be politically feasible.
Second, even assuming that these issues are solved, there remains the question of how realistic it is for legislators to make judgments about what does and does not increase shareholder value. At some level, this is what legislators are supposed to do, but Delaware’s corporate code is remarkably devoid of governance dictates, and its legislators have little experience in this policymaking area. The conceit of the current code is to leave it to the parties to contract from a bare base to those changes that will improve value. Although the supposition that parties will actually bargain or have incentives to strike efficient deals may be questioned, it is not at all clear that legislative incentives are the problem. After all, how are legislators to measure the merits of various academic studies suggesting governance improvements? The literature is rife with claims that doing X, Y, or Z will improve shareholder value, but also counterclaims on the merits or on theoretical grounds. Empirical scholarship is increasingly impenetrable by non-specialists, and, in the event of hearings on the merits of X, Y, or Z, we can be confident that there will be as many adamant pros as cons, and even more estimates on the potential impact on firm value.
In addition, there is no way for legislative judgments about the impact of X, Y, or Z to be evaluated ex post, since numerous other variables, like general economic conditions, competition in the industry, and other regulations, may impact firm profitability. This means that there will be no (or very noisy) feedback on the efficacy of governance changes and the merits of the proportional tax scheme. This may undermine the political support for the tax or for particular governance changes, since causation will be so uncertain.
These problems simply raise the question of why legislators should prefer making these judgments, instead of leaving them to firm owners and managers. It is unlikely that the sum of decision costs and error costs is less for legislators and courts than for managers and shareholders. Legislators simply have no experience with this kind of analysis, as shown by the history of corporate legislation in Delaware. Moving to a new paradigm where legislators make governance choices will require overcoming this deficiency in skills and information, as well as the inertia of the current system. This means that the incentives, especially at the beginning of the new regime, will have to be much higher than would be necessary in equilibrium. As a result, the political resistance to getting such a plan started may be greater than one would think if one were simply evaluating the steady-state case, and a new regime less likely to emerge than it would be if Delaware legislators routinely made corporate governance calculations.
Unlike legislators, firm stakeholders are betting their own money, careers, and reputations, and are likely to know the idiosyncratic circumstances of their firms. Firm-specific changes in governance are more likely to be narrowly tailored to firm and/or industry circumstances, are more likely to be capable of ex post analysis and reconciliation, and are more responsive to market forces that will weed out good from bad governance choices. In addition, the firm may be the only sensible locus of judging governance. Studies showing that certain governance changes (for example, smaller boards or separating the chair and CEO roles) will increase firm value may be biased by omitting unobservable variables at a firm level,8 and therefore may be yielding false results or ones that are not generalizable across all firms.
In light of these problems, delegating this job to legislators seems sensible only if the other mechanisms for enacting governance changes that will increase firm value (while not doing other harms) are irreparably broken. Instead of giving the power to legislators, who know less than managers, shareholders, and creditors, why not advocate repealing the Williams Act, changing the rules about how firms repay costs in proxy battles, or instituting any number of other reforms that would keep the burden on firm stakeholders to make these decisions?
Professor Barzuza’s insight makes an important contribution to how we think about the interplay between taxes and regulation, and what we view as the most appropriate ways to optimize corporate default rules. The changes she envisions are unlikely to overcome the managerial power she presupposes, however, and in any event are a clear second best to a world of few mandatory rules and robust freedom of contract. Focusing on improving mechanisms of private ordering, rather than getting caught in legislative battles, seems to be a more sensible method of improving corporate governance.
Preferred citation: M. Todd Henderson, The Impotence of Delaware's Taxes: A Response to Barzuza's Delaware's Compensation, 95 Va. L. Rev. In Brief 49 (2009), http://www.virginialawreview.org/inbrief/2009/05/16/henderson.pdf.
[Back to this issue]