IN "Taxing Punitive Damages," Gregg D. Polsky and Dan Markel argue that defendants paying punitive damages are under-punished relative to juries' intentions, because tax-unaware juries do not take into account the fact that the deductibility of punitive damages significantly reduces defendants' after-tax costs. 1 They note that the Obama administration has proposed addressing the under-punishment problem by amending the Internal Revenue Code to disallow deductions for punitive damages (and for settlements paid on account of punitive damage claims) 2 They conclude, however, that the proposal would be ineffective because defendants could avoid its impact by disguising nondeductible punitive damage settlements as deductible compensatory damage settlements. 3 They argue that a superior approach would be to leave federal tax law unchanged and to change jury instructions in punitive damage cases instead. 4 If juries were explicitly told that punitive damages were deductible, they could "gross up" the awards to impose the desired level of after-tax punishment on defendants. In contrast with the Obama administration's proposal, this non-tax, non-federal solution to the under-punishment problem would not be undermined by pre-trial settlements: "Gross ups, in addition to increasing jury verdicts, would increase settlement values because litigants determine these values in the shadow of what a jury would be expected to award." 5
Their argument is powerful and original. It may have dramatic real-world effects, if it inspires plaintiffs' lawyers across the nation to request the jury instructions required to produce tax-aware juries, and if courts grant those requests. In this brief Response, however, I raise two possible objections to their analysis. The first objection is that they do not consider the alternative of a nondeductibility rule applicable to punitive damages but not to settlements of punitive damage claims. This narrower nondeductibility rule is arguably superior to both broader nondeductibility and tax-aware juries. The second objection is that they do not consider how their analysis would change if deterrence, rather than punishment, were viewed as the primary function of punitive damages. Although these are considerably more than quibbles, they do not detract from my view of their article as a major contribution to the scholarly literature on the intersection of torts and taxes, with the potential for significant real-world impact. The Response closes with a brief observation on the relationship between the article, plaintiffs' attorneys, and ten dollar bills on sidewalks.
I. SHOULD THE PROPOSED NONDEDUCTIBILITY OF PUNITIVE DAMAGES EXTEND TO SETTLEMENTS?
As explained above, Polsky and Markel argue that tax-aware juries are preferable to the Obama administration's proposal to make punitive damages nondeductible because defendants could avoid the nondeductibility of punitive damages by mislabeling punitive damage settlements as compensatory damage settlements. The administration's proposal would formally apply to settlements as well as to judgments: "No deduction would be allowed for punitive damages paid or incurred by the taxpayer, whether upon a judgment or in settlement of a claim." 6 Polsky and Markel argue persuasively, however, that settling parties would almost always claim the entire settlement was with respect to actual damages, and that the Internal Revenue Service ("IRS") would have great difficulty in challenging the parties' allocations. 7 Because jury tax awareness suffers from no comparable weakness, Polsky and Markel argue that it is a superior response to the under-punishment problem caused by the deductibility of punitive damages under current law. 8
Their argument is persuasive as a critique of the administration's proposal; under the proposal punitive damage settlements are supposed to be nondeductible, but that supposed nondeductibility is largely unenforceable. Consider, however, an alternative proposal, under which punitive damages would be nondeductible, but a settlement would remain fully deductible—even if the parties explicitly acknowledged that a portion of the settlement was on account of punitive damages. There is a plausible argument—based on principle, rather than merely on the impracticality of identifying the punitive components of settlements—for this alternative proposal.
When a case goes to trial and results in an award of punitive damages, the jury (or judge) has determined that the defendant engaged in conduct sufficiently reprehensible to warrant punishment. If the jury is not tax-aware, allowing the defendant to deduct its punitive damages results in the defendant suffering less punishment than the jury intended. Assuming the courts continue to operate with tax-unaware juries, disallowing deductions for punitive damages is necessary to produce the jury-intended level of punishment. When a case is settled, however, there has been no determination by the legal system that the defendant engaged in punishable conduct, let alone a determination of the proper dollar amount of punishment. It is surely true that, in many settlements with a punitive component (whether acknowledged or unacknowledged by the parties), the likelihood that a jury would have awarded punitive damages if there had been a trial is well below fifty percent. A defendant will often be willing to make a substantial settlement payment to avoid even a small risk of a large punitive damage award at trial. Studies have found that plaintiffs are awarded punitive damages in only about two or three percent of all cases that go to trial—plaintiffs prevail in roughly half of the litigated cases and are awarded punitive damages in four to six percent of the cases in which they prevail. 9 Although settled cases may not closely resemble litigated cases in their potential for punitive damages, 10 the very small percentage of punitive damage awards in litigated cases suggests that a large proportion of defendants making payments to settle punitive damage claims would not have been found liable for punitive damages at trial.
Even if the IRS could identify with perfect accuracy in every instance how much of a settlement was paid with respect to punitive damage claims, in many—probably most—cases of punitive damage settlements the jury would not have awarded punitive damages if the case had gone to trial. If (1) the legal system has made no determination that the taxpayer engaged in conduct warranting punishment, (2) more likely than not, the legal system would not have made such a determination had the case proceeded to trial, and (3) there is no way of knowing what the amount of punitive damages would have been if punitive damages had been awarded, then it is far from obvious that the taxpayer will be under-punished if it is allowed to deduct the entire amount of the settlement. There is a strong argument that disallowing a deduction to avoid under-punishment should be limited to cases in which a jury (or judge) has actually determined (1) that the taxpayer engaged in punishable conduct, and (2) the appropriate dollar amount of punishment. If that argument is accepted, the Obama administration's proposal should be revised to exempt settlements from its scope—for reasons that have nothing to do with any concerns about enforceability of the broader disallowance.
There is a close precedent for such a limitation on the scope of a deduction-disallowance rule intended to prevent under-punishment. Section 162(g) of the Internal Revenue Code, enacted in 1969, denies a deduction for two-thirds (that is, the punitive portion) of antitrust treble damages, but only if the taxpayer has been convicted of a criminal violation of federal antitrust laws (or has pleaded guilty or nolo contendere in such a case). According to the 1969 Report of the Senate Finance Committee, "This means that the deduction is to be denied only in the case of 'hard-core violations' where intent has been clearly proved in a criminal proceeding." 11 The argument here is similar. Only if it has been "clearly proved in a [civil] proceeding" that the taxpayer has engaged in conduct warranting punitive damages has the legal system made a punishment determination the integrity of which must be ensured by the denial of a tax deduction.
Because of the design and rationale of this alternative proposal—applying only to punitive damage judgments, because the logic of the proposal extends no further—the proposal is not undermined when defendants avoid the disallowance rule by settling their cases. Unlike the administration's proposal, the revised proposal would function in practice just as it is intended to function in theory. Polsky and Markel claim that jury tax awareness is superior to the administration's proposal as a means of achieving the goals of the administration's plan, because of the unenforceability of the settlement aspect of the proposal. Theirs is a powerful critique of the administration's proposal, but it does not apply to the revised proposal suggested here. In the absence of any disconnect between the purpose and the actual operation of the revised proposal, there would be no reason to look to jury tax awareness as a non-tax, non-federal substitute for deduction denial. In short, if one is persuaded that the policy concern about under-punishment is adequately addressed by a deduction denial that does not extend to settlements, then there is no reason to consider a non-tax solution. Indeed, the tax solution is far superior, if for no other reason than Congress's ability to enact it in one fell swoop, rather than depending on the jury-awareness decisions of the courts of the fifty states.
II. PUNITIVE DAMAGES AS DETERRENCE
Polsky and Markel base their analysis on the assumption that punitive damages are fundamentally about punishing defendants. In the first sentence of their article they state, "As the name implies, punitive damages are principally awarded to punish defendants for torts committed with a malicious or reckless state of mind." 12 The preceding Part of this Response accepted that characterization for the sake of argument. This Part, however, questions that characterization and explains that the analysis changes dramatically under a plausible alternative characterization.
In a footnote supporting the claim of their first sentence, Polsky and Markel include quotations from three Supreme Court cases, all of which indicate that the purposes of punitive damages are to punish and to deter. 13 This view of punitive damages as serving a deterrence function is not peculiar to the Supreme Court. The standard account of punitive damages as a deterrent focuses on a defendant engaging in a dangerous course of conduct. The defendant's conduct harms a number of persons, but for one reason or another the defendant escapes liability for most of the harm. Suppose, for example, the defendant's course of conduct imposes $1 million of harm on each of ten victims, but only one of the ten successfully sues the defendant. If the defendant owes that successful plaintiff only $1 million of actual damages, the defendant will pay for only ten percent of the total harm and will be greatly underdeterred. Adding $9 million of punitive damages to the successful plaintiff's actual damages, however, would force the defendant to internalize the cost of all the harm it caused, thereby producing the appropriate level of deterrence. 14 The punitive damages serve as a substitute for the $9 million of actual damages the defendant does not pay its other victims.
If the defendant had paid $1 million actual damages to each of its ten victims, the defendant would have been entitled to deduct the entire $10 million. That is true under current law, and would remain true under either the Obama administration's proposal or the proposal suggested above. That result—full deductibility of the $10 million—is also appropriate as a matter of deterrence. In deciding how much to spend on accident prevention, the defendant will be comparing the cost of tax-deductible safety precautions with the alternative of damage payments. Given the clear deductibility of safety precautions as business expenses, the damages alternative must also be deductible if the defendant is to have the incentive to take safety precautions up to—but not beyond—the cost-effective level. Nondeductibility of damages would create a tax preference for safety precautions over damages, resulting in overdeterrence (that is, excessive spending on precautions).
Nothing in the above analysis changes if the defendant pays the entire $10 million to one victim (as $1 million actual damages and $9 million punitive damages), instead of paying $1 million to each victim. The proper level of deterrence will be achieved if, and only if, the defendant is allowed to deduct the entire $10 million. In short, under the deterrence theory of punitive damages, the current deductibility of punitive damages (and settlements) is entirely appropriate and should not be legislatively altered. Moreover, under the deterrence theory there is no reason to inform the jury that the defendant will be able to deduct punitive damages. If a tax-unaware jury is provided with the evidence (relating to the number of uncompensated victims of the defendant and the extent of their harms) and the jury instructions (explaining the deterrence function of punitive damages) necessary to implement the deterrence theory of punitive damages, it will produce the correct $9 million punitive damage award in the hypothetical case. Informing the jury that the defendant can deduct the $9 million would serve no purpose; it would only create a risk that the jury would incorrectly increase the award above $9 million to produce a $9 million after-tax cost to the defendant.
Polsky and Markel assume that punitive damages are about nothing but punishment, and argue—quite persuasively, if one accepts their premise—that either punitive damages should be nondeductible or (better yet) juries should be made tax-aware. The discussion here has argued that deterrence is a significant purpose of punitive damages, and that neither nondeductibility nor tax-aware juries would be appropriate if punitive damages were about only deterrence. In reality, of course, punitive damages are about both punishment and deterrence. The relative significance of the two purposes is crucial to the evaluation of the merits of proposals for both nondeductibility and jury tax awareness. If one agrees with Polsky and Markel that the punishment rationale is far more important than the deterrence rationale, then one should favor either nondeductibility or tax-aware juries. If one believes that the deterrence rationale dominates the punishment rationale, then one should favor neither reform proposal. If one believes that the two rationales are of roughly equal importance—unfortunately, a very reasonable thing to believe—then formulating policy prescriptions will be considerably more difficult.
III. IS THAT TEN-DOLLAR BILL REALLY THERE?
There is an old joke about two economists walking down the street. "One sees a $10 bill lying on the sidewalk and asks, 'Isn't that a $10 bill?' 'Obviously not," says the other. 'If it were someone would have already picked it up.'" 15 Polsky and Markel claim that state courts would generally be receptive if plaintiffs' lawyers were to request jury instructions designed to produce grossed-up punitive damage awards, 16 but that it has never occurred to any plaintiffs' lawyer to make the request. 17 If they are right, the oversight by plaintiffs' lawyers is truly astounding. The argument for grossing up punitive damages is straightforward enough, and plaintiffs' lawyers are not known for their propensity to leave money on the table (or sidewalk). The oversight is particularly remarkable given that a closely analogous issue—whether juries should be informed that plaintiffs' personal injury damages are not subject to federal income tax—has been litigated in a number of jurisdictions. 18
One of two things must be true—either Polsky and Markel are wrong in predicting that courts would be receptive to plaintiffs' lawyers' arguments for tax-aware juries, or plaintiffs' lawyers have been stunningly remiss in failing to make those arguments. The latter is certainly possible. After all, the old economists' joke is funny precisely because sometimes there really is a ten-dollar bill on the sidewalk. On the other hand, this is not a matter of a ten-dollar bill being on the sidewalk for a few minutes; this is a matter of plaintiffs' lawyers having failed for decades to pick up millions upon millions of dollars. If that turns out to have been the case—if courts are receptive to the tax-aware jury arguments when plaintiffs' lawyers finally make them—the puzzle of the plaintiffs' lawyers who did not pick up the millions should be high on the research agenda of anthropologists of the legal profession.
Preferred citation: Lawrence Zelenak, Of Punitive Damages, Tax Deductions, and Tax-Aware Juries: A Response to Polsky and Markel, 96 Va. L. Rev. In Brief 61 (2010), http://www.virginialawreview.org/inbrief/2010/10/25/zelenak.pdf.
[Back to this issue]