Monday, May 3, 2010
Geistfeld on "Tax breaks for bad behavior? On the relation between income taxes, punitive damages, and liability insurance"
They make the nice point that if punitive damages are not deductible, then plaintiffs and defendants have an incentive to “disguise punitive damages as compensatory damages in pre-trial settlements.” Doing so decreases the (after tax) cost of settlement for defendants, creating a gain that can then be shared by the settling parties. By way of extended analysis, Polsky and Markel go on to conclude that the best way to solve the “under-punishment problem” created by deductibility is not to eliminate the tax break, as everyone had previously concluded, but instead to apprise juries of the deductibility issue so that they will “gross up” the punitive award to offset the tax break.
Largely missing from the analysis, however, is discussion of how liability insurance affects the incidence of tort liability. Once this dimension of the problem has been recognized, it becomes apparent that there is a much stronger case against the deductibility of punitive damages.
Consider a world (largely like our own) in which every defendant worth suing has liability insurance covering at least a portion of a tort judgment (or any other form of civil liability that permits the award of punitive damages). Suppose our insured defendant has incurred punitive damages liability. Perhaps surprisingly, this form of liability is not expressly excluded from coverage under the standard-form liability-insurance contracts. Whether the defendant can actually collect on the insurance, however, depends on whether the jurisdiction permits the insurability of punitive damages as a matter of public policy.
Nine or so jurisdictions, including California and New York, prohibit the insurance of punitive damages. In these jurisdictions, any settlement between an insured defendant and the tort plaintiff presumably will allocate the appropriate amount to punitive damages. Regardless of how the defendant and plaintiff would otherwise like to characterize the proportion of compensatory and punitive damages covered by the settlement, the insurer is obligated to indemnify only the former category and accordingly will seek to maximize the portion of the settlement attributable to punitive damages (and excluded from coverage). The insurer usually can police the terms of the settlement directly (the insurance contract gives the liability insurer the right to settle the case). But if the insurer does not fully participate in the settlement, the terms of the settlement would not have preclusive effect in a subsequent coverage dispute with the tort defendant/policyholder regarding the amount of the settlement that is covered by the policy and properly allocable to compensatory damages. The liability insurer, therefore, presumably will monitor the portion of the settlement allocable to punitive damages, effectively precluding plaintiffs and defendants from otherwise manipulating settlements in a manner that would thwart efforts to restore the full “sting” of punitive damages by making them nondeductible.
The argument against deductibility then largely generalizes to the remaining jurisdictions that permit the insurability of punitive damages. The standard-form liability-insurance contracts do not cover liabilities for “expected or intended harms.” In these cases, the insurer can deny coverage altogether—for both compensatory and punitive damages—and so it will not monitor the portion of any settlement properly allocable to punitive damages. In light of the settlement problem identified by Polsky and Markel, the best approach would be to deny deductibility for the entire liability. These instances of intentional wrongdoing clearly implicate the retributive concerns that would create a problem of “under punishment” in the event that the punitive award receives a tax break. Rather than let the litigants manipulate settlements for tax reasons, why not eliminate the tax break altogether for liabilities of this type? Why should these intentional wrongdoers be able to deduct any of their liabilities as a cost of doing business?
Regardless of how one answers this question, the case for nondeductibility remains intact. The public policy concerns implicated by the insurability issue are substantively identical to those posed by the deductibility issue: each allows the tort defendant to distribute the cost of the punitive award to a wider group (other policyholders; other taxpayers). In deciding to permit the insurance of punitive damages, a jurisdiction has concluded that the redistribution afforded by liability insurance does not create any public policy problem of “under punishment.” So, too, in these jurisdictions the redistribution afforded by the deductibility of punitive damages does not create any public policy problem of “under punishment.” Consequently, even if a tax rule of nondeductibility could be largely circumvented by the settling parties as Polsky and Markel conclude, there is no “under-punishment problem” created by the de facto deductibility of punitive damages. This does not mean that the deductibility issue is largely irrelevant. The tax rule against deductibility is still desirable as a federal matter because it furthers the public policy of those states that reject the insurability of punitive damages on the ground that wrongdoers should not be able to redistribute their punishment to others.
Admittedly, I live in a state where punitive damages are not insurable, and the analysis of Polsky and Markel has much more to offer than I have indicated. They artfully unravel the surprising complexity of what appears to be a rather straightforward issue—whether bad behavior deserves a tax break. Clearly, I should have put this article into my “read right away” pile (although that pile, of course, also ends up getting shuffled around at the close of the academic year).
- Mark Geistfeld
Sheila Lubetsky Birnbaum Professor of Civil Litigation
New York University School of Law