Irreparable Benefits: An Example

Thanks to everyone who emailed or posted with comments on yesterday's first entry.  Tomorrow, I hope to join in on several of those threads and talk about the issues they raise.  Today, however, I had promised an example that might make my argument more concrete.

To that end, suppose that the plaintiff in a given case holds a patent on a chemical process shown to significantly reduce the rate of genetic mutation in a certain type of animal cell. The process at the moment has no specific medical application in humans, but the plaintiff believes that the process will ultimately mature into an important human therapy. The defendant, meanwhile, recently began work on a similar chemical process, also hoping in the end to find applications related to human ailments. The plaintiff’s legal allegation is that the defendant’s process infringes the patent and the defendant therefore should not be permitted to engage in further research without permission. The defendant’s response is that its research is permissible, either because its process does not fall within the scope of the patent’s claims or because the patent is invalid in light of the prior art.

If the patentee moves for preliminary relief,  the first hurdle will be to show that something irreparable is at stake. Patent harms are not literally irreparable—most patent-related injuries can be fully compensated by some ex post cash payment—but they are typically deemed irreparable because they are difficult for courts to value. I will say more about this argument later, but for now note that almost every “irreparable” harm is actually irreparable only in this limited sense. Speech harms are an obvious example. Limitations on speech are routinely characterized as harms for which there can be no adequate compensation.  However, a civil rights activist hoping to stage a peaceful demonstration on a particular Sunday would surely willingly trade that opportunity if compensated by the funding necessary to sponsor, say, ten demonstrations the following week, or, better yet, a mixture of demonstrations, public hearings, and other means through which his message might be heard and thoughtfully evaluated. The real problem with most speech harms is therefore not that after-the-fact cash remedies cannot make the relevant victim whole, but instead that courts cannot reliably estimate the size of the necessary cash payment. The same difficulty applies to most patent harms, and patent harms are for this reason today presumptively eligible for preliminary relief.

With irreparable harm shown, the next step in the analysis is to apply the three classic factors and therefore to consider: (1) the likelihood that the plaintiff’s patent is valid and infringed; (2) the irreparable harm that would be imposed on the defendant by a wrongful court order to halt its research; and (3) the irreparable harm that would be imposed on the plaintiff were the court to deny relief only to later rule that the patent was in fact valid and infringed. The first factor requires little explanation. If the plaintiff’s case is a slam-dunk, the injunction should issue immediately, regardless of the relative irreparable implications. Conversely, if the plaintiff’s allegations are preposterous, no injunction should issue, again regardless of the irreparable consequences. Both of these conclusions follow from the simple fact that preliminary relief is not meant to contradict the outcome on the merits. If the merits are clear, the court’s decision with respect to preliminary relief is also clear—at least in all but the most unusual circumstances.

Things are more interesting in cases where the merits are murky, because in those cases a court must confront the possibility that its ruling with respect to preliminary relief will ultimately turn out to be inconsistent with its final judgment on the merits. This is where the standard for preliminary relief does its heavy lifting. Start with the possibility that the court will refuse to issue the injunction at the preliminary stage but then, after a full hearing on the merits, conclude that the defendant’s research did in fact infringe. As the traditional analysis suggests, one cost associated with this errant denial is any irreparable harm that might be suffered by the patentee. This is a private cost suffered by the complaining patent holder, but it is also a social cost in that mistakes like this will over the long run dampen the ex ante incentive to pursue patent-eligible research, discourage patent holders from litigating even valid claims, and likely drive inventors to invest more heavily in costly self-help protections.

There is another cost associated with this errant refusal to enjoin, however, and that is the irreparable benefit that accrues to the infringer. This cost is one that the traditional analysis overlooks, yet—like the irreparable harm normally considered—it, too, has unintended private and social consequences. Undeserved irreversible gains skew the defendant’s incentives with respect to the question of whether to litigate or settle.  They also encourage the defendant to invest further in his research, a wasteful outcome in cases where that research will ultimately turn out to be impermissible. Most importantly, undeserved irreversible gains undermine the defendant’s incentive to “invent around” the patent rather than infringing it. This latter implication is of particular consequence given that the patent system is designed to encourage innovation not merely by rewarding patent holders for their accomplishments but also by forcing rivals to discover comparable, non-infringing substitutes for patented inventions.

Turn now to the opposite category of court error, namely an instance where the court issues an injunction at the preliminary stage but then, after a full hearing on the merits, concludes that the accused research was in fact permissible. The traditional analysis focuses exclusively on any irreparable harm imposed on the defendant. For example, the defendant might be forced to substitute an alternative chemical process for the disputed one, or the defendant might have to freeze its research project for the duration of the litigation. Again, to the extent that ex post remedies will suffice to undo the effects of these changes, the court can safely ignore them.  However, to the extent that these harms have lasting implications—private harms suffered by the defendant and societal harms associated with the unintentional disruption of research that the patent system meant to allow—the conventional approach to preliminary relief will rightly count them as reasons not to issue the injunction.

My point, predictable at this stage, is that similar private and social concerns arise with respect to the plaintiff’s wrongful but irreversible gains. An errant injunction could translate into several years of exclusive rights that the patent system did not intend to recognize. Maybe the implications of those bonus years can be undone by some accurate ex post financial penalty. Where valuation is a problem, however, this irreparable benefit is yet another cost associated with an errant decision to enjoin. It is a deviation from the intended ruling on the merits, and, like the irreparable harm that might be wrongly imposed on the defendant, it has distributional implications for the affected party and dynamic implications for society more generally.

With irreparable harm shown, the next step in the analysis is to apply the three classic factors and therefore to consider: (1) the likelihood that the plaintiff’s patent is valid and infringed; (2) the irreparable harm that would be imposed on the defendant by a wrongful court order to halt its research; and (3) the irreparable harm that would be imposed on the plaintiff were the court to deny relief only to later rule that the patent was in fact valid and infringed. The first factor requires little explanation. If the plaintiff’s case is a slam-dunk, the injunction should issue immediately, regardless of the relative irreparable implications. Conversely, if the plaintiff’s allegations are preposterous, no injunction should issue, again regardless of the irreparable consequences. Both of these conclusions follow from the simple fact that preliminary relief is not meant to contradict the outcome on the merits. If the merits are clear, the court’s decision with respect to preliminary relief is also clear—at least in all but the most unusual circumstances.

Things are more interesting in cases where the merits are murky, because in those cases a court must confront the possibility that its ruling with respect to preliminary relief will ultimately turn out to be inconsistent with its final judgment on the merits. This is where the standard for preliminary relief does its heavy lifting. Start with the possibility that the court will refuse to issue the injunction at the preliminary stage but then, after a full hearing on the merits, conclude that the defendant’s research did in fact infringe. As the traditional analysis suggests, one cost associated with this errant denial is any irreparable harm that might be suffered by the patentee. This is a private cost suffered by the complaining patent holder, but it is also a social cost in that mistakes like this will over the long run dampen the ex ante incentive to pursue patent-eligible research, discourage patent holders from litigating even valid claims, and likely drive inventors to invest more heavily in costly self-help protections.

There is another cost associated with this errant refusal to enjoin, however, and that is the irreparable benefit that accrues to the infringer. This cost is one that the traditional analysis overlooks, yet—like the irreparable harm normally considered—it, too, has unintended private and social consequences. Undeserved irreversible gains skew the defendant’s incentives with respect to the question of whether to litigate or settle.  They also encourage the defendant to invest further in his research, a wasteful outcome in cases where that research will ultimately turn out to be impermissible. Most importantly, undeserved irreversible gains undermine the defendant’s incentive to “invent around” the patent rather than infringing it. This latter implication is of particular consequence given that the patent system is designed to encourage innovation not merely by rewarding patent holders for their accomplishments but also by forcing rivals to discover comparable, non-infringing substitutes for patented inventions.

Turn now to the opposite category of court error, namely an instance where the court issues an injunction at the preliminary stage but then, after a full hearing on the merits, concludes that the accused research was in fact permissible. The traditional analysis focuses exclusively on any irreparable harm imposed on the defendant. For example, the defendant might be forced to substitute an alternative chemical process for the disputed one, or the defendant might have to freeze its research project for the duration of the litigation. Again, to the extent that ex post remedies will suffice to undo the effects of these changes, the court can safely ignore them.  However, to the extent that these harms have lasting implications—private harms suffered by the defendant and societal harms associated with the unintentional disruption of research that the patent system meant to allow—the conventional approach to preliminary relief will rightly count them as reasons not to issue the injunction.

My point, predictable at this stage, is that similar private and social concerns arise with respect to the plaintiff’s wrongful but irreversible gains. An errant injunction could translate into several years of exclusive rights that the patent system did not intend to recognize. Maybe the implications of those bonus years can be undone by some accurate ex post financial penalty. Where valuation is a problem, however, this irreparable benefit is yet another cost associated with an errant decision to enjoin. It is a deviation from the intended ruling on the merits, and, like the irreparable harm that might be wrongly imposed on the defendant, it has distributional implications for the affected party and dynamic implications for society more generally.

Naturally, there is much more to say on all of these topics. For instance, there are interactions among the various long-run incentives I consider above, and those interactions amplify some concerns but mitigate others.  Moreover, the concept of irreparability is actually significantly more complicated than I have thus far let on; some errors, for example, turn out to be irreversible when they manifest themselves as undeserved losses but fully reversible when they manifest themselves as unearned gains.  For now, however, I want to stake out only one basic proposition: when evaluating a motion for preliminary relief, any deviation from what will be the ultimate resolution on the merits is relevant, no matter whether that deviation is perceived to be a benefit or a harm. All that matters is that the deviation is unintended and irreversible.