« Judge John T. Noonan, Jr., RIP | Main | Justice Alito, funny man »

Thursday, April 20, 2017

No Simple Algorithm

In his article Libertarian Separation of Powers, Aziz Huq asks whether a libertarian should prefer one instantiation of interbranch structure over another.  His answer is no:  What institutional arrangements maximize liberty is deeply contingent.  “Confident prediction and prescription require a high degree of historical and circumstantial tailoring. There is no facile algorithm.”

Something loosely similar, I’ve been suggesting in previous posts, is plausibly true of the class action.  For libertarians, there’s no simple algorithm for regulating  class certification.

In this post, I’ll flesh out the point further by turning to the current version of the Fairness in Class Action Litigation Act ("FICALA"), Congress's latest attempt at class action regulation. To blog-simplify, I’ll assess the Act from the standpoint a crude libertarian—someone whose strong preference for private ordering leads him to embrace very crude constraints on state intervention in the marketplace. 

The virtue of this heuristic is not that it tells us whether reforms in the Act are good, bad, or indifferent, but that it makes the following claim at least colorable:  Even confined to a single category of Rule 23 class action, Rule 23(b)(3), it’s not clear that there’s one set of optimal libertarian class certification rules.

I’m going to focus on one portion of the FICALA:  its provisions on partial certification or “issue classing.” Partial certification, grounded in Rule 23(c)(4), involves slicing class claims into their component issues and then certifying discrete common issues (most often, issues related to primary conduct or general causation), leaving individualized issues relating to specific causation and damages calculation to later proceedings brought by individual class members. 

FICALA, following circuits like the Fifth, forbids partial certification—class certification, it provides, should hinge on the certifiability of the class claims considered as a whole.  This turns Rule 23(c)(4) into the most banal of housekeeping provisions—one that allows the court to segment and hold separate trials on a series of common issues in a thoroughly cohesive class. 

How might the crude libertarian think about FICALA's partial certification ban? 

The crude libertarian has spent a lot of time reading, although not carefully, the literature on “regulation by litigation.”  The literature defines “regulation by litigation” as the “use of the threat of a catastrophic loss in litigation to coerce agreement to forward-looking, [quasi-administrative] regulatory provisions in a settlement.” The problem with regulation by litigation, according to this literature, is that it evades the usual checks on regulation built into the political process, imposed by administrative procedure, or what have you.

Our crude libertarian sees class certification through the lens of this literature and, characteristically, does so in blunt, un-nuanced terms:  Damages class actions are, in his mind, one thing—an opportunity for regulation by litigation (by, he likes to tell you, “politicized” judges.)  The crude libertarian isn’t picky about how to deal with that opportunity.  He's happy so long as judicial discretion to take advantage of it is locked down.

And so the crude libertarian embraces the contemporary version of the Rule 23(b)(3) predominance test, which, as Robert Bone laments here,  requires a very high degree of class cohesion in a way that imposes a rule-like constraint on certification. The same impulses, of course, lead the crude libertarian to reject partial certification, because it allows judges to evade the predominance requirement.  In the crude libertarian's view, partial certification uncorks the judicial ”regulatory” discretion that the modern predominance test tries to bottle up.

Or, at least that’s how things look to crude libertarian when he focuses on the corporate wrongdoing class action.  What, though, happens if the crude libertarian turns his attention to, say, constitutional tort litigation? 

The crude libertarian, it turns out, is pretty hawkish on monetary claims for constitutional torts.  Here, after all, he thinks, litigation protects rather than disrupts autonomy and private ordering.

Of course, he knows that imposing monetary liability on municipalities has costs as well as benefits.  He’s not sure where the tipping point, where costs outstrip benefits, is.  But he’s sure we haven’t reached it yet.  Indeed, recent political trends have given him the distinct fear that the pendulum may have swung too far in the anti-liability direction.

And so when he turns his attention to class actions based on constitutional torts, the Fairness in Class Litigation provisions on partial certification seem. . .  problematic.   Sure, section 1983 class litigation for damages is especially rare—but it exists and frequently depends on the use of partial certification.

Applied in this setting, then, partial certification seems not bad but, plausibly, good.  Sure, partial certification occasionally misfires.  But the occasional misfire is a small price to pay for  giving the deterrent threat of monetary liability for constitutional torts some needed extra bite.

The virtue of the crude libertarian isn’t that his judgments are all spot on.  They're not.  The crude libertarian is a crude simplification.  But it's a simplification with a purpose, e.g. pointing up a possibility:  Much as there is no “facile algorithm” for maximizing liberty through macro interbranch structure, there is no single “libertarian” algorithm  for regulating damages class actions. The best class certification rules, from a libertarian perspective, may vary--alternatively constraining and empowering damages classes--in different contexts. 

Thanks to his crudeness, the crude libertarian can’t  give us firm answers about how to structure the class action in these contexts.  But, in a future post, I’ll suggest he helps us see some common ground between progressives and libertarians about who ought to make that design decision. 

Posted by Mark Moller on April 20, 2017 at 11:40 PM in Civil Procedure | Permalink


The comments to this entry are closed.