Tagged: worst case scenarios


The Most Interesting Part of Driesen’s Economic Dynamics of Law is the Focus on Change Over Time

In his new book The Economic Dynamics of Law, David Driesen works hard to focus the reader on how policies shape and are shaped by change over time. This is what I find most helpful and distinctive about Driesen’s approach: that he really rolls up his sleeves and tries to think out the intertemporal implications of legal policy. This is a difficult and complex task, and Driesen makes some real inroads, particularly in his treatment of the challenges of systemic risk. He also makes a valuable start at interrogating the distributional questions of how policies impact power dispersal—a start that could be refined still further by adopting Outka’s suggestion of extending this analysis backwards in time, and by addressing the possibility of intergenerational power dynamics (which Driesen ignores even in his recent clarifying comment).

Beyond these observations, I have a few suggestions to make in further refining his temporal approach, and then a few additional reflections on the structure of his argument.

The first suggestion is also a shameless plug for a work-in-progress of mine called Reregulation and the Regulatory Timeline, where I too am struggling with how to operationalize policy choice over time. My approach is to contextualize policy decisions along a timeline, which can then be used to illustrate not only the path-dependence that arises from past decisions’ impact on current decisions, but what I call “intertemporal dependence,” or the fact that current policy decisions can also strategically account for future path dependence. The world isn’t the only thing that changes over time; legal policy also changes, and policymakers should be encouraged to recognize that their decisions today will impact the costs, benefits, and implications of policy decisions far into the future.

Driesen recognizes that costs and benefits shift over time, but he makes this point merely to trivialize the usefulness of equilibria (56). But another—more dynamic?—view of the fact that a policy’s distribution of benefits and costs often change over time would be to note that equilibria shift with costs and benefits, and to prescribe that a dynamic policy would do well to allow behaviors to shift with the landscape of costs and benefits. In this sense—and I return to this point later—Driesen’s fixation on problematizing the underpinnings of neoclassical economics leads him to overlook a valuable application of his approach to cost-benefit analysis.

What is Driesen’s approach—or at least, what kind of decision procedure is dynamic economic analysis? As Driesen himself recognizes, it is essentially a narrow and future-oriented form of the precautionary principle (59). I would add that his focus on avoiding systemic risk strikes me as very similar to the narrow precautionary principle that Cass Sunstein calls the “Catastrophic Harm Precautionary Principal” in Worst Case Scenarios. Sunstein describes his principle this way: “When risks have catastrophic worst-case scenarios, it makes sense to take special measures to eliminate those risks, even when existing information does not enable regulators to make a reliable judgment about the probability that the worse-case scenario will occur.” As Sunstein points out, whether this principle is helpful is a case-specific question that depends on three things: when information will trigger the principle, the role of costs, and the role of any probabilistic information that does exist (WCS 119-120). I worry that Driesen’s anti-systemic-risk precautionary principle is subject to the exact same concerns.

That said, Driesen’s approach is more complex than a single narrow precautionary principle, because he would also have policymakers adopt a second precautionary principle, one that seeks to minimize the risk of shutting down future important opportunities for economic development. As Frischmann has noted, Driesen unfortunately spends less time developing this theme than on systemic risk. But its addition adds significant nuance to Driesen’s approach, even as it adds the potential for difficult value tradeoffs. Driesen’s recognizes these challenges (70-73 + some later discussion in examples), and tries to deal with them by subordinating the economic-development principle to the systemic-risk principle (72). But for me this was insufficient: I was still left wondering particularly about the role of costs. Driesen urges policymakers to consider the shape of change over time, and (building on Douglass North’s work on adaptive efficiency) to leave options open in the face of uncertainty (e.g. 148). But how much should policymakers invest in purchasing those options?

Flexibility is rarely free: at the least, it increases current and future decision complexity, both for policymakers and for industries and people affected by the flexible policies. And often the option itself requires some investment to be established. How much investment is justified in options to reduce systemic risk, given that the cost of these options may well cut into economic opportunities? Again, I think that considering both sides of this ledger—the systemic risks of a policy as well as what Frischmann calls the systemic benefits—makes Driesen’s approach far more nuanced than a unilateral normative commitment that would not necessarily invoke these conflicts. But without some idea at least of how to manage costs (e.g. through an application of option pricing theory?), it is difficult to know how to mediate the tensions here.

All of this makes me agree with Livermore that DEA as a standalone procedure, and as it is currently operationalized, is a bit squishy. But I don’t think this means that the larger theory underlying DEA, understood as the combination of two narrow future-oriented precautionary principles, fails to give any useful guidance to policymakers. I just think that the primary benefit of its guidance is that it encourages systematic reflection about impacts of policies through time. In this sense, Driesen’s preoccupation with cost-benefit analysis and neoclassical economics is largely a distraction from what is most novel and interesting in his approach. I would have far preferred to see an account for dynamic analysis—for analysis sensitive to the change over time that occurs in both legal structures and in world conditions—that was severed from a treatment of cost-benefit analysis.

What’s the real connection between DEA and CBA, if there is one? I think it’s actually this: the more critical a person is of how CBA works as a decision procedure, the more likely she is to welcome alternative decision procedures of any kind. And conversely, the more satisfied a person is with how CBA works as a decision procedure, the less need he is likely to see for additional procedures. But this doesn’t tell us much of anything about DEA: it just tells us about how people felt about CBA before DEA showed up on the scene. That suggests DEA just isn’t giving us much traction on the CBA question. Which is fine. But is also a reason that we may all get further if we recognize, as Livermore points out in in his first post, that DEA can be used either in concert with CBA or not. And that either way, DEA is valuable for promoting further systematic thinking about policymaking in the face of change over time.