[Ed. Note: On Friday, May 2 and Saturday, May 3, 2014, the Petrie-Flom Center hosted its 2014 annual conference: “Behavioral Economics, Law, and Health Policy.” This is an installment in our series of live blog posts from the event; video will be available later in the summer on our website.]
In this next installment of today’s live-blogging of the conference (and with all of the caveats of live-blogging mentioned by my colleagues and my apologies for any errors or misrepresentations) we have Professors David Hyman (DH), Mark White (MW) and Andrea Freeman (AF) in a panel moderated by Glenn Cohen (GC) on the “Potential Problems and Limits of Nudges in Health Care”.
The panel began with DH, H. Ross & Helen Workman Chair in Law and Director of the Epstein Program in Health Law and Policy, University of Illinois College of Law, and a talk entitled, “what can PPACA teach us about behavioral law and economics” (Patient Protection and Affordable Care Act). DH began with the observation that nudges often work quite well… “unless they don’t”. While many nudges are “sticky”, i.e. they influence behavior in the way they were intended, others are “slippery”, i.e. they fail to influence behavior in the way they were intended. His talk set out to illustrate the phenomenon, and to pose two questions. The first was an empirical question: what makes a nudge sticky vs slippery? The second was philosophical: is it meaningful to talk about a “failed nudge” or when we do, do we really just mean failed marketing? He focused on an analysis of PPACA as a case study.
First, he pointed out that PPACA contained several components structured as opt-in but with various incentives as nudges: exchanges, Medicaid expansion, the community living assistance services and support act (CLASS), the employer mandate, and the option for coverage of those under 26 on their parent’s health plan. Did any of these nudges “work”? 35% of states opted into and 8 million individuals enrolled in the exchanges, 51% of states opted into Medicaid expansion, CLASS was killed for political reasons, it is too early to evaluate the employer mandate, and though we know little about the utilization of extending the amount of time young people can stay on their parents’ plans we know that passing that part of the act went smoothly. DH pointed out that it is not at all straightforward to tell from the above figures whether these opt-in choices with incentives “worked” and we don’t know how much of a difference it would make if we had flipped the default to opt-in.
DH’s conclusion was that we “need a better theory of nudging” if we hope to be able to tell when they work and when they don’t. He pointed out that while there have been good studies on individual nudges, we need more work on general characteristics of nudges that work and those that don’t. Without this data, perhaps, we should be modest in how much we think we can accomplish with nudges.
He concluded with an interesting philosophical critique of the whole idea of “failed” vs “successful” nudges. Where’s the libertarian in libertarian paternalism, he asked, if nudge designers will just re-tool the nudge to make it “work better” if you are not making the choices they want you to make?
Next, Mark White (MW), Professor and Chair of the Department of Philosophy, College of Staten Island, CUNY, considered whether there are special circumstances in health care that make health-nudges more acceptable than other sorts of nudges. First he clarified that he understood nudges as rearrangements of choice architecture to guide people to make better choices according to their own interests (not the interests of public health officials, etc.). This talk expanded upon the general framework of his recent book, which posited three main objections to the use of nudges:
1) Epistemic: designers don’t know what our true interests are, so can’t make good nudges; this is especially true if we assume complex heterogeneity in individual interests.
2) Ethical: once we acknowledge the epistemic objection, then we grant that the interests promoted by nudge-attempts are imposed by the designers and infringe on autonomy.
3) Practical: If we accept MW’s definition of nudges, the epistemic objection and the ethical objection, we must concede that actual nudges cannot be done.
He argued that even considering how imperfectly we know our own interests, we are still in the best position to make choices and should be free of attempted nudges. But,he asked, does this conclusion hold when we look at the special case of health-care?
MW posited four ways that nudges regarding health might be different: 1) health is an interest to all of us, 2) choices in healthcare are difficult and complicated, 3) they are emotional wrought decisions, and 4) the decisions must be implemented over time (e.g. follow drug regimen). These are great reasons for the uniqueness of decisions in healthcare, he observed, but what he disputed was their relevance to his objections to nudges.
While the global interest in health is true, he argued, it is not our only interest, is often in balance with other interests, and many people have different conception of what sort of health they want promoted. Unless the healthcare provider knows the patient’s interests, he cannot know how best to nudge. In the end, he proposed that because health decisions are more personal, our autonomy must be guarded more in that context and we should be even more critical about what sort of nudges we allow.
Finally, Andrea Freeman (AF), Assistant Professor of Law, University of Hawaii at Manoa William S. Richardson School of Law, spoke about the limits of nudging in relation to food policy and equality. She highlighted that some institutions have conflicting interests, and that sometimes these interests are balanced in a way that disproportionately takes a health toll on the already worse-off in health. She pointed out as an example that the USDA is responsible both for making healthy food guidelines AND for deciding which agriculture to subsidize and thus which industry to promote. She posited that in such conflicts, institutions will prioritize easily measured and incentivized outcomes, which will often promote sales over health outcomes. When the USDA had to decide what to do with a dairy surplus, she noted, they founded a company to create and promote products that contained LOTS of cheese. Instead of such a direct role in promoting health, the USDA merely funds research on nudging and promotes product labeling. The problem is that nudges work less well in regards to food preferences than elsewhere. She cited research that indicated that labeling only benefits the already health-obsessed, which tend to be already better off; many people will choose the low cal sandwich but then add an unhealthy desert and the labeling actually leads to overall less healthy choices.
She concluded by saying that the true causes of poor health are not individuals’ choices but food availability and cost structures stacked against healthy options. The USDA should shift its priorities in favor of health over industry: Make healthy food universally available, take fast food out of schools, and re-evaluate subsidies according to health. While research on nudges is good, it should not come before these structural changes.
The Q&A included questions on a) what alternatives are there to nudges if we do not like them b) whether we can ever truly avoid nudging c) whether healthcare is indeed special because the stakes of errors are higher or because we have a positive right for the government to provide it. While responses varied, the consensus was that we can control how we nudge and what intentions are behind the nudges.
At one point in the Q & A, Professor Cohen injected a bit of humour through making a failed attempt to promote MW’s recent book by trying to project its amazon page onto the main screen. I will help him attain delayed success by putting the link to the book here.