Taking Opportunity Costs Seriously in Public Health Law

In a recent Fordham Law Review article, Rosenberg and Cohen made an interesting argument about popular policies to increase the number and distribution of retail supermarkets to eliminate “food deserts” (ie, neighborhoods without adequate access to food stores).¹ These policies, which are in the form of tax breaks and zoning provisions, are promoted as ways to enhance equitable access to healthful food and reduce rates of obesity. Contrasting the popularity of this policy approach with the “overwhelming evidence that they fail to improve health or dietary outcomes,” the authors lament the possibility that the attention and resources invested in proximate access to healthful food options (ie, building supermarkets) is at the cost of alternative upstream interventions, such as raising the minimum wage, strengthening labor protection, improving Supplemental Nutrition Assistance Program benefits, and protecting and expanding free and reduced-price school lunch programs.

It should come as no surprise that a widely praised health policy initiative is actually a naked emperor. Similar to other policy interventions, the policy of improving access to healthful food by supporting new food stores started as an idea in one place (the authors trace it back to Tony Blair’s Britain) and spread more or less by happenstance as new people adapted the idea to address the problems and priorities they perceived in other places. Many such innovative legal interventions turn out to be good ones; however, all things being equal, some innovative legal interventions turn out to be duds. Regardless of whether or not Rosenberg and Cohen are right about the failure of food desert policies or their impact on their preferred alternative policies, the larger point is a good one: adopting any particular law reform campaign can have substantial opportunity costs.

A basic concept in economics, opportunity cost refers to the benefits of other options that are lost when one option is selected. It thus “expresses the basic relationship between scarcity and choice.”² Public health generally, and in any particular place, has limited resources: limited time, limited funds, and limited political capital. Ideally, resources for developing, enacting, and deploying law reforms are used where the bang for the buck is the greatest. Squandering resources on a policy proposal that has more fashion than fact behind it is a waste. Along with the pure opportunity cost, one could well dilute one of the most important advantages of public health in the policy realm: a reputation for following the science and getting things right. Particularly when laws and regulations that raise costs, limit choices, or reduce some industry’s profits, being right—and having the evidence and arguments to convince others that you are right—can make the difference between better population health and (well-deserved) public skepticism and political defeat. And being right is, unlike many other determinants of policy, one over which we have direct influence.

Concern about opportunity costs does not require that we should pursue only policies that have a strong evidence base and every other indicator of being a sure winner. That is a basic chicken-and-egg problem: a new legal approach will not be “evidence based” until it is no longer an innovation—when it has been around long enough in enough places to have been thoroughly evaluated. Innovation is always a guess; a sense of opportunity costs dictates only that the guess be as educated as we can make it. The better we understand how mistakes happen, and the better we use our core tools of public health science to avoid mistakes, the better educated our guesses will be—and the easier it will be to minimize opportunity costs.

A policy innovation might be a dud by virtue of simple failure to achieve its goals. More subtly, we might think of it as a dud if the opportunity cost includes not mobilizing the local health community around an alternative that would have been more effective. The article by Rosenberg and Cohen got me thinking about the risk factors for mistakes in policy innovation. It might be a good idea for more people to weigh in on the list, but what follows is a start. How do dud ideas get started and spread? Some of the risk factors include the following:

• The basic cognitive biases that impair everyone’s assessments, predictions, and choices: we choose from ideas that are available and favor those that fit our existing beliefs and reinforce our group identity. Once we feel good about an idea, we are prone to unjustified optimism nurtured by confirmation bias as we consider evidence and alternatives.³

• The wrong people in the room: diversity of viewpoints, expertise, and experience is a primary means of reducing bias and increasing available information.⁴

• The allure of chalking up a legislative win: achieving a win comes with the attendant risks of framing “problems” or policy “solutions” in ways that promise political success⁵ but do not necessarily correspond to epidemiological or social evidence of the phenomena at issue. (This is part of the claim about food desert policies.¹)

• Political pressure to act on an issue at a particular time or in a particular way: some argue that this pressure led to unnecessary state and federal quarantine provisions during the Ebola scare.^6,7

• The uncritical application of methods that were once believed to be effective to a new problem for which they are inapposite: one example is reviving mandatory premarital testing to prevent HIV infection.⁸

• The grapevine: word of policy innovations spreads unsystematically, and as word spreads, repetition may become validation. If we hear about something new and exciting, we may assume it is working. One might see Health in All Policies as an example of this grapevine effect.

Some important practical qualifications are in order. Every choice has opportunity costs, and choosers will inevitably make mistakes. In real life, opportunity costs may be, at least in part, theoretical only. For example, resources may be locked in budgetary silos, so that if a grant is restricted to emergency preparedness, a health officer can try to use it as effectively as possible within that domain but does not have the option to spend the money on a clean indoor air effort that would have greater population impact. Politically experienced people may also point out the value of a legislative win to future policy, as exemplified by laws on concussions in youth sport that swept the nation during 2009-2013.⁹ Although the laws were confined to secondary prevention (ie, focused on stopping second traumatic injuries) and were even criticized as conceding too much to the political influence of the National Football League,¹⁰ these laws could well be responsible for some of the growing awareness of the issue of concussions among young athletes and the need for more substantial legal interventions.¹¹

Awareness of opportunity costs and the risk posed by our own lying brains is an important part of avoiding mistakes. Awareness should inspire caution. Then we can take maximum advantage of the tools provided by the science of public health for eliminating biases and testing predictions of impact. In a chapter we wrote for a health policy book, Evan Anderson and I canvassed some of the useful tools available, thanks to science, to check our biases and assess new policy ideas in the absence of direct evidence.¹² Although policy innovators usually lack substantial direct evidence about an innovative intervention, they usually have good data on the problem, and these data are often sufficient for critically assessing the fit between a problem and a proposed solution. Rosenberg and Cohen claim that policy entrepreneurs defined the problem of poor nutrition as one of proximity to good food rather than willingness or ability to pay, which, they say, is the real problem.¹ Even if a measure seems sensible, it may not be the optimal choice. The Haddon Matrix, a tool for organizing intervention options of various kinds at various entry points to reduce injuries, is a good way to overcome tunnel vision. Causal modeling (ie, using robust social science theory and behavioral or epidemiological evidence) can be used to expose and test assumptions about the problem being addressed and the mechanism(s) through which the policy is expected to have its effects.¹³ Drawing on various people with unique experiences of a phenomenon (eg, community members, lawyers, or researchers) brings in new information. Thinking about causal processes in the social environment also allows consideration of well-evidenced, analogous interventions: a tax on the consumption of sugar-sweetened beverages is a tax, and much evidence is available of how consumption taxes work. When we are conscious of opportunity costs and decide the benefits justify incurring them, we can call them “trade-offs” and retire in the evening with a clear conscience.

Innovation precedes evidence, but it does not have to precede it by much. Sometimes new policies have dramatic, immediate impact, as demonstrated by the effect of clean indoor air laws in reducing heart attacks.¹⁴ In general, more emphasis is needed on a systematic and strategic legal epidemiology approach to monitoring and evaluating the laws with which we treat millions of Americans at a time.^15,16 The National Institutes of Health, the Centers for Disease Control and Prevention, and legislators could substantially increase funding for legal evaluation, which at the moment constitutes a fraction of health research dollars.¹⁷ If a city or state is trying something new and promising, evaluators should gather preintervention data and study implementation as it unfolds. Early evidence of practical problems can improve the innovation and inform other potential early adopters. Early evidence of impact (or lack thereof) from one place can also improve the innovation and inform potential early adopters. As an idea is adopted elsewhere, it should be possible to have multisite, quasi-experimental evidence in a matter of years rather than decades.

In a world of finite resources and incomplete information, it pays to take opportunity costs seriously. We certainly cannot tackle all threats to life and health, nor can we expect that many threats will ever be eradicated. That means making choices with, and valuing, our scarce resources. The scarcer the resources get, the more the choices matter.