Chapter 8 of Escaping Paternalism provides a vigorous public choice critique of paternalism in general and the “new paternalism” in particular. While they deploy most of the standard public choice arguments, Rizzo and Whitman (RW) also ably turn the tools of behavioral economics against paternalism itself. Real-world paternalists really should proverbially cast the beam out of their own eyes before they decry the motes in the eyes of the populations they aspire to rule.
How can Rizzo and Whitman justify their appeal to behavioral findings after spreading so much skepticism about the underlying research?
Given that we have argued in previous chapters that many alleged biases may not be biases at all, that acting upon them might sometimes be rational, and that people have various means of controlling their biases when necessary, we need to justify why here – in the public policy context – we are seemingly changing our tune. What justifies our invocation of cognitive biases to explain government failure?
Our answer is twofold. First, we are to some degree offering an immanent critique: if and to the extent that people have genuine and significant cognitive biases, then we should expect those biases to affect public policy. Thus, any justification for paternalist policy that refers only to the biases of people in the private sector and not the public sector is incomplete at best. When genuine biases exist, they will operate in both arenas… Behavioral paternalists could respond by arguing that cognitive biases are not really that severe and therefore should not have deleterious effects on policymaking – but to do so, they would also have to abandon much of their case for paternalist intervention in the first place.
Second, the public sector is rife with spillover effects, also known as externalities. Government policies automatically affect thousands of people who did not choose them. People making political choices – including voters, legislators, regulators, and judges – rarely, if ever, bear the full benefits and costs of their actions.
[I]t matters whether the effects of alleged biases fall primarily on the decision-makers or on the population as a whole. For instance, suppose someone is affected by optimism bias that makes them willing to take greater risks. On an individual level, as we’ve indicated in earlier chapters, it’s not clear this constitutes irrationality; it could reflect the individual’s unusual preferences, or perhaps a positive mindset that is conducive to their overall life success despite greater exposure to risk. But on a social level, if that optimism results in policies that expose the whole body politic to greater risk, that fact is worrisome whether the optimism is rational (for the individual) or not.
After reviewing the public choice staples of rational ignorance, concentrated benefits/diffuse costs, and self-interested regulators, RW introduce readers to “bootlegger and Baptist” stories – then provide a slew of matching case studies of actually-existing paternalism.
1. Official obesity standards are largely based on interest-group maneuvering, not medical science.
The new, lower thresholds increased the number of Americans classified as overweight or obese by more than 35 million overnight (Kuczmarski and Flegal 2000). What motivated the WHO and NIH to adopt the new recommendations? We cannot say for sure, but we can venture an educated guess. The WHO report, which heavily influenced the NIH’s later decision to change the BMI thresholds, “was drafted and written under the auspices of the International Obesity Task Force (IOTF),” an organization “primarily funded by Hoffman-LaRoche (the maker of the weight-loss drug Xenical) and Abbott Laboratories (the maker of the weight-loss drug Meridia)” (Oliver 2006, 28–29). These companies stood to benefit from more people being qualified for coverage of their products by private and governmental health insurance.
2. The war on vaping is absurd:
While e-cigarettes may not be entirely free of risk, there is good reason to believe they constitute a far superior alternative to regular cigarettes from a health
perspective. A report from Public Health England, a division of the UK’s Department of Health, found that “[w]hile vaping may not be 100% safe, most of the chemicals causing smoking-related disease are absent and the chemicals which are present pose limited danger,” and concluded that e-cigarettes are “around 95% safer than smoking” (McNeill et al. 2015, 12).
Yet the absence of evidence has not stopped anti-tobacco advocates and anti-smoking organizations from pushing for vaping regulations and bans immediately; the mere suggestion of possible harms is sufficient. To take just one example, the website for the United States’ Action on Smoking & Health (ASH) presents a series of unknowns – not proven harms – related to e-cigarettes, and then concludes that “while the safety of e-cigs remains in doubt, they should be s [sic] treated like cigarettes” (Action on Smoking and Health n.d.). Other abstinence-focused organizations, such as the Campaign for Tobacco-Free Kids and the Truth Initiative, have also supported regulation of e-cigarettes in a manner similar to that of regular cigarettes.
In May 2015, a year after the US Food and Drug Administration proposed new regulations on e-cigarettes, a coalition of “31 health and medical groups including the American Academy of Pediatrics, the American Academy of Family Physicians and the American Heart Association” sent a letter to the president urging the federal government to finalize the regulations (Sifferlin 2015). Although the letter to the president emphasizes the health risks associated with e-cigarettes, no such risks are cited…
3. U.S. nutritional guidelines are driven by politics, not science:
Light describes some of the notable differences between the unpublished food pyramid and the final version. The number of recommended servings of fresh fruits and vegetables fell from 5–9 to 2–3, while the recommended servings of whole-grain breads and cereals rose from 3–4 to 6–11. (The recommendation for fruits and vegetables eventually rose to 5–7 through the efforts of an anti-cancer campaign by the National Cancer Institute.) Light further notes that white-flour baked goods, which the experts had placed at the pyramid’s peak for items to be eaten sparingly, had been moved to the pyramid’s base (Light 2004)…
…In recent years, many have begun to reevaluate the conventional wisdom that emphasizes caloric restriction and reduced fat intake, instead arguing for restriction of carbohydrates and greater consumption of protein and fat – a position most commonly associated with the names John Yudkin, Gary Taubes, and Robert Atkins (Leslie 2016). If these authors’ position is true, then it could be that the meat and dairy lobbies deserve less blame, and the grain and sugar lobbies more blame, for American obesity. We do not feel qualified to weigh in on this debate, except to observe that there exists plausible evidence in support of both sides… Our purpose here is to shine a light on the process by which government nutritional guidelines have been created, which stands in sharp contrast to the humble, cautious, and disciplined approach imagined by behavioral paternalists.
Then RW ponder the specific biases that taint paternalistic policy-making, including:
1. Action bias: like most people, paternalists yearn to “do something” but feel no strong urge to “do something that demonstrably works.”
2. Overconfidence: like most people, paternalists hastily conclude that they know better than the rest of mankind – and even more hastily conclude that they know how to help.
3. Confirmation bias: like most people, paternalists latch on to pieces of evidence that bolster their worldview, but gloss over counter-evidence.
[B]ehavioral paternalists do not merely wish to inject behavioral thinking into existing regulatory agencies. They have proposed the creation of whole new domains of regulation. In the United Kingdom, their work contributed to the creation of the Behavioural Insights Team, also known as the “nudge unit,” a government group (now partly privatized) whose raison d’être is to find public sector applications of behavioral research. In the United States, behavioral research provided the impetus for President Obama’s Social and Behavioral Sciences team. What are the odds that groups such as these would not find justification for their existence?
4. Availability bias: like most people, paternalists put too much weight on vivid examples.
5. Affect heuristic: like most people, paternalists put too much weight on their own impulsive emotional reactions.
6. Present bias: like most people, paternalists put too much weight on immediate gains relative to long-run costs.
On the benefit side, politicians who support them can expect to gain credit for taking action to solve pressing problems such as obesity and insufficient savings. When and if the chosen policies generate unexpected costs through unintended consequences and side effects, those costs will occur at a substantial delay – after the policies have been implemented and results measured. (And if results are never measured, so much the better.)
Chapter 9 delivers a full-throated defense of slippery-slope arguments, arguing that slippery-slopes prosper in the presence of continuity and vagueness. Consider the policy fruit of research on “present bias”:
[T]he target rate of time discount involved in any attempt to “help” agents with hyperbolic preferences is inherently vague. We do not know where “reasonable” impatience ends and “excessive” impatience begins. There is no sharp dividing line between them. There is little promise that future research will resolve these questions. This vagueness itself is sufficient to create a gradient. There is no clear line to resist the gradual creep of higher savings requirements, higher fat taxes, and the like.
The same goes for the literature on “risk narratives”:
For risk narratives to be effective, they have to be sufficiently frightening or visceral. And therein lies the problem: there is no objective line between “not frightening enough” and “too frightening.” Jolls and Sunstein admit that excessively frightening narratives could be counterproductive, inducing too little risk-taking, and their response is telling: “Of course there are line-drawing problems here, but the basic point is straightforward” (p. 214). Line-drawing problems are, of course, the telltale sign of a gradient. What is the right level of emotional response needed to qualify a decision as rational? As an empirical matter, there is simply no way to know whether customers who engage in a risky activity are doing so rationally – with a full understanding of the risks – or have simply not been exposed to a sufficiently scary narrative. The gradient goes from missing narrative to mildly compelling narrative to worst-case-scenario narrative.
Another factor that further greases the slope: state regulation crowds out self-regulation, leading to demands for more state regulation:
[I]ndividuals may treat external control and self-control as functional substitutes, which has important implications for paternalist policy-making. Suppose a policymaker decides to place a tax on a “bad” activity or a subsidy on a “good” activity. Insofar as the new policy is perceived as a form of external control, the evidence suggests that targeted agents will respond by decreasing their level of self-control… Then arguments will be made for increasing the subsidy or tax and expanding the degree of paternalistic intervention. The changing behavior of the target agents generates a new round of incentives for policymakers to intervene.
Idle worries? Hardly:
[W]e have already seen how the failure of soft interventions to generate expected results can lead the supposed experts to get behind more aggressive intervention. In a New York Times op-ed, George Loewenstein (a behavioral economist) and Peter Ubel (a physician and behavioral scientist) observe that mandatory calorie posting at restaurants has had little impact on eating choices – and immediately jump to the conclusion that policy needs to increase the relative price of unhealthy foods to have a real impact (Loewenstein and Ubel 2010). In their analysis, the justifiability of the policy goal of reducing overweightness and encouraging better health is simply taken as given. The remainder of the op-ed considers a number of other policy arenas in which, they suggest, the failure of a behavioral “nudge” provides grounds for more aggressive intervention.
RW go on to argue that experts’ aforementioned cognitive biases further grease the slippery slope. Ponder:
Action bias, in addition to encouraging initial paternalistic interventions, has the potential to encourage subsequent interventions as well. As long as the problem does not appear to have been solved by the initial intervention, policymakers will feel pressure to take further action.
Escaping Paternalism then rebuts recurring objections to slippery-slope arguments.
The idea that we should “make progress” on the initial interventions, and then do what we can to “pour sand” on the slope, is a variant of the usual (and, we think, hackneyed) response to all slippery-slope arguments: that we can simply “do the right thing now, and resist doing the wrong thing later.” But if the slope argument is correct, there is a causal (albeit probabilistic) connection between initial interventions and later ones. Saying we should move forward on those initial interventions is akin to saying we should do something because it promises present benefits while ignoring the potential costs in the future. Ironically, it is just this sort of error in private decision-making that behavioral paternalists think cries out for correction. The slope risk must be counted among the costs of the initial intervention.
1. Other critics of behavioral economics have pointed out that behavioral economists could fall prey to the very biases they study. However, no one – but no one – has fleshed out this doleful observation more convincingly than RW. Once you read Chapters 8 and 9, you will forevermore see the famed behavioral economists as the arrogant, domineering, conventional, and otherwise flawed human beings that they are. They think they’re fit to rule, but they’re not.
2. RW’s “bootlegger and Baptist” case studies are powerful – and the policy issues are too big to dismiss as availability bias bait.
3. The discussion of action bias is particularly good. Seriously, can you imagine a politician saying, “Yes, people are hurting themselves. But helping them would cost a lot of money, inconvenience many innocent bystanders, and probably wouldn’t do much good. So let’s do nothing”? Tragically, the voice of reason sounds bad.
4. Once again, I wish the book contained a detailed section on opioids. The usual view, of course, is that Big Pharma’s lobbying prevented wise regulation from paternalistically saving hundreds of thousands of American lives. This is clearly a public choice story, but not one in the spirit of Escaping Paternalism. So is the conventional story flat wrong – or what?
5. RW’s defense of slippery-slope arguments is the best I’ve seen. Focal points and bright lines are the best defense we have against slippery slopes; once you frame everything as a continuum, policy really does flow in predictably statist directions. (Note the parallel with hyperbolic discounting, where people make decisions they predictably regret. When there is a slippery slope, policy predictably becomes more draconian than policy-makers initially wanted… or at least officially initially wanted).
6. “The slope risk must be counted among the costs of the initial intervention.” Hardly anyone affirms it, yet who could deny it?
7. On reflection, Escaping Paternalism contains an implicit intellectual history of the new paternalism. The heart of the history: The new paternalists created a playground full of new slippery slopes which they predictably traversed, all the while denying the policy relevance of slippery slopes. If this seems unfair, dwell on the Loewenstein-Ubel discussion a bit longer. Or the war against vaping.
8. RW embrace glorious candor. If we fully embrace their “inclusive rationality,” RW freely admit that we can no longer impugn paternalists as irrational. I’d add, however, that RW expose their opponents’ irrationality so convincingly that we should reject inclusive rationality. Many beliefs are arrant folly, and relabeling this arrant folly as “inclusive rationality” is bending over backwards for no good reason.
9. RW appeal heavily to the dangers of irrationality externalities rather than irrationality per se. Hence, like me, they worry far more about political irrationality than personal irrationality. Yet to be fair, RW could use most of their arguments about knowledge problems – culminating in humanity’s lack of telepathy – to foster skepticism about our ability to detect, measure, or mitigate externalities. The lesson: A restrained appeal to knowledge problems is great, but if you push knowledge problems too far, you paint yourself into an agnostic corner.
10. The depressing thing about Escaping Paternalism is realizing that the book is too intellectually scrupulous to sway policy. Alcohol will probably remain legal because of (a) status quo bias, and (b) most people think alcohol is fun. Yet if alcohol were invented today, almost no government would allow it. Even ironclad appeals to knowledge problems and public choice concerns would be impotent in the face of paternalistic demogoguery. Alas.