What behavioralism lacks, on its own, is a theory of politics and power.
For instance, Sunstein posits that exogenous crises help explain the emergence of legislation and the overcoming of bounded rationality problems. (But where do the crises come from?) Specifically, “outrage” or “availability” entrepreneurs stoke the cognitive fires and convince lawmakers to take action. This can lead to policy solutions that are tailored more or less well to correct probabilities. (But what motivates these entrepreneurs?) He notes that culture can also play a role in attitudes towards risk. Societies (and groups and individuals within society) can be individualist, hierarchist, egalitarian or solidaristic. The latter groups may be more likely to embrace costly climate change policies that will benefit others more than themselves, while strongly individualist cultures may be less likely if the benefits and risks accrue elsewhere. (Sunstein, 2009, at 57, 66) (But who shapes these cultures?)
Without a theory of how groups and cultures are formed and shaped, behavioralism cannot direct its sophisticated (even arguably prediction-capable) view of economic agents’ internal environment to a view of how agents’ perceptions are structured by the external environment, and how their actions play out in the external environment.
The omission of politics and the factors shaping preferences would seem to be congenial to some evolutionary-inspired behavioralists like Simon, who have encouraged economists to return to a study of history. (Simon, 1996, at 47) But other behavioralists have responded to the gap by seemingly plugging in their own subjective judgments about politics, or by trying to use administrative devices to “nudge” agents’ subjective, error-prone cognitive frameworks closer to the “true” probabilities. (Thaler and Sunstein, 2009)
If the widespread cognitive failure of citizens is a rationale for paternalism, this would seem to suggest a robust role for government. In fact, Sunstein and colleagues are skeptical of government as well – especially the more populist it is, i.e. pandering to (and therefore replicative of) the cognitively deficient citizenry. (Sunstein, 2009, at 205) Elsewhere, they worry not only about outright populist government, but even that “the very accountability of bureaucrats” could lead them to replicate cognitive biases. (Jolls and Sunstein, 2006, at 233)
They therefore embrace the very conservative public choice framework for analyzing governmental actor motivation. The medicine for these widespread cognitive failures and biases from within the state is to create a specialized CBA-focused agency that insulated from “populist pressures” and trained in behavioral economics.[i] (Jolls et al., 1998, at 1543-1544)
(The authors unsurprisingly replicate other aspects of neoclassical thought as well. For instance, social outcomes are superior when derived fundamentally from willingness of citizens to pay for certain benefits or avoidance of certain costs.[ii] Private preferences (“corrected” through analysis of behavioral shortcomings) are the correct foundation for policy, in at least most of the cases.[iii] Sunstein is willing to give private ordering a fairly comprehensive role in addressing the payment for massively complex policy endeavors like climate change, but apparently not in the identification of the underlying problem in the first place.)
But once the cognitive failures have been generalized so as to include every individual on the planet, how would it be possible to step outside of the matrix? These specialized behavioral units would themselves be subject to cognitive failures. Perhaps the dislike of populism also comes from some sort of hidden preference. Indeed, without some sort of theory of the external environment, it is difficult to say what style of governance should be preferred. Sunstein, for one, seems to naturalize the notion of cost-benefit analysis (CBA) as an economistic (and therefore, preferable) mode of decision making. (Sunstein, 2009) But Simon notes that CBA is – like balloting, bargaining and planning – only provides “administrative” [not market] answers and is simply another non-market coordinating function in complex societies where the “price equivalents” are assigned by someone or something. (Simon, 1996, at 31, 42)
Postscript on Behavioralism
Theoretical agnosticism and incompleteness does not stop the world. Just because Sunstein doesn’t have a theory of the drivers of the external environment structuring CBA does not mean that he does not know how to play in this space. For instance, while Sunstein headed up regulatory affairs for the Obama administration, he expertly inserted corporate lobbyists into the role of fleshing in the higher costs to climate change policies, while sidelining voices within and outside the administration that argued for the benefits of action. (Broder, 2011) (Tucker, 2012a)
Elsewhere, Sunstein lauds the Reagan administration’s cost-benefit, corporate-supported approach to reducing ozone-depleting chemicals through the Montreal Protocol, and praises the Obama administration’s enactment of auto tailpipe measures that were “widely supported by the automobile industry.” (Sunstein, 2012) But there is no inherent virtue in including corporate voices at the table, especially if they will tend to exaggerate the costs and minimize the true benefits of action. Moreover, as Sunstein himself knows, how and when corporations decide to support regulatory measures is determined by multiple complex situational factors, not all of which are exogenously determined. The “availability entrepreneurs” that Sunstein dislikes (but of which he himself is one) can manipulate the perceptions of both governmental and corporate actors, and change how they characterize the costs of change. (Tucker, 2012b)(Hacker, 2010)[iv]
Simon goes a little further towards developing a theory of the interaction between the external environment. Because of the difficulty of correctly assigning probabilities to uncertain events, humans rely on feedback loops that adjust to unforeseen events. Systems ideally use some form of feedforward, to also predict events. But feedforward can be destabilizing if actors are constantly and rationally attempting to outguess one another (mutual expectations).[v] Scenarios like the prisoners’ dilemma and other games, so confounding to the neoclassical economist through the creation of multiple unstable equilibria, can be much more stable in situations of bounded rationality and norms. Simon suggests that, when faced with uncertainty, it may be more preferable to simply imposes expectation through organization, rather than chancing “major events that will affect many parts of the organization in the same direction.” Humans’ ability to utilize markets and hierarchies to respond to local problems may explain their dominance. (Simon, 1996. at 37, 42-43)
While Sunstein primarily views cognitive failures as a rationale for depoliticizing policy processes, Simon appears to be much more optimistic about humans’ ability to evolve norms like docility and organization that enable humans to thrive – in spite of, and perhaps because of, politics.
[i] Sunstein actually attempted to operationalize these concepts once in the Obama administration.
[ii] Sunstein makes some allowance for the fact that the poor may be less able to pay (and that their views will thus be underrepresented in CBA), and that some critics have noted the difficulty (if not offensiveness) of assigning monetary values to human lives. But he views the alternative to CBA as a form of precautionary principle, which he characterizes as “paralyzing.” He responds, in neoclassical fashion, by discounting “rights-based” claims, and saying that we should subsidize poor people’s income so as to correct the CBA (Sunstein, 2009, at 204-206, 218, 226)
[iii] Sunstein also accepts a view of regulatory interventions that they are inefficient will definitionally cost jobs, because firms pass on these costs to the consumer or employee. The only reason for regulation rather than general taxation is where there are some international public good or where there is some insurmountable distributional question (i.e. the beneficiaries of a regulation are not paying for its cost) (Sunstein, 2009, at 219, 231, 236)
[iv] Given Sunstein’s general favorability to incrementalism, paternalism, and corporate consultation, it is ironic that another agenda he has pushed – trade liberalization (Tucker, 2012c) – is tending to undermine all of the above. (Tucker, 2012a). (Tucker, 2012b)
[v] This poses some interesting corrollaries with Adam Smith’s Theory of Moral Sentiments, where “one’s individual merit is contingent and dependent on the happiness and merit of others” (Malloy, 1987, at 229), i.e. their mutual expectations of one another. Could morality in this context be an unstable mode?