The Wallis Report, the Cooper Review, and the limitations of Behavioral Economics

In a recent special session on Superannuation at the Australian Conference of Economists, David Gruen — an economist working for the Australian Treasury and one of the authors of the Cooper Review (see here) reflected on the intellectual paradigm shift that informed its deliberations and recommendations (see here). 
According to Gruen, and his co-author Tim Wong, the Wallis report was built on the fiction commonly referred to as homo economicus, that mythical animal that is able to make even complicated decisions optimally, or at least better than government. According to Gruen, the Cooper Review reflects the insight that for many financial products, consumers lack (and cannot efficiently obtain) the knowledge to make informed decisions and hence some good old-fashioned paternalism is justifiable: “Perhaps one way of understanding the differences between Wallis and Cooper reports is that, rather than treating these cases as ‘exceptions to the rule’, we in the Cooper Review considered them to be more widespread – indeed of central relevance when it comes to decisions about retirement savings.” (pp. 6 – 7)

Arguing that a substantial body of work has emerged in recent decades in the field of Behavioral Economics, Gruen then sells unabashedly that field’s alleged insights as the foundation on which to build the choice architecture of the Australian superannuation scheme.

The result is known: it is likely — if indeed the Cooper Review recommendations will be implemented in its essence, which seems likely at this point — that a mandated one-size-fits-all default option (“MySuper”) will be imposed on superannuation members. Gruen, clearly, see this as progress. After all, superannuation members still have the choice to opt out. Or so he argues.

The argument is problematic for at least two reasons.

First, the paternalism of the MySuper default might indeed save poorly informed consumers from themselves. Unfortunately, the existence of such a default option is also likely to result in consumers being poorly informed and disengaged with their supers in the same way as insurance without deductions leads people to be less careful about the things they have insured. Gruen acknowledges some version of this argument in a footnote on p. 16 but dismisses it in passing in a somewhat snotty manner. Well, like it or not, there is an endogeneity problem here. And to dismiss it out of hand is wrong-headed and troubling. Especially if it comes from as influential a person as Gruen. Recent evidence has shown that even optimal defaults may not be optimal and that, in particular when consumers are fairly heterogeneous, requiring individuals to make explicit choices for themselves may dominate optimal defaults (e.g., Carroll et al, Optimal Decisions and Active Decisions, Quarterly Journal of Economics 2009, pp. 1639 – 1674). The result by Carroll et al. is hardly surprising in light of what we know empirically about the — sometimes quite dramatic — effects of financial literacy and peer effects.

Second, it has become bad habit of highly visible policy and/or opinion makers to appeal to alleged insights from Behavioral Economics. None of the people I have in mind here has ever actually done an experiment and it is clear from their uncritical sampling of the evidence that they do not know the relevant literature or the controversies over the production of the — mostly – laboratory evidence that undergirds much of Behavioral Economics. The simple fact is that pretty much every cognitive bias that psychologists, and behavioral economists, have allegedly identified is contested in the relevant literature. (There’d be too many references to list here but feel free to ask me for the syllabus of my Behavioral Course.) Building the superannuation choice architecture on disputed evidence is problematic at best.

To summarize: Has there been an intellectual paradigm shift? Yes, probably. Is that a good thing? Maybe. People often are not particularly good at financial decision-making (and regulators and/or employers do not help them much). Are unreflected translations into policy decisions of alleged insights from Behavioral Economics the way to go? I doubt it. While Behavioral Economics does have some interesting things to say, much of the evidence it allegedly provides is contested. Plus the obsession with one-size-fits-all default options is likely to crowd out strategies that entice people to become engaged with their supers as well as the incentive-compatible design of disclosure and delivery systems that could actuallu be understood and used. The design and implementation of such systems better rely on standard economic theory about information asymmetries and mechnism design. It is surely reasonable to assume that homo economicus rules superbly on the supply side.

10 thoughts on “The Wallis Report, the Cooper Review, and the limitations of Behavioral Economics”

  1. Where are these super funds encouraging members to become actively involved in their savings plans? I’d like to see that. At least we can say that Mysuper will provide greater choice in the market, and not rip off the disengaged so much as current default options. Other than that, agree entirely.

    Like

  2. Taken from Bryan Caplan at http://www.econlog.econlib.org

    P.S. I wish this point were original to me, but it’s not. I first heard it in high school while reading The Ayn Rand Lexicon. Surprisingly successful Randian psychologist Edwin Locke explains it elegantly:

    Behaviorism’s substitute for the mind is certain entities in the environment called “reinforcers.” A “reinforcer,” say the Behaviorists, is an event which follows a response and makes subsequent responses of the same type more likely. “What type of events change the probability of responding?” we ask. “Reinforcing events,” we are told. “What is a reinforcing event?” we inquire. “One which modifies response probability,” they reply. “Why does a reinforcer reinforce?” we ask. “That’s not a relevant question,” they answer. . . . To understand why a “reinforcer” reinforces, Behaviorists would have to make reference to the individual’s mental contents and processes-i.e., they would have to abandon Behaviorism.

    Like

  3. Hi Andreas,

    welcome!

    Despite your doubts about the validity of David’s stated reason for the one-size-fits-all superannuation program, I am firmly in favour of it. Mainly, I expect the transaction costs to be much lower with a simplified system, partially because it will be easy to get competition when there is a homogenous product all insurers have to provide.

    Dont forget, David Gruen is at the Treasury, which worries first and foremost about costs.
    If policy really had to be decided by a full knowledge of all the highly fragmented results in the behavioural econ field, not much would ever get decided.

    Like

  4. While only a fool would suggest that homo economicus actually existed, it would be rash to conclude that central planners in Canberra should step in and make these decisons for people. Do bureaucrats have the information required to make good decisioins? And what incentive do they have to really care?

    Like

  5. “… it would be rash to conclude that central planners in Canberra should step in and make these decisons for people”

    But bene_diction, the whole of compulsory super is based on precisely that. If you don’t believe that Canberra bureaucrats are better placed than individuals to tell each of those individuals how they should distribute their income over their lifetime, then the whole case for compulsory super falls apart.

    Which is why I still contend that compulsory super has reduced the wellbeing of the Australian population.

    Like

  6. derrida derider,
    There is alternative justification (based on the OLG framework) for compulsory super based. Central planners might care about the social security burden placed on future generation under a PAYG system (i.e. non-compulsory super). A fully-funded system (i.e. compulsory super) avoids this issue. Incidentally, one of the disadvantage of a fully-funded system is that people may not make the optimal long-term saving decisions, hence the justification for further government intervention which is touched in the above post.

    There is additional political economy issue attached to social security. Under our political system, the current generation has a monopoly on policy making via elections (until the children grow up anyway). Of course if we were really selfish we could’ve voted against compulsory super a long time ago …

    Like

  7. What fun. I will respond to the comments summarily and in due course. Bring it on. Matthew, your email please?

    Like

Comments are closed.