“Irrationality” is not a sufficient condition for intervention

One thing that irritates the hell out of me is the fact that many people believe the slightest sign of cognitive limitations implies that we should have government intervention.

Hell I’m centrist, I see myself as left-leaning in somethings and right-leaning in others.  But this attitude leads to ridiculous, poorly thought out, interventions.

Several things we need to think about before intervening are.

  1. What is the problem?  Compared to what other practical outcome?
  2. We have to say “why the problem is occurring” this occurs before we can figure out policy,
  3. We have to figure out if government can practically implement a scheme, given their own cognitive and institutional constraints,
  4. We need to ask if the implied problems are observable,
  5. We need to figure out what the consequences of such a policy are – how do people respond over time.

The most common area where this occurs is in the realm of saving.  More people than I care to count have told me that people are irrational, so save too little, so some form of compulsion in savings makes sense.  Some of the most intelligent, articulate, and thoughtful people I know have been known to occasionally move into this school of though.  However, lets see what needs to be ticked off for this to make any sense:

  • We need to figure out why people are saving too little relative to how much they do save.
  • We need to make a moral judgment regarding the impact on themselves (if they save too little and end up poor it is really their own problem – why do we need to intervene?)
  • If we don’t have that, we need to figure out how it impacts on “other people” – and likely make a value judgment for why this is suboptimal.
  • We need to figure out how well government can observe this issue.
  • We need to make a policy based on this observability, how well we can target “problem savers”, and the cost of implementation.
  • We need to look at the long-run implication of the choice – for example, if we make people “save” they could likely borrow on the basis of this asset … leading to no change in net savings rates.
  • We need to look at the distributional and efficiency impact of this policy, and the error bands around it – given uncertainty there would need to be a significant net positive impact for intervention.

I genuinely believe that people understate the intelligence of other people in society – there is no other explanation for how often I am told “people are stupid we need to make them save”.  It may surprise many people to find out that other people in society do think, and do make choices.  Attacking a “straw man” version of people and mentioning “irrationality” does not justify regulation.

Often I will say “it doesn’t matter if people are irrational, how is that a justification for compulsory savings” – too which people tell me I don’t understand.  I guess the failure to actually provide a logical argument does make it hard for me to understand 😉

8 replies
  1. Joe
    Joe says:

    Can I paraphrase your post and edit your final quotation to say “it DOES matter if people are irrational, it’s just that there are other things that matter too” – see your list of questions.

    Intervening because of ‘irrationality’ is and will always be paternalist in some way shape or form. It requires the government to advance the position that it knows better than its citizens. So the sorts of questions you pose about savings aren’t, I would say, fundamentally different than those that might be asked when the government passes usury laws or bans party pills or poligamy or anything else.

    I think the idea that individuals can look after themselves better than the government can is a good starting point – and note that’s different from assuming individuals are rational (cos they’re not). But that’s a rebuttable presumption.

    I’m not sufficently familiar with the issue of savings to make a call as to whether it meets this test. But the other thing I would say is that government intervention need not necessarily amount to compulsion. Some might say that the opt-out Kiwisaver arrangements are basically cut from the same cloth, but you can envisage interventions even lighter – compulsory financial education of some kind, a mandatory choice (rather than opt out) when signing for a new job (like having to fill in your tax code)…

  2. rauparaha
    rauparaha says:

    I think it depends what you mean by ‘irrationality’. Certainly it is insufficient to point to a cognitive limitation or an information cost to suggest that people are irrational. However, I think the irrationality that is often being referenced as a justification for intervention is different: it is a situation in which a person systematically fails to take an action that would maximise their utility. If the government can help them to take action at a very small cost then I think there may be a prima facie case for intervention.

  3. Matt Nolan
    Matt Nolan says:


    Indeed. The main difference with nudging and direct regulation is that at least you have verifiable choice – which implies that we have some additional idea regarding whether the policy is good or not.


    I agree, systematic biases often require an institution to improve outcomes – such as government. And if that case is made, that is nice.

    My issue here is when people just say “irrationality” and then sell their pet intervention 😉

  4. Joe
    Joe says:

    I certainly see where you’re coming from Matt. there is a risk that ‘irrationality’ will become the new market failure’ – that label describing a ‘problem’ that is generically slapped on anything a policy analyst, or as likely, a politician, wants to meddle with.

    I think @rauparaha has done a good job of indicating why the baby shouldn’t be thrown out with the bath water there.

    On another note, as a first time commenter I was suprised that there was such a lag while moderation took place. Can you advance an economic arguement to support that? I would have thought the marketplace of ideas would be sufficent to sort things out.

  5. Matt Nolan
    Matt Nolan says:


    “On another note, as a first time commenter I was surprised that there was such a lag while moderation took place”

    Fair enough. Moderation is done manually, so times vary strongly. There are many many spam comments to sort through – and the spam filter appears to block comments almost at random.

    I have also been away from the blog quite a bit – I prefer to reply to comments fairly regularly but haven’t had the opportunity over the past couple of months (between being away and my internet at home not working).

  6. Eric Crampton
    Eric Crampton says:

    @rauparaha: You still need that, expectationally, policy makers are less prone to irrationality than are individuals. And voters are the ultimate monitors of whether policy makers are getting things right. And voters are irrational when acting as voters.

    Irrationality all the way down, and worse than just leaving folks be, I’d say…..

  7. bLack
    bLack says:

    I agree, systematic biases often require an institution to improve outcomes – such as government. And if that case is made, that is nice.

Comments are closed.