Offsetting recently posted about a tweet by Gareth Morgan on eating and control, including a reply I popped up. Essentially, Gareth’s tweet implied that the way individuals make choices indicates we have no choice over how much they eat. I disagreed talking about precommitment – he stated I assumed perfect information, which is both a touch untrue and (surprisingly to many) irrelevant.
@TVHE You are assuming perfect information
— Gareth Morgan (@garethmorgannz) February 23, 2014
It did get me thinking though. The two of us actually have almost exactly the same model of choice in our heads for this issue, and as a result any differences of view of on the appropriateness of policy that we might have are not due to differences in the underlying model. Recently I saw this paper via Neuroskeptic on twitter, and my reply to that also was:
@Neuro_Skeptic While I agree in large part, the author downplays commitment to future rules – dangerous when discussing addictive behaviour
— TVHE (@TVHE) March 4, 2014
Yes we can think of matters through a “conscious” and “unconscious” mind, where the unconscious mind bear similarities to a computer – and where we ex-post rationalise choices to satisfy ourselves. This model has empirical backing, is logically consistent, and suits my priors – so I’m comfortable with it. And it is this context that the article Gareth discussed, and his own view on the obesity epidemic, are based. In this way, yes we have the same model of fundamental choice.
Where we might differ is in terms of our belief that people can, will, and should precommit their actions and influence the rules followed by their unconscious mind.
I take a revealed preference approach to this. If people don’t precommit, it is because precommitment itself is costly and so the benefits from doing so are presumed to be too low. Their may be a failure in their expectations or information – in which case government can help. Government may be able to offer low cost commitment mechanisms, or even at a push use framing effects to “nudge” (note, these framing effects should not be directly costly – something that often gets forgotten). But that is that.
It may be possible to take a more paternalistic line, and say that there is a certain way people should be. I disagree with this, and I don’t know if this is actually Gareth view, and this is why I said this “might” be a difference – in may not be. However, if it isn’t a difference and we have the same view of choice, I am uncertain why he would have disagreed in the way he did with my tweet.
What does this have to do with the title of the post?
I’m getting there. One thing Gareth has been pushing is a sugar tax, a recent tweet about it was here:
Sugar tax may be necessary, England's chief medical officer says http://t.co/egTCcp6IPV
— Gareth Morgan (@garethmorgannz) March 5, 2014
Now I touched on some of the limitations of the tax idea in this post. However, the limitations I pointed out do not imply that the best tax itself will be zero – I leave myself more open here. They are just issues we should consider. Geoff Simmons touched on these things (he is writing with Gareth on these issues) in this tweet:
@TVHE Interesting. What do you think mechanism to justify smoking tax? I would say time inconsistency, health costs, plus productivity loss
— Geoff Simmons (@geoffsimmonz) February 14, 2014
I understand the argument for a tax based on time inconsistency issues and a version of “user pays” for health care – a linear tax when the cost is highly non-linear is a rough solution, but it is at least worth looking into.
However, if Gareth disagrees with my point on precommitment and does not think there is any “control”, this in part undermines the case for a tax.
If people are not making choices due to “substitution” (the change in the relative price of food) because they are programmed to just eat, the only way a tax changes behaviour is through the income effect – essentially making people poorer. So a tax program will exacerbate poverty. Furthermore, if people are not making choices, we don’t get a substitution towards a healthier diet.
The part of the externality associated with “user pays” is still relevant, independent of a change in behaviour. However, given how imprecise the tax is at actually targeting this ‘policy externality’ it might be a bit more appropriate to just target it directly – say by having user pays for obesity related disease. If we don’t want to do this, as we think society should pay, then there is actually no externality!
The part of the externality associated with “time inconsistency” is completely irrelevant without a change in behaviour. It is a transfer between “different versions of the self”, so if we don’t change behaviour it just nets out.
Note: Externalities from “loss of productivity” aren’t externalities. There is a market price for labour. Claiming this as an externality is becoming increasingly fashionable among economists, and it makes me very uncomfortable – it deserves a post of its own to be honest!
Choice issues are hard. Policy issues are hard. Both Gareth and Geoff know this, I’ve talked to them in the past. Furthermore, I think this sort of tweet indicates that they realise, and care about the fact that, these are delicate complicated issues:
The use of financial incentives for weight loss is promising, but tricky to get right. http://t.co/m82buQiVD7
— Geoff Simmons (@geoffsimmonz) March 5, 2014
However, I am uncertain why the rhetoric that sometimes comes out places little faith in individual choice – given we are using the same choice framework to understand the issues. Yes people make bad choices and regret them, yes most people don’t seem to make choices that we feel we would have made in their shoes, but fundamentally people are all incredibly different.
Most importantly though, if we are going to use a rhetoric about choice that takes away individual responsibility, we have to be careful selling a policy that demands changes in individual choice – this is a logical inconsistency. If we do so because the “natural” conclusions of our rhetoric, towards controlling choice, appear abhorrent, then perhaps the rhetoric we are using is downplaying the propensity of individual choice!
I genuinely like the fact that Gareth and Geoff are looking into these issues, and trying to describe the role of policy and choice. Furthermore, my comments are a call to do analysis – not to say we shouldn’t! I am just always cautious of the role of policy within an environment with individual agency, namely:
@garethmorgannz I will do, thanks! I just worry that focuses on nudges, without individual agency, can lead us a bit astray!
— TVHE (@TVHE) February 23, 2014