There were a bunch of excellent comments on yesterday’s food post. I thought I could answer them by adding a bit of clarity here.
First off to answer this tweet:
@TVHE I love it when you call me names. Particularly ones I don't understand.
— Geoff Simmons (@geoffsimmonz) September 2, 2014
My comments were certainly not intended to be an insult – “extreme social constructivist” was equivalent to call of “perfect information” which both Gareth and Geoff throw out there fairly often (including at me). I used this based on the near constant focus on how firms are “tricking” us, ergo shifting our preferences. Given the theoretical construct being used to justify what was written is that firms are doing this I’m happy with the description I gave.
Potentially it was seen as insult as these issues are often painted as “positivists vs constructivists”. Many in public health would see themselves as strongly, strongly, positivist – relying on data to estimate and falsify like a true science. However, data alone can’t give us anything – we need a conception of theory. A theory that relies strongly on firms and institutions setting people’s preferences IS constructivist.
This is why the positivist vs constructivist debate can be a bit of a red herring – hence why I follow a “scientific realist” view of economics and problem definition. In this way, I don’t see constructivist as an insult – it is a type of framework that has value as long as we try to consider “many models”. My preference is for methodological individualistic models instead – but this journey is a bit beside the point.
Think about the kids
Now the people commenting all asked whether I believed in a “rational choice model for kids”. Of course I don’t, as I said:
Social constructivism does have a nugget of truth, and in the formative years of childhood there are important elements too it. However, there are two key concerns.
First let me state down where our frameworks meet. Advertising to and information for kids does matter. We build up a stock of habits, rules of thumb, and physical attributes as children which have a fundamental impact on the type of life available to us as adults – and we are ill equipped to make choices for ourselves. In that case we do need to think about these types of issues.
My concerns were about how we use the framework to analyse and think about choice – but like almost all economists I heavily appreciate these points around habit formation etc ESPECIALLY with regards to kids.
So I wasn’t saying that we shouldn’t consider these types of explanations – what was I saying then? I was saying this shouldn’t be treated as our only explanation – and if we accept a “preference shifting” argument as an important part explanation that is fine. But be aware it is a debate you need to be able to have, on reasonable terms (eg recognising alternative explanations and the difficulty of balancing evidence) – the shrill attack on kids toys at supermarkets comes off as extremism here!
Now I noticed in the comments that Geoff states they aren’t anti the toys – I’ve read the posts and media releases and there is a lot of harsh language in there, and a very anti-firm rhetoric that can only be explained through a social constructivist view of firm actions influencing individuals. I can only answer the words that are put in front of me, and those words seemed relatively anti the program.
The discussion on choice and our “explanation of stylized facts”
When it comes to building an explanation, there is in fact a trade-off between an explanation from social constructivism and one from heterogeneity of choice.
If we see choices that differ from what we would choose we can explain these choices either with regard to different models. The same “variability” can be explained by people wanting different things – the question then is, is this due to individuals inherently valuing things in a different way to us, or is it due to them being “tricked” into valuing things differently.
The answer to this question is a central point of contention, and my goal was to bring it out. Especially given that each form of explanation has specific shortcomings – with constructivist models subject to multiple equilibrium with varying forms of value that may not be fully transparent.
By focusing too much on only a single explanation, we are inherently placing less weight on other explanations. A key set of other explanations in this case involve “people being different”. If we aren’t careful, such a bare focus will be implemented in policy and will restrict choices.
This is no small point, not in the slightest. The fact that most of our actions are unconscious, that habit formation is key, and that even our costly attempts at conscious choice are bounded does not offer a slam dunk argument for the “firms are controlling your preferences” argument being thrown around.
Weak wills and time inconsistency
There is a feeling that time inconsistency in itself justifies action. Of course it may – however, once again we are discussing a shifting “preference” as our explanation, implying that there are multiple explanations of our phenomenon with different policy recommendations depending on the one you pick!
Time inconsistency is especially intuitive though – which when it comes to “picking a model” this is sometimes a pretty important factor. At a minimum it makes it persuasive.
But, if we pick time inconsistency we need to ask “why don’t people find costly ways to make themselves commit”. Or if we go to use policy “how can we make it cheaper and more accessible to precommit?”.
This is an especially good concept as people can “self-select”. If they are suffering from precommitment issues, then making it easier to precommit improves their lives. If they are simply willing to trade-off future health more heavily than we are, then they won’t precommit and they’ll make choices.
Now we may say here “people are stupid, they’ll make choices outside their own interest”. I’d ask us not to use the term stupid, and to again define and try to measure the ways people mess things up – the concept of help here is about working with people to improve their lives (as we can’t observe their preferences), not so much to decide what we think their preferences SHOULD be
These issues are relatively simple/obvious
There is no such thing as a simple or obvious issue in social sciences. It is incredibly common for social scientists to call results obvious, trivial, or simple but it is a narrative I’ll fight till my last breath.
Social issues are far too important, and the inherent differences and depth of individuals too much, for any social “problem”/”issue” to deserve any of these terms.