Just a quick clarifying post, following the previous PhD wiffle, where I picked on the transition movement to come up with some modelling questions. I want to say this now so I don't keep on repeating myself in later posts: it's not my intention to try and disprove the relevance of what the transition movement is doing. Whatever the merits of any assumptions its actions are based on, I have nothing but the highest respect for people working to reclaim some direct control over their own economic destinies. I am acutely aware that one of the historic roles of quantitative modelling (whether implicitly or otherwise) has been as a tool to justify robbing people of agency. This is especially true in its most pervasive form, finance-ministry-condoned economic methods: incredibly consistent across the world, and something no-one has a great deal of choice in since the vast majority of political parties do little more than tinker at the edges.

That use of models is, unsurprisingly, not something I have any desire to contribute to - but I don't think that should mean rejecting quantitative methods as a tool for helping steer our direction of travel. Economic self-determination is a good thing - I see no reason why quant methods shouldn't support it. We should have a future where quant planning tools work with the grain of democratic decision making and public action. There are reasons why, theoretically, quant tools have tended to go against that grain; again, that's a topic for later.

It's not an easy problem to solve. Here's an instructive example I hope to study in a bit more depth, off the back of this paper (a colleague of mine is a co-author). It aims to probe the idea of `smart cities'. The concept has, it seems, gained a lot of traction in US planning circles; David Roberts has argued strongly in support of the idea and his articles via that link give a good overview.

But the `takeaway for practice' from the JAPA article is:

Urban form policies can have important impacts on local environmental quality, economy, crowding and social equity, but their influence on energy consumption and land use is very modest: compact development should not automatically be associated with the preferred spatial growth strategy.

That's quite a modest set of conclusions: `compact development' is not necessarily a carbon and energy cure-all. The fundamental reason the paper finds this is that it actually adds some economics of land use to the problem. I need to get a special symbol for "I'll come back to this"... but the crucial part of this story has been the reaction: it seems to have caused a pretty intense ruckus among those with a deep commitment to the smart growth / smart cities idea.

Which leads me to wonder about the problems involved in linking quant/economic modellers and decision-makers - not just policymakers, but the kind of people working at a local level in the transition movement. It's easy enough to envision some perfectly healthy relationship between the two, but the reality is (and has always been, actually) pretty dysfunctional.

One solid reason for that: it's much harder to roll a boulder up a hill when someone's following you up questioning your rolling method the whole time. Social action benefits from having an agreed set of assumptions to work with. Despite these thoughts on the transition movement I am still, when it comes down to it, quite unsure about some of the fundamental assumptions that drive it. But like that climate cartoon ("what if its a great hoax and we create a better world for nothing?") dense economic webs made up of a froth of small-scale activity stand entirely on their own merits. Cop out? Hmm.

Update: but I reserve the right to change my mind about quant models. There's a small but definite risk that, on further investigation, it might become obvious that the democratic downsides outweigh any actual insight they might supply, and we should thus break the fingers of all coders attempting to model society. Harsh but fair, and I'll offer my own fingers up for the hammer first.