Tony Curzon Price (London, openDemocracy): Laura Sandys - chair of my board here at openDemocracy - sends out a newsletter as part of her campaign to take South Thanet at the next election. (see telllaura.org.uk). Today, the day her leader got concrete with his localist policies, she sent round a fascinating bit of sleuthing on the local effects of post office closures.
The post office was told by its masters in Whitehall to close 2,500 branches. How does a centralised organisation respond? By getting a few operations research nerds to crank the handles on the computer models. I've done enough of it myself to imagine exactly how this happens: for each branch, look at how much they are costing, and of the very expensive ones, try to figure out whether any have close substitutes. In this day of massive Geographic Information Systems, you can find neat little algorithms to define "closeness" and "accessibility". A few enjoyable weeks later, you have a gem of a model ,which, with a tweak here and a tweak there will give you some small number of variants of the list of 2,500 victims.
But what Laura found was that the models weren't so good. Unsurprisingly, the nerds working on GISs couldn't know that, for example, there were steep steps between this road and that post-office: and the absence of small, on the ground detail led to bad decisions. The story is an allegory for de-centralisation. Where there is and is not a post-office is a question for local politics, not for an operations research department in a national monolith. The monolith won't have the information it needs, and will substitute a political process of weighing of costs and benefits for an industrial information-processing alternative. Not only does it make bad decisions, it makes them in a way that gives localities no responsibility for making the necessary trade-offs. What I like about the new localism is that, once it has been let out of the bag, it will change many institutions (and their decision making processes) profoundly.
