As we approach yet another crunch week for Brexit, it looks like Richard Thaler & Cass Sunstein’s “choice architecture” is about to take centre-stage. Those of us in jobs that come with some responsibility for determining HOW choices are made recognise this slippery concept, and its contingent dilemmas and pitfalls.

I’m hearing that MPs might be given a series of indicative votes on the various alternatives if the Prime Minister’s deal fails or doesn’t take place. But how will that happen? Let’s assume, however optimistically, that MPs will at least be able to agree on what those competing options are. Their next problem is that there is apparently no standard methodology for facilitating the choice, and so parliament will first need to decide how to decide!

One approach is a kind of musical chairs, where MPs vote in a series of rounds with the least popular option dropping out each time, until only one remains. But what happens if an MP’s preferred choice gets knocked-out early? Are they allowed to skulk off to a Westminster pub and thus avoid having to register preferences for alternatives that they despise? I can’t see many being thrilled if forced to vote in each round, and asking them to do so probably just tempts ever more perverse wrecking tactics.

Another method is for MPs to indicate their preferences via a one-off question. But will they be asked to indicate their top choice and a reserve; or their top three in rank order; or to rank all of the options; or some other variant? There are many ways of capturing preference in this way, each of them making a different appeal to logic. It is not really about which is best, the bigger point is that they would each produce different outcomes purely because of how they work. The way the options are framed – their choice architecture – will be fundamental to the outcome.

Questionnaire authors and other choice architects know that sometimes seemingly small differences in how we ask can bring about big differences in outcomes. As Thaler and Sunstein told us in “Nudge”, we simply can’t be neutral when presenting a choice. We inevitably bring some bias with us, whether generated by our own perspectives, or resulting from the mechanics of a question. Add to that the long-established psephological observation that non-binary problems cannot be solved by binary votes. If there is no majority in parliament for any single option then all a series of binary votes will achieve is to spell out that reality..

If that isn’t worrying enough, what if – rightly or wrongly – politicians end up throwing the decision back our way? What options would go on the ballot paper this time; how many options would we be given; how would they be phrased; and how much supporting information would accompany them? Anyone setting off to achieve fairness would feel obliged to present multiple options and enough background reading to allow an informed decision. But not only is that next-to-impossible in the highly charged Brexit atmosphere, as a designer of surveys I know that very often the more involved you get, the poorer the response you get back. People get tired and confused; and although we might declare otherwise, the reality is that we all tend to prefer choices that are couched in simple terms. Unfortunately that means that the issue at hand has to be summarised, which in turn can easily lead to misrepresentation.

And that, I think, was the problem in the first place!

Good luck everyone.