A lot of government effort is focused on changing behaviour, whether it’s getting people to do more exercise, find work or to save for retirement. One of the key tools in the government’s arsenal(in large part thanks to The Behavioural Insights Team) is behavioural psychology – small, low-cost nudges that focuses on subtle changes in people’s environments. This is contrasted with traditional government levers for behaviour change, many of which are much more resource and cost-intensive ‘shoves’ e.g. increased regulation.
While BIT is often, for better or worse, referred to as the ‘Nudge Unit’, we deal in both slight prods and and larger shoves because both can be effective. However, we also know they often have different cost-benefit profiles: very crudely, nudges tend to be simple, cheap to implement but have small effects, whereas shoves tend to be complicated, expensive and have bigger potential effects while also carrying greater risks of failure.
A classic example of the latter is the Family Nurse Partnership, an intensive home visiting programme for new mothers, which seemed to have no effect on its key outcomes despite costing the taxpayer £4,270 per mother (and a total of just over £3m for the 719 women in the Building Blocks trial).. On the other hand, as Bob Putnam noted when speaking at BIT last week, the introduction of universal, free secondary education in the US accounted for most of America’s GDP growth over the 20th century – rendering it a huge success and worthwhile endeavour despite the initial costs.
On the nudge side, BIT tested the impact of sending behaviourally-informed text messages to adults who booked appointments to get career advice from the Education Development Trustand reduced the failure to attend from 28% to 21%, a pretty big bang for not a lot of buck. In contrast, a message that aimed to encourage volunteers to take a more active role in a West Java ‘Eco Village’ programme didn’t lead to either increased engagement or retention of volunteers.
In some cases there’s a clear reason to prefer a nudge to a shove or the other way round e.g. small budgets will favour cheap interventions. In other situations policy-makers have to make a choice. This poses an important question: do we have a rigorous and systematic way of predicting which choice is likely to be best? I think the answer is probably no, and that may be one of the next challenges policy wonks, behavioural scientists and economists need to turn their minds to.
