Can a pill stop you drinking? Why short-term nudges won’t change behaviour
THE NATIONAL Institute for Health and Care Excellence (Nice) has been the butt of much ridicule over the past week. A pill designed to reduce alcohol consumption among problem drinkers is set to be made available across the NHS. But the concept of problem drinkers is so wide that it embraces people who enjoy a couple of modest glasses of wine a day. Indeed, the treatment is not really aimed at truly serious alcoholics, who are more likely to knock back vodka by the litre.
There are now vast swathes of behaviour which Western authorities attempt to modify. The government has a so-called “nudge” unit dedicated precisely to this end. Obesity, smoking, the amount of exercise people take, voting registration, recycling, and energy consumption are all examples of areas where policymakers are keen to see change. And on the latter, it is not just the amount of energy we consume but the mix that governments care about. Hectored for years that diesel fuel was morally superior to petrol, some unfortunates followed the advice and switched their cars to diesel. They now find themselves on the receiving end of a volte face by the bureaucracy.
There is a literature in top-ranking economics journals on the impact of such interventions. In general, there is a short-term effect which gives policymakers what they want. But, gradually, reactions become muted and people revert to their old patterns. There are exceptions, but most attempts to change behaviour fail.
An interesting paper in the latest American Economic Review by Hunt Allcott and Todd Rogers shows the enormous efforts which are needed to alter decisions that people make in the long term. In the US, nearly 100 utility companies hire a firm called Opower to send monthly home energy reports to millions of households. These reports include information on personal energy use, social comparisons and energy efficiency.
The real interest in the Opower work is that some of its programmes were set up as controlled scientific experiments. Allcott and Rogers examine three of the longest running, which started in the late 2000s. Highly sophisticated metering devices were installed. Households, from a very large sample, were selected at random to receive the information. And after two years, some of those getting the reports were randomly assigned to have them stopped. This way, both post-intervention persistence and the incremental effects of continued treatment can be measured.
Unsurprisingly, after receipt of the report, there is an immediate reduction in energy consumption, though this impact decays rapidly. In households discontinued after two years, the subsequent decline is much lower. The sheer frequency of the reports does seem to alter behaviour. There are further reductions in energy consumption in households that continue to receive the information, suggesting that people take a very long time to completely change their habits.
In the UK, attitudes towards wearing seat belts and drink driving did eventually change, but it took a very long time. Short-term trendy campaigns to “nudge” behaviour are just not going to work. Governments have to be in it for the long haul.
Paul Ormerod is an economist at Volterra Partners, a visiting professor at the UCL Centre for Decision Making Uncertainty, and author of Positive Linking: How Networks Can Revolutionise the World.