The coalition government has talked up the benefits of behavioural economics. Early on it promised to shun “the bureaucratic levers of the past” and help “people to make better choices”. More recently a Downing Street “nudge unit” has been set up to examine how better information can change behaviour, and the economist Richard Thaler (who co-wrote the book Nudge) flew in last week to launch it. But, as yet, it is hard to see much evidence of these ideas being applied in practice.

This is a shame. Behavioural economics offers a more plausible account of how people and companies behave than the old models that pervade economic theory. As the Nobel prize winner Daniel Kahneman observed: “Humans reason poorly and act intuitively.” In theory, “nudging” should help governments change the behaviour of people and companies. Cruder changes, such as altering prices, should matter less.

But this is not what this government is doing in practice. Take plans to help workless families, for example. Here the coalition has focused almost entirely on changing behaviour by cutting or changing benefits. Or take binge drinking, an area seemingly perfect for nudge-type solutions, but where debate has concentrated on increasing the price of alcohol.

The government needs to show far more imagination. In particular, it must recognise how social networks can unlock the power of behavioural policy. Both mainstream and behavioural economics mistakenly assume that we act as autonomous individuals. But we do not: we are hugely influenced by the actions of others. The implications of this insight can be dramatic, with small initial changes delivering large final outcomes.

The financial crisis is the perfect example. The first sign of impending doom was a rapid loss of confidence among banks that their counterparties could repay their loans, even though economic fundamentals had not changed. Politicians and regulators then gazed into the abyss of the cascade of failures that could easily have followed the collapse of Lehman Brothers two years ago. Only now have figures such as Charlie Bean, the Bank of England deputy governor, begun to recognise that old approaches based on markets and “optimal” pricing of risk are not the way to deal with similar crises. Instead, approaches that look at the networks that connect financial institutions are needed.

Many other social phenomena can be explained by people simply copying the behaviour of others. Look again at problem drinking – 54 per cent of those who binge have friends who do likewise, compared with only 17 per cent for those who do not. This is essentially a fashion-related phenomenon, almost entirely unrelated to conventional incentives, such as the plan by the Scottish government to introduce a minimum price for alcohol.

Other problems facing the coalition could benefit from this “nudge plus networks” approach. Ian Duncan Smith, welfare minister, is grappling with the problem of generations in the same family who have never worked – a cultural rather than a purely economic problem. But it is the family and friendship networks in which many of the long-term workless are embedded that make it hard to shift their behaviour by conventional incentives alone.

At the opposite end of the spectrum, the huge discrepancies that have opened up between executive pay – and not just in the banking sector – and the rest of society reflect the spread of a set of values across the networks of this particular group. They believe they deserve these vast payments. A 50 per cent tax rate is not going to change this – indeed it may even reinforce an attitude that the group is somehow special. Instead we need to experiment with a whole series of nudges, to find one that spreads across this network, and makes it again socially unacceptable to be paid vulgarly large amounts.

This approach is disconcerting for policymakers raised in the command and control era. But a series of experiments of small nudges, harnessed to the power of networks genuinely to change behaviour, offers a potentially much more effective way of tackling seemingly intractable problems. The coalition is halfway there. But it now needs to push the power of networks to realise its vision of a Big Society.

The writer is author of ‘Why Most Things Fail, and also N Squared: Public Policy and the Power of Networks’, published by the RSA

Copyright The Financial Times Limited 2018. All rights reserved.