10 April 2017

Governments Want Citizens to Transact Online: And This Is How to Nudge Them There

A randomized control trial that “nudged” users of a disability parking scheme to renew online showed a six percentage point increase in online renewals. Image: Wendell (Flickr).

In an era when most transactions occur online, it’s natural for public authorities to want the vast bulk of their contacts with citizens to occur through the Internet. But they also face a minority for whom paper and face-to-face interactions are still preferred or needed — leading to fears that efforts to move services online “by default” might reinforce or even encourage exclusion. Notwithstanding these fears, it might be possible to “nudge” citizens from long-held habits by making online submission advantageous and other routes of use more difficult.

Behavioural public policy has been strongly advocated in recent years as a low-cost means to shift citizen behaviour, and has been used to reform many standard administrative processes in government. But how can we design non-obtrusive nudges to make users shift channels without them losing access to services? In their new Policy & Internet article “Nudges That Promote Channel Shift: A Randomized Evaluation of Messages to Encourage Citizens to Renew Benefits Online” Peter John and Toby Blume design and report a randomized control trial that encouraged users of a disability parking scheme to renew online.

They found that by simplifying messages and adding incentives (i.e. signalling the collective benefit of moving online) users were encouraged to switch from paper to online channels by about six percentage points. As a result of the intervention and ongoing efforts by the Council, virtually all the parking scheme users now renew online.

The finding that it’s possible to appeal to citizens’ willingness to act for collective benefit is encouraging. The results also support the more general literature that shows that citizens’ use of online services is based on trust and confidence with public services and that interventions should go with the grain of citizen preferences and norms.

We caught up with Peter John to discuss his findings, and the role of behavioural public policy in government:

Ed.: Is it fair to say that the real innovation of behavioural public policy isn’t so much Government trying to nudge us into doing things in subtle, unremarked ways, but actually using experimental techniques to provide some empirical backing for the success of any interventions? i.e. like industry has done for years with things like A/B testing?

Peter: There is some truth in this, but the late 2000s was a time when policy-makers got more interested in the findings of the behavioural sciences and in redesigning initiatives to incorporate behaviour insights. Randomised controlled trials (RCTs) have become more commonly used more generally across governments to test for the impact of public policies — these are better than A/B testing as randomisation protocols are followed and clear reports are made of the methods used. A/B testing can be dodgy — or at least, it is done in secret and we don’t know how good the methods used are. There is much better reporting of government RCTs.

Ed.: The UK Government’s much-discussed “Nudge Unit” was part-privatised a few years ago, and the Government now has to pay for its services: was this a signal of the tremendous commercial value of behavioural economics (and all the money to be made if you sell off bits of the Civil Service), or Government not really knowing what to do with it?

Peter: I think the language of privatisation is not quite right to describe what happened. The unit was spun out of government, but government still owns a share, with Nesta owning the other large portion. It is not a profit-making enterprise, but a non-profit one — the freedom allows it to access funds from foundations and other funders. The Behavioural Insights Team is still very much a public service organisation, even if it has got much bigger since moving out of direct control by Government. Where there are public funds involved there is scrutiny through ministers and other funders — it matters that people know about the nudges, and can hold policy-makers to account.

Ed.: You say that “interventions should go with the grain of citizen preferences and norms” to be successful. Which I suppose is a sort-of built-in ethical safeguard. But do you know of any behavioural pushes that make us go against our norms, or that might raise genuine ethical concerns, or calls for oversight?

Peter: I think some of the shaming experiments done on voter turnout are on the margins of what is ethically acceptable. I agree that the natural pragmatism and caution of public agencies helps them agree relatively low key interventions.

Ed.: Finally — having spent some time studying and thinking about Government nudges .. have you ever noticed or suspected that you might have been subjected to one, as a normal citizen? I mean: how ubiquitous are they in our public environment?

Peter: Indeed — a lot of our tax letters are part of an experiment. But it’s hard to tell of course, as making nudges non-obtrusive is one of the key things. It shouldn’t be a problem that I am part of an experiment of the kind I might commission.

Read the full article: John, P. and Blume, T. (2017) Nudges That Promote Channel Shift: A Randomized Evaluation of Messages to Encourage Citizens to Renew Benefits Online. Policy & Internet. DOI: 10.1002/poi3.148.


Peter John was talking to blog editor David Sutcliffe.


Note: This article gives the views of the authors, and not the position of the Policy and Internet Blog, nor of the Oxford Internet Institute.