Nudging taxpayer behaviour

Nudge theory has now been around for 10 years. The idea behind this theory is that policy doesn’t always have to require people to do one thing or another. Offering them gentle directional guidance, like putting healthy foods on eye level shelves in shops, can encourage people to make the ‘right’ choice when decisions loom, while still allowing people choices to opt for the ‘less healthy’ options if they so wish. While clearly not suited to all policy areas, tax is one area in which ‘nudge’ has found a home alongside other notable success stories in areas such as health policy, organ donation and charitable giving.

In the April 2017 issue of Tax Adviser, Diana Onu and I explored a selected history of academic psychologists’ work as it has been applied to individuals’ tax compliance behaviours. We highlighted how this work stretched back to the 1950s in exploring how to develop co-operative approaches to voluntary tax compliance to sit alongside the more long-standing deterrence approaches then more widely applied – with varying success.

A modern outworking of this research has been encapsulated in so-called ‘NUDGE’ theory (an acronym describing various approaches to gently motivating desired choices put forward by Richard Thaler and Cass Sunstein in their 2008 book entitled, ‘Nudge – improving decisions about health, wealth and happiness’ – work ultimately contributing to Thaler’s 2017 Nobel Memorial Prize in Economic Sciences for his wider work in this area). The key principle of this theory is, in essence, that sometimes people need a gentle push to do the right thing.

Along with wider insights from behavioural economists suggesting the need for a greater role for persuasion over regulation in policy setting, this theory found lots of support amongst those involved in various public compliance areas, including in the area of tax, in the UK Government – even to the extent of the book becoming required Conservative MPs’ reading that summer. This resulted in 2010 in the creation of the ‘Behavioural Insights Team’ (BIT) as part of the Cabinet Office. It was claimed at the time, this was the world’s first government institution dedicated to the application of behavioural science. While Nudge is only one part of a wider field of behavioural insights, so central was Nudge theory and principles to this Team’s perceived activity it even became nicknamed the ‘Nudge Unit’.

A review of the work of this Team since its creation is provided in an article by Simon James of the University of Exeter published in 2015 (in the International Journal of Applied Behavioural Economics, April–June 2015, Vol. 4 No. 2) in which he charts their not insignificant successes in the four years to the point of the drafting of that paper (and prior to this Team becoming a social purpose company in that year still partly owned by the Cabinet Office but now operating outside of Government).

James describes the various areas in which their work had significant impacts – particularly highlighting health policy, organ donation and charitable giving – but attributed less success to their impact in the area of finances and taxation. At one level, he argues, the tax system has always provided an element of nudging behaviours and made use of behavioural insights, even before we specifically referred to it as such – with various tax concessions and reliefs providing financial incentives to ‘do the right thing’ (e.g. save into a pension) over other options (e.g. spending all your income now) without providing specific penalties for not so doing if the taxpayer choses not to. This feature of being ‘easy and cheap to avoid’ is in fact one of the core principles of what Thaler and Sunstein deem a nudge to be.

A pure nudge, in the Thaler and Sunstein sense, does not inhibit freedom of choice. This does somewhat limit is use as a policy tool as it can only strictly be used where freedom to disregard the proposed better path is allowable (i.e. not strictly directly applicable therefore to tax compliance issues). There are many instances where policy interventions are proposed that carry a desire to be much stronger than this would allow – pension auto-enrolment perhaps being a recent such example where to opt out is no longer a costless option but where specific effort has to be undertaken to achieve that result, and where this effort has to be repeated (once every three years in this case). As such, auto-enrolment pension policy, as operated in the UK currently, is arguably not a nudge policy in the strictest sense.

However, where choice can be allowed (even if not considered desirable), nudge can have success over outright deterrence strategies. For example, an article in the Evening Standard in 2014 discussed a policy intervention in tax (using white envelopes instead of brown and handwriting addressees’ names to nudge more favourable compliance results out of taxpayers being contacted) produced a 10% increase in response rate and a 16% rise in people paying tax.

There is a more fundamental criticism of nudging – namely the extent to which any government can be trusted to know best for the multitude of decisions an individual should make across any number of things to do with their life. However, assuming that we have to accept that is an unavoidable aspect of living together in a society that is government led, a nudge principled policy does at least leave the recipient with a choice whereas other policy approaches may not do so. Thaler and Sunstein tackle this challenge head on referring to the need to provide what they call ‘libertarian paternalism’ in policy making – paternalism with liberty still preserved. As such, they would argue, this is the acceptable face of paternalism and as such should be a welcome addition to policy setting toolkits.

So where has nudge theory affected recent tax policy making? Rational economics perspectives (such as arguing that the presence of penalties will induce people to pay their full tax obligations) cannot alone explain why so many taxpayers are largely voluntarily compliant. There are clearly psychological (or other) factors at play in the compliance decision-making of individuals that drives some of the motivations for their choice of behaviour. Here is where ‘nudge’ or behavioural insight motivated policies can be effective in encouraging people towards making the ‘right’ choice – in this case to be more voluntarily compliant than a taxpayer may otherwise be. At the same time doing so in a cost effective way compared to equivalently achieving deterrence approaches that create resistance in at least some taxpayers.

So has this been proven in practice to work? A second piece by Simon James helpfully assists with this analysis (Taxation and Nudging – Chapter 17 in the Handbook of Behavioural Economics and Smart Decision Making, Morris Altman (ed), Edward Elgar 2017). He highlights that tax authorities have long included a range of factors and attitudinal differences in developing their compliance strategies and taxpayer responses – even before Nudge encapsulated these ideas in their neat acronym. James highlights examples in Australia and New Zealand from the early 2000s where this was expressly advertised as such and where different responses were provided by the tax authorities to different taxpayers on similar compliance issues where it was considered likely that their motivations were different and that such different responses would elicit the desired outcome. This might range from jumping directly to using the full force of the law on those deemed to have already decided to be non-compliant, to making it very easy to comply where it was deemed to be likely the taxpayer was willing to be compliant, with just a little persuading.

Another area in tax compliance under development in the UK, and specifically related to nudge approaches, can be found in the drive towards pre-populating tax return information (a system pioneered in Nordic countries in the 1990s). Where such information can be alternatively sourced than from the taxpayer directly, and then shared with the taxpayer at the point of their tax filing, this will induce compliance by default (by ‘default’ in fact being the ‘d’ approach in the acronym NUDGE). A taxpayer has to make a positive choice to alter such data they know the tax authority then must hold on them, and are much less likely to do so unless that data is proposed in error. They are ‘nudged’ into including data they might otherwise have left out, or that could provoke confrontation if revealed to be known about by the tax authority later once a tax return is filed without that information. Pre-population is now widely used across many other countries including Australia (from 2006), Austria, France, Italy (from 2015) and Spain (from 2003).

The use of norms to nudge behaviour is another area closely related to the use of these principles. Several examples can be found of tax authorities using different emails or letter messages to encourage compliant behaviours. The Behavioural Insights Team, working with the World Bank and Government of Costa Rica demonstrate one such case with a large scale randomised control test in Costa Rica for their 2014 filing return period. The trial involved supplying different mixes of email and letter based messaging and third-party information to induce higher rates of small company tax filing to occur – with impressive results (Brockmeyer, A, Hernandez, M, Kettle, S. and S. Smith (2016) ‘Casting the tax net wider’, World Bank Group Policy Research Working Paper 7850). The BIT had also carried out similar work for HMRC in 2011 and 2012 using ‘behaviourally-informed’ social norm messages to more than 200,000 late filers across those years (the most successful of which producing £4.9m in extra revenue across a 23 day trial in the first year and £9m in the second year – Hallsworth, M, List, J., Metcalfe, R and I. Vlaev (2017) The behaviouralist as tax collector: using natural field experiments to enhance tax compliance, Journal of Public Economics, 148, 14-31). BIT’s 2016-17 Update Report also details their largest yet intervention using social norms reaching almost 750,000 taxpayers in Mexico moving into the formal sector from previously operating in the informal sector. This work involved encouraging them to register their employees for social security where they were failing to do so traditionally, leaving their workers exposed to remaining outside the social security safety net.

A similar study by the French tax authority in the same year is highlighted in a report by the Joint Research Centre (the European Commission’s in-house science service) in their 2016 report, Behavioural Insights Applied to Policy (EUR 27726 EN). This involved nudges to use the internet for filing by those known to have online access but who weren’t using online filing. In this case social norms and loss aversion was used to try and create a conversion to online filing. The campaign led to a 10% increase in online use for tax filing.

It seems that Nudge therefore is likely to be here to stay as it moves into its second decade of life. Going beyond Nudge itself into the use of behavioural insights to inform policy choices is now well embedded in policy setting process. It is playing an effective role in adding to the policy arsenal of tax authorities to help them develop co-operative solutions to encouraging compliance and to sit alongside continuing use of deterrence solutions, which are themselves being used perhaps now in different ways and on different taxpayers than might have been the case in the past in part due to behavioural insights into their most effective use. The role of academics in helping develop these approaches has been important both to develop the theories initially, to lead the way in showing how they can be implemented, and in providing evidence of their effectiveness when applied.