Creating innovative change through new legislative pilots

The sandbox piloting approach being explored is defined in the consultation document as ‘a temporary environment where HMRC could conduct tests of new policy and processes which suspend, implement, ease and/or harmonise legal obligations. This could be for a distinct group of people or sector for a defined period of time and would be accompanied by appropriate safeguards and guidance.’ The outcomes would then be evaluated to inform future decisions on adopting the change permanently.

CIOT response

The CIOT supports in principle the idea of HMRC developing an innovative approach to the development of new tax policy and legislation and trialling new processes. However, we think the ‘sandbox’ piloting approach could give rise to problems, particularly if it was run as a parallel tax or penalty system, rather than being treated as a trial involving a small number of taxpayers. In our view, participation in a pilot should be voluntary, no-one should end up in a worse position because they participated in it, and any adverse tax, interest and penalty position should be promptly corrected by HMRC. Also, participants (and their advisers) should be compensated for any costs incurred as a result of them being in the pilot. It is essential that existing taxpayer safeguards are protected within the sandbox environment.

If the results of a pilot indicate to HMRC that legislative change is desirable in a particular area of the tax code, that change should be follow the usual consultation process, and then be considered by Parliament during the legislative process. Equally, we think sandbox testing could be undertaken as part of the consultation process itself. But a sandbox should not be a short cut to enacting or introducing new rules without any external scrutiny, nor should it remove the need for proper piloting where there is a large scale system change, such as Making Tax Digital. Its value potentially lies in exposing issues in the early stages of policy development and helping HMRC to assess whether a policy idea is workable and likely to have the desired effect before it becomes ‘baked in’ and consequently harder to deviate from.

LITRG response

The Low Incomes Tax Reform Group’s (LITRG) response highlights concern that implementing and operating sandboxes in practice presents challenges which do not have easy solutions. For example, the response states that sandbox participation should, in general, require taxpayer consent – but notes that this might mean the population of sandbox participants is too small and/or not representative enough for HMRC to draw useful conclusions.

There may, however, be statistical approaches to compensate for this – and in very limited circumstances there might be an argument for some form of mandatory participation depending on the pilot concerned (for example, where a policy was certainly going to progress in broadly similar form and where ultimately the whole population will be affected). However, mandatory participation would need to be accompanied by robust safeguards, support and controls.

The response also discusses the difficulty of maintaining fairness between sandbox participants and non-participants, and highlights the necessity of providing suitable support to sandbox participants (especially those who are unrepresented). It also notes that the question of fairness might be considered both ways: the sandbox environment might be favourable or not favourable compared with the position outside it. Finally, any safeguard that protects the taxpayer from additional tax, penalties or interest as a result of being in the sandbox should not rely on HMRC discretion – the safeguard should be an accessible legal right.

ATT response

The ATT agrees that might be helpful to test new ideas or processes on smaller groups before expanding to the wider population. However, we are yet to be convinced that the proposed approach of legislative sandboxes is the next step in solving the problems of testing new systems. There are a number of other aspects of the process of testing and developing policy which we think could be usefully explored and developed, with more time spent on policy design, supported by thorough consultation, before the testing of systems and processes commences.

One way in which we think this approach could be more usefully used is to protect those who cannot immediately access HMRC systems on launch. Instead of a testing sandbox, this would be a shielding sandbox, giving HMRC the powers to grant exemptions or exclusions for those who cannot immediately access new systems or processes.

We agreed with the CIOT that participation should be voluntary and suggested that there should be permitted opt outs, as are currently provided in jury service. We also consider that an independent oversight board would be a welcome safeguard.

The CIOT’s response can be found at: www.tax.org.uk/ref1144

LITRG’s response can be found at: www.litrg.org.uk/ref2785

The ATT’s response can be found at: www.att.org.uk/ref427

Margaret Curran mcurran@ciot.org.uk
Tom Henderson thenderson@litrg.org.uk
Helen Thornley Hthornley@att.org.uk