.png)
.png)
KEY POINTS
- Taxation’s submission to the tax agents strategy consultation was based on readers’ views.
- If enrolment went ahead, it should be for all agents.
- Independent tribunal should decide if HMRC can cease to deal with an agent.
- Views were split on whether agents should have to be qualified.
Back in June, we launched our response to the tax agents strategy consultation, inviting you to either write in, or, ideally, complete the online survey to let us have your views on the strategy proposed by HMRC for dealing with tax professionals in future.
The intention was to allow all our readers, and especially those who are either not members of institutes or who did not want to get involved in their own body’s response to the consultation, to have their say. We promised to bring these views together into a formal response to the consultation.
As well as receiving a small number of printed replies we had a large number of responses online. Not everyone replied to all the questions (one respondent answered a question a quarter of the way through with ‘oh my goodness what a long survey’), but we generally received between 50–100 responses to each.
The responses were open, so no statistics are given below, but the overall weight, and some of the specific issues raised, are set out. An edited version of this article has been submitted to HMRC as our (or, rather, your) response.
Enrolment
HMRC’s consultation document started by asking for comments on the ‘enrolment’ process, by which HMRC would have a single record for each agent, enabling them (they say) to ‘better target support and reduce the risk of fraud’.
Respondents generally thought that all agents should be subject to the enrolment process – if they thought that it should happen at all, since a small number disagreed with the whole idea. There was some support for a limitation based on whether the agent was in business, and some suggested a threshold of having ten or twenty clients.
There was also some confusion in answers that distinguished between qualified and unqualified accountants. Those who wanted only qualifieds to be enrolled were implicitly saying the unqualified should not be part of the system, those who said only unqualifieds should be enrolled seemed to think that qualifieds should only be subject to their institutes’ procedures.
One person suggested ‘a two or three-year transitional period for “grand-fathered” agents to become qualified’, and several emphasised the importance of a CPD requirement.
Most respondents wanted a multi-office firm to have one enrolment, although one person suggested a sub-registration within that so that HMRC could identify the individual office concerned. A minority either wanted office-by-office registration, or a system that was flexible enough to cope with both.
The consultation had proposed a set of basic data which would have to be provided under enrolment. A clear majority of respondents simply answered ‘yes’ when asked whether it was reasonable to provide this set of information.
The isolated disagreements mostly complained that it would be onerous and that much of the information was already held by HMRC.
Perhaps one point to draw from this latter comment is that once the information has been gathered it should not be asked for again.
While most thought that the data set proposed was sufficient, a common further suggestion was for the agents to give details of professional institute memberships.
Various suggestions for how the enrolment process should be rolled out were made, but the overwhelming comment was to avoid the tax return season of November–January.
Opinion was then divided as to whether it should be February–March or July–August, with some going for other combinations. One clearly thought-out comment is worth printing in full:
‘The obvious answer to the time of year is between February and April with a further sector in say August. I have considered whether it should be in order of size, over a number of years, so that the bigger firms are the pilot managers, but this may be unfair on smaller firms. So after rigorous testing, and pilots in Big 4, large, medium, small, and micro businesses, I suggest that it is rolled out alphabetically over a two-year period within the months suggested above.’
Two years was one of the longest periods suggested, quite a number of you thought that it should be done over a period of a few months after a trial. Several respondents emphasised the need for a good online process of enrolment, with data prepopulated if at all possible.
There was an overwhelming rejection of annual re-enrolment, although some of the longer answers accepted that re-enrolment on a prepopulated form, or by confirming existing details would be acceptable:
‘If it took ten minutes to check existing information and submit any changes then fine but we could do without lots of extra form filling.’
The consultation asked what performance indicators should be used to assess the success of the enrolment process, apart from savings in the administrative burden. Although one respondent said pithily, ‘Having to enrol is hardly going to reduce an administrative burden is it?’, many others were looking for increased efficiency – fewer letters unanswered at HMRC, greater accuracy in tax collected under PAYE, etc.’
Self-service
The next big element of the proposed strategy was allowing agents to ‘self-serve’ in certain areas – interacting directly with HMRC’s systems.
The two main areas where HMRC proposed this should happen were authorisations to act as an agent (thereby replacing the 64-8 process) and amendments to notices of coding.
In general, the self-serve options were thought to be right, although there were suggestions for additions either now or in the future – access to information held by HMRC such as P60, P11D, etc, setting off one type of tax against another, and speeding up VAT registration.
The consultation document asked what safeguards would be needed if self-service was introduced. The most common safeguard suggested was to generate a letter to the client saying that the agent had self-authorised.
Other suggestions included random auditing and spot checks, watching for constant mistakes on reductions to payments on account, etc., and the (by now familiar) call for only qualified agents to be permitted to self-serve.
Asked whether agents welcomed the increased control, and whether it could have a detrimental effect on client relationships, one respondent answered, ‘Yes and no’ and another answered,‘No and yes’.
Overall, you gave the proposals a guarded welcome, with a clear awareness that there could be problems ahead.
Respondents did not really feel they had the expertise to answer the question about the level at which IT security processes should be set. ‘Can you ask this in English, please?’ probably sums it up best.
There was no clear view as to whether the professional bodies would be able to monitor this, as the consultation document proposed.
The consultation asked whether there should be any restrictions or transitional arrangements for self-serve access. This mainly threw up the responses about qualified agents again.
The only other issue raised by a small number of respondents was to ensure the robustness of HMRC systems before implementation.
Agent view
The third, and most controversial, main element of the consultation document is the ‘agent view’.
While HMRC want to emphasise the ability to provide support and help to agents who need it, they also admitted that it would be used to identify agents who might be falling below the standards which should be expected of them.
This might involve looking at the compliance record of clients, but the consultation recognised that an unintended (and unwanted) side-effect of this could be that agents would not be prepared to take on the very clients where their professional intervention would be most useful to HMRC.
Understandably, a majority of you were clear that simply looking at client behaviour was dangerous, for the reasons that the consultation document itself identifies.
Feelings in this area were strong, with comments such as ‘this is where it becomes pernicious’, and ‘I find the notion of viewing an agent by client compliance nauseating... we will only take on clean clients and the rest can go hang’.
This was the point at which the consultation specifically asked whether agents should have to be qualified in order to act for clients on tax matters. The question split respondents roughly 50-50, with most being definite in their views one way or the other.
One respondent made the interesting comment that ‘I am concerned about the desire for qualifications when so few HMRC staff are professionally qualified and subject to institute standards and disciplinary proceedings’, which may be food for thought.
Those who answered ‘yes’ to the requirement for being qualified generally answered ‘no’ to the suggestion in the consultation document of allowing those with a lower level of qualification to submit returns online but not to self-serve – and those who had answered ‘no’ to the requirement for qualification generally answered ‘yes’ here.
While there was a range of responses to the question about the length of any transitional period for unqualified agents to qualify, those who thought that qualification was necessary generally thought that around three or four years was a sufficient time.
There was some limited support for ‘grandfathering’ those who could show they were qualified by experience, and that ‘it is not acceptable to ask those over 50, say, to take exams’.
Those who supported qualification saw a general upholding of standards and ethics as the main advantage, together with monitoring which could give HMRC some comfort.
This also fed through to the question about sanctions for ‘bad’ agents, where reference to the professional body was seen as the main sanction.
There was general distrust of HMRC being able to decide on their own that they could refuse to deal with an agent without a right of appeal: ‘Allowing HMRC to refuse to deal with an agent is already giving HMRC too much power’.
In such cases, there was an overwhelming view that the decision should be made by an independent panel.
Other issues
Most respondents were not expecting to make changes as yet based on the consultation document, though some were planning to look at their IT capability and one was ‘looking forward to retirement after 40 years in tax’.
Most were not expecting significant additional one-off costs, although some thought there could be software implications. Even fewer were expecting significant ongoing costs, and a couple of respondents specifically expected the costs to be outweighed by the savings.
There was more of a split on whether smaller practices would be proportionally harder hit – ‘Large firms can absorb these costs, the smaller ones have less play in the system’ probably sums up the views of the doubters.
By now, having answered several questions relating to the qualified versus unqualified debate, respondents’ attitudes had hardened, ‘Possibly more impact. Tough’ being one of the more uncompromising views. A more measured response was:
‘Unqualified agents... will be particularly hard hit if professionalisation is instituted at the same time. So it is better to cope with the HMRC proposals first then consider an all-exam-based qualified profession afterwards.’
Respondents were asked for the number of clients they served. Some took this to mean themselves personally, some clearly meant their whole practice. A typical number given was between 350 and 500, though the lowest was five and the highest 1,200.
The consultation period has now ended, and HMRC are analysing the responses. If the views of our readers are anything to go by, there is a readiness to take these proposals seriously and engage with HMRC on them, with the expectation that they may provide genuine benefits for advisers in practice.
However, there is a real concern about the implications of the ‘agent view’, particularly for those who try to deal with the more difficult clients, and there is a complete split of opinion on whether self-service should be limited to qualified advisers.
