Comments from ACCA to HM Revenue and Customs (HMRC), September 2011.
ACCA has observed with growing concern the apparent worsening in the relationship between HMRC and tax agents over recent years. This decline in the standard of dialogue has to a very large extent been driven by the restructuring of HM Customs and Excise and the Inland Revenue into HM Revenue and Customs, and the related headcount reductions and refocusing on automated procedures rather than human intervention in taxpayer affairs.
We cannot undo what has been done, and HMRC and the tax agent community need to try to move forward from the current position. Much of what is proposed in the consultation is sensible and constructive, recognising the changing practices of tax agents as well as the change in circumstances for HMRC. Never the less, there are still areas of concern in the proposals, and work to be done before the new models will meet the requirements of both sides. In particular our members are concerned that HMRC should address shortcomings in its own performance, and while this should not be allowed to delay progress in improving matters for agents it must not be ignored.
If agents are to have access to taxpayer information, whether on line or through other means of contact with HMRC, it seems reasonable that they should be ‘enrolled' so that HMRC and taxpayers can be certain that taxpayer security is maintained. The process should apply, with appropriate modification, for anyone who seeks to receive information from HMRC in connection with taxpayers.
Larger practices will pose significant difficulties for HMRC's current model of registration and enrolment, and the HMRC engagement proposals are equally likely to cause significant procedural difficulties for large firms. Whatever the solution settled upon to control the HMRC side of the transaction, large practices will face the need to implement management and control procedures to ensure that internal client confidentiality is maintained, especially in commercially sensitive areas such as VAT and corporation tax.
Different solutions will work best for different advisers, depending on their business and administrative models, so the best solution will be flexible to respond to their differing needs. As so little of the detail is known, it is impossible to put forward concrete suggestions, but it seems likely that some kind of two-tier process, or a facility for linking registrations, would address the needs of large firms to have effectively ‘subcategories' of enrolled credentials to allow different parts of the firm to access different sets of taxpayer information.
In the context of an agent in business the details are broadly appropriate. However it should be noted that in many cases a practice will not want to receive client monies. Doing so can raise a number of security and audit independence issues, and many practitioners would not feel comfortable with HMRC having details which might enable them to (albeit inadvertently) create such difficulties.
For most reputable businesses the information listed should be adequate to confirm identity to the required standard. In the vast majority of cases the new enrolment will simply be a reregistration of an existing well known relationship and there would be no benefit to HMRC in requesting more, while there would be a cost to businesses in providing it and HMRC in processing it.
HMRC has access to a wide range of background information in relation to taxpayers and to businesses which it uses to target suspected tax evaders of the sort who should not be enrolled. We would not expect HMRC to publicise details of its precise activities in such areas, but would expect that if inconsistencies or risk flags show up in the information already provided then HMRC should be able to investigate further, if necessary contacting the business directly to seek confirmation of details.
One further piece of information which might usefully be collected at enrolment is details of the agent's professional indemnity insurance (‘PII') cover. ACCA requires all members in practice to be covered by appropriate PII, and believes that this is a useful indicator to prospective customers of any tax agent in business of the agent's probity. PII is of course not appropriate for ‘friends and family' agents, and should not become compulsory for all registrations as agent. However, if the benefits of self-serve access to HMRC systems are to be exploited by a firm or individual for commercial gain at the expense of the general public, some measure of protection for consumers is in order, and PII cover will provide an instant filter with less need for other complex grandfathering provisions, or the need to wait for unaffiliated advisers to attain professional qualifications. Requiring PII to grant system access would of course necessitate annual re-enrolment, and issue which is addressed below.
In cases where the information provided under the standard data set gives HMRC cause for concern face to face interviews should be conducted before registration is completed.
The enrolment process must not be started in December or January. The vast majority of the agent community are fully engaged with completion of the tax compliance cycle during that period.
If re-enrolment is to be an annual process then HMRC should provide prepopulated forms to each agent so that revisions are by exception, save for details of PII cover. HMRC already engages in information sharing with the mortgage industry; implementing similar links with the insurance industry would present no insurmountable technological challenges.
Ideally HMRC would be able to assess whether enrolling agents, and dividing them into categories, leads to a reduction in the number of errors identified in tax returns submitted by the various categories of agent. However, recent experience has suggested that HMRC's measurement of statistics in this area is not sufficiently robust to support a meaningful analysis.
While most of the responses received by ACCA indicated a general willingness to consider engaging with the self serve proposals, members generally viewed them as being more appropriate to simple procedural matters than more complex areas.
However, there was a consistent theme that members do not believe that the potential administrative costs savings are ‘worth' the risks associated with the ‘agent view' proposals and the proposals for HMRC monitoring of agents, a step which is still widely perceived as little more than regulation by another name.
‘I also think increased access to dealing with certain matters for clients would be helpful, purely because it will be cheaper for us to do their job for free than to waste more time trying to get them to do their job themselves.'
The penalty for wrongful self-authorisation should be significant and clearly communicated to agents using the self-authorisation portal. HMRC should also undertake to alert taxpayers to the self appointment of a new agent. In most cases HMRC will have an email address for the taxpayer via their Government Gateway credentials, and the notification could be automatic and at effectively zero marginal cost (although the function would need to be incorporated into the initial design and build of the process). For other taxpayers, the notification would be by post. In time, taxpayers should be able to opt to receive any such notification either by SMS text message or email.
The motivation for fraudulent self-authorisation will in most cases either be financial (with the aim of triggering spurious refunds to the ‘agent's' account) or political/prurient, with the aim of accessing confidential taxpayer information.
In the former case, amongst other existing criminal sanctions, the individual responsible for making the fraudulent self-authorisation would fall clearly within the proposed ‘Dishonest Conduct' legislation, have both dishonest intent and the aim of unlawfully reducing the tax liability of the target (albeit for the benefit of the ‘agent' rather than the ‘client').
In the latter case, many of the most likely targets (MPs, celebrities, HNWIs) can be identified as a class and should be specifically warned by HMRC of the possible risks attendant to the new procedures. Taxpayers in these categories could be given the option to require two part authorisation (as happens now) for their accounts, although the extra administrative burden imposed on them, and also on HMRC in running two separate systems, is unlikely to make such an option attractive. Provided that taxpayers are aware of the need to monitor communications from HMRC, and those seeking to self authorise are aware of the processes and safeguards, and penalties for abuse of the system, the number of fraudulent self authorisations should be kept to a minimum.
Our members have expressed grave reservations about the potential impact on their relationship with clients under the new proposals. Concerns are both that the legal burden of liability might shift and also that clients' expectations might change unreasonably.
Currently, if an error occurs in processing of client information then the agent will have a trail indicating what information was received from the client, and what was submitted to HMRC. From this it will be a simple matter to establish whether the client provided the wrong information, the agent somehow changed it, or HMRC generated the wrong result having been given the correct information. The agent can be held liable only for their own errors, and it is a relatively simple matter for the agent to demonstrate what they have done. However, in a situation where the agent is directly responsible for inputting data into HMRC systems, the agent will need to have some means to evidence what they have done, along with absolute faith in the system not to in some way misinterpret their operations.
Moreover there is the potential that HMRC will view errors on the client account as stemming from the agent, rather than the client. Agents will need to maintain records to protect themselves not only from client accusations in the event of an error in HMRC's systems, but also to protect themselves from HMRC enquiries in the event of an error (or deliberate misinformation) in instructions received from their client.
From the client side, the knowledge that agents have a new ‘enhanced' status in the HMRC system may lead clients to expect more of their agents than can realistically be provided. Clients are also likely to bring more pressure to bear on agents to make amendments to the system which are not necessarily justified or in the client's best long term interests.
Security is fundamental to tax transactions. Confidentiality of taxpayer data is paramount, and to this end HMRC are best placed to ensure that their systems are secure and access to them restricted to authorised and recognised users.
The structure of the internet and the ways in which information is transmitted allow for tracking of data, and for the recipient to establish to a fair degree of certainty the origin of the transmission. Provided HMRC's systems are developed to the appropriate level, development and dissemination of appropriate guidelines on industry practice to ensure that agent credentials cannot be hijacked or imitated by other users should be relatively uncontroversial. Many agents will use specific tax or accountancy software which can be developed (following dialogue between HMRC and the developers) to interact in a recognised and predictable manner with HMRC's systems.
No. While ACCA can provide guidance for members on good IT practice it is neither practicable nor desirable for ACCA to attempt any meaningful level of monitoring of its members' IT practices and proficiency.
HMRC is under an obligation to all taxpayers to provide a safe and secure online environment in which tax matters can be addressed. While it will no doubt be challenging for HMRC to develop a robust yet useable model which is both secure and accessible for all users, this is the path that government has chosen to travel through the ‘Digital by Default' agenda.
While restriction of access to self serve to members of professional bodies appears to offer a simple route to restrict self serve access to appropriate agents it presents a number of practical and legal obstacles. While ACCA believes that membership of a professional body provides the best protection for HMRC and tax payers, imposition of such a requirement on all advisers (while it would be welcomed by some of our members) would require a significant transitional period during which existing unaffiliated advisers could attain the necessary affiliations.
As an alternative, all agents proposing to undertake tax advice on a commercial footing should demonstrate their bona fides through PII cover.
ACCA is concerned that the proposed pilot is too restricted in scope, and runs the risk of wasting the time and effort of all involved on both sides if it goes ahead without a clear understanding of the level of engagement and safeguards required for a wider scale rollout of self serve in the future.
The consultation document states that ‘the strategy does not seek to change ht relationship between client, agent or HMRC…' and yet it is clear that the self serve proposals would lead to a fundamental shift in the risks faced by agents, and the need for them to change administrative procedures to respond to those changes. Introduction of automated nationwide monitoring of agent transactions would change the relationship between the three parties still further.
HMRC have made the point that whatever level of monitoring may have taken place prior to the Inland Revenue/HM Customs & Excise merger, this would have been on an ad hoc local level, and there would have been no consistency nationally, or indeed even a single viewpoint on nationwide firms, given the number of their contacts with HMRC (as it became) in various districts. However, this highlights the value of local knowledge, and more importantly the limitations of any attempt to gather meaningful nationwide data.
The vast range of business models, client profiles, firm sizes and taxes dealt with suggest that without significant work on HMRC's part to better understand the entities in respect of which they are gathering data, that data will be very nearly as useless to them as it is now, but with the added disadvantage that HMRC will waste their own time acting on it and agents will have to waste time responding. For example, firms which specialise in handling disputed enquiries are likely to register a disproportionate number of penalties, appealed notices and other ‘negative' signals in respect of their client base, despite performing an essential function in bringing to a close disputes which have become unnecessarily costly in administrative terms for both HMRC and the taxpayer.
Meanwhile, firms which operate primarily in the sphere of providing tax and investment advice would register comparatively few ‘hits' on the HMRC system from which to form any opinion of the standard of their work, even if much of the advice given ultimately turned out to be below the standards expected by HMRC. Indeed, in the absence of suitable filters in HMRC's process, a single ‘rogue' office or department in a large practice might escape notice if they dealt with a sufficiently small proportion of the firm's work, even thought that ‘small proportion' might constitute a significant level of tax compared with the vast majority of small practices.
If the monitoring process is to have any real value, then the dataset used by HMRC for comparison will need several years to populate and settle in. Furthermore, for comparisons between firms to be valid, HMRC would need to be comfortable that it is comparing like with like. The proposals give no clear indication of how this information would be gathered, or loaded into the system. In many cases, the agents themselves will be the best, if not the only, source of that knowledge, but relying upon a self-assessment of the agent's status would impose a further administrative burden upon them, and also introduce the opportunity for dishonest individuals seeking accreditation through enrolment to misrepresent themselves at the very beginning of the process.
HMRC has been remarkably and refreshingly frank in acknowledging the lack of useful intelligence, as opposed to information, it currently has about tax agents. In view of the low base from which it is starting in attempting to create a set of data against which to compare tax agents' performance, it seems pre-emptive to discuss safeguards and appeal procedures before we can even reasonably consider the realms of what might be possible in terms of monitoring and reporting.
ACCA believes that requiring all tax advisers, and accountants, to hold some degree of formal qualification and regulation before charging the public for their services would serve to protect the public and serve as a guarantee of quality. That having been said, there are many unaffiliated advisers who have no formal professional qualifications but do provide an excellent service for their clients. There are still more advisers, professionally qualified and otherwise, whose activities involve some interaction with tax matters and who are currently authorised as tax agents for one or more of their clients. ACCA believes there should be a presumption in favour of existing agents being able to enrol into the new system, but subject to the basic requirement of demonstrating PII cover to be in place before being able to access the self serve elements of HMRC's systems.
Future registrations should be restricted to those who are able to prove sufficient knowledge and experience of the UK tax system. The simplest test would be to restrict access to those holding the appropriate professional qualifications, subject of course to direct mutual recognition of other European qualifications.
Allowing two tiers of access (for example a ‘view only' and ‘full read /write' split) could potentially be helpful to both HMRC and agents. The ability to read information from HMRC records would be useful both for ‘friends and family' agents and for book-keepers acting in a commercial capacity.
ACCA is concerned that a requirement of qualification of the agent before allowing any access to HMRC's systems on behalf of clients would be counter-productive. It is widely recognised that there are many advisers who act as agents for tax purposes with no formal qualifications or affiliation who are nevertheless are able to provide a valuable and professional service to their clients based on many years of experience. From the taxpayers' point of view, there may be little practical difference between a professionally qualified and regulated adviser with 10 years of post qualification experience and an unqualified, unregulated adviser with the same level of practical experience – at least, until things go wrong. At this point, the difference between the two may become all too painfully (and expensively) apparent to the taxpayer. All the professional bodies require their members to hold professional indemnity insurance, a consumer safeguard which is lacking in the unregulated sector. Access to taxpayer information in a business capacity should be restricted to those who can show PII cover is in place. This simple requirement would protect the public both in the event of a claim, but also before the event by reducing the scope for the criminal and the uninsurable to become agents in the first place.
‘As a sole practitioner, I believe that it would be more beneficial to carry out book keeping duties rather than being a tax agent under the new proposals.'
If full access and monitoring applies only to those with a recognised professional qualification, then the agent will be subject to the disciplinary processes of their own body, up to the ultimate sanction of losing their membership, at which point they would no longer be able to act as an agent.
For agents enrolled to the lower level of systems access, and subject only to PII requirements, we would expect that insurers would require sight of HMRC's assessment of the agent's performance, and that any demonstrable failings on the agent's part would have an adverse affect on their access to cover. However, with no access to any independent scrutiny of their relationship with HMRC, many agents might feel uncomfortable with such a situation. Many of our members expressed concern that HMRC's decision making can be inconsistent and that responses may be disproportionate and are insistent that there should be some independent oversight of HMRC's activities.
An independent body would in any event be essential if there were to be no controls on the qualifications of new agents, as there is no other way to hold unaffiliated agents to account. It is the clear view of a significant proportion of our members that appointing HMRC as an oversight body in respect of their professional work would compromise their independence, and put an unacceptable strain on their relations with both HMRC and their clients. Grave concerns have also been aired about the ability of HMRC to create a sufficiently experienced, objective and consistent oversight organ, particularly in times of funding cuts in the department. If there is to be any further oversight of tax agents' activities, it must be undertaken by an independent body which enjoys the trust and respect of all parties, staffed by individuals who can demonstrate a knowledge not only of the tax system and its administration but also of the commercial realities faced by agents in dealing with clients.
There is of course no reason why such an independent body could not also have oversight of HMRC's behaviour and performance, to act as a safeguard against poor performance by a minority of HMRC staff.
‘As professionals accountants are duty is to our clients, whilst acting in accordance with professional standards and the law. We must maintain our independence and avoid being considered as part of the bureaucracy of the state.'
Even in a well functioning tax system where agents respected and had full confidence in the abilities and impartiality of the revenue authorities, there would be strong arguments in favour of introducing an independent panel to supervise any decision to refuse to deal. In the current atmosphere, poisoned by years of mistrust and failing service provided to agents by HMRC, there are few who would welcome the prospect of HMRC having unilateral powers to destroy an agent's livelihood without any recourse to external moderation. As discussed above, there should be an independent body to oversee all aspects of the HMRC/agent relationship, and as well as the ability to strip an agent of its authority to act for taxpayers, it should have the power to review the activities of HMRC officers and report in public to the Commissioners where failings are so serious that an individual should no longer be in the position of trust and power that an officer of the Revenue enjoys.
‘It seems wholly inappropriate that [HMRC be] in the position of holding a sword of Damocles over a professional firm who may be dissuaded from acting for certain clients, or even defending taxpayers against HMRC's increasingly aggressive view of taxation and enforcement.'
The lack of detail and certainty in the current proposals makes a sensible discussion of the likely cost impacts impossible, particularly in view of the many other changes to the tax system and the commercial world currently under way. Developments in technology are changing the way firms work, with the move to online systems, cloud computing, the imposition of iXBRL and the rise of mobile working. Changes to the legal environment will allow multidisciplinary practices to offer a wider range of services, and much of the specialised advice formerly the province of ‘boutique' firms may be offered by high street names through captive advisory practices. The rise of cross border trade and changes to the EU VAT and company law regimes may change the way clients operate. E-invoicing and electronic payment methods will change the way transactions are carried out and evidenced, giving scope for long term reform of VAT and other transaction tax administrative procedures. Even without the changes prompted by the current proposals, we cannot make clear predictions about the shape or scope of accountancy practices in five or ten years' time.