Research and analysis

Understanding 'Unpaid Tax Agents' and their motivations to register with HMRC

Published 20 July 2023

Mixed method research with Unpaid Agents and people who receive help from Unpaid Agents.

HM Revenue and Customs (HMRC) Research Report 703.

Quantitative research conducted by Kantar Public between June and July 2022. Qualitative follow-up research conducted by Kantar Public between September and November 2022.

Prepared by Kantar Public (Cindy Ndejembi, Jan Paul Schlindwein, Kirstin Meudell, Matthew Holt, Richard Matousek and Simon Maydew) for HMRC.

Disclaimer: The views in this report are the authors’ own and do not necessarily reflect those of HMRC.

Glossary

Term	Definition
Unpaid Agents	Unpaid Agents are individuals who help others with HM Revenue and Customs (HMRC) tasks. In contrast to Paid Agents, Unpaid Agents are not paid for this help with HMRC tasks by the person they help. Unpaid Agents include friend or relative helpers, individuals working in voluntary and community sector organisations, employers, managers, colleagues, capacitors (including those holding Power of Attorney and executors), and pro bono professionals specialising in finance or tax matters including agents, accountants and lawyers.
Person who receives help, people who receive help, people who use Unpaid Agents	A person who receives help in their dealings with HMRC from an Unpaid Agent.
Semi professionals	Unpaid Agents who help in a more formal capacity, that is as a manager or colleague, as part of a volunteer position, helping the vulnerable as a job, professionally but pro bono, or as an executor, Power of Attorney or trustee. It should be noted that the category of semi professionals is for reporting purposes to allow for a distinction between the non professional capacity of friends and relatives help, compared with those helping in a more professional capacity.
Friends or relatives	Unpaid Agents who help as friends, relatives or both.
Direct contact	Direct contact is defined as helping someone in any of their dealings with HMRC and in doing so, interacting directly with someone at HMRC or with a HMRC system (for example, speaking to HMRC on someone’s behalf over telephone or via email).
Indirect contact	Indirect contact is defined as helping someone in any of their dealings with HMRC but not interacting directly with someone at HMRC or with a HMRC system.
Trusted Helper Service	The Trusted Helper Service is currently offered by HMRC and allows individuals to register online to help a friend or relative with a number of HMRC related tasks. One individual can register as Trusted Helper for up to 5 other people.
64-8 form	The ‘Authorising your agent (64-8)’ form provides authorisation for someone to act on behalf of someone else in their dealings with HMRC, for individual or business tax affairs.

1. Executive Summary

1.1 Research background

HM Revenue and Customs (HMRC) and HM Treasury’s 10 year Tax Administration Strategy sets out the vision of a single digital account, bringing together data across different taxes and different data sources to provide personalised services for taxpayers. The single digital account aims to also improve parallel services for agents and representatives, working towards HMRC’s vision of services designed for all agent types from the outset, with agents able to see and do what their clients can.

As part of the process of designing the single digital account, HMRC is looking at how best to serve the needs of different types of helpers or representatives. For example, those needing assistance can formally appoint someone to help them with HMRC related tasks by using paid services provided by organisations such as accountants and known as ‘Paid Agents’. Another option would be to seek help from an unpaid source, for example a friend or family member who might help free of charge and would be known as an ‘Unpaid Agent’.

The Trusted Helper Service is currently offered by HMRC and allows individuals to register online to help a friend or relative with a number of HMRC related tasks. One individual can register as Trusted Helper for up to 5 other people. This research was designed to uncover more about Unpaid Agents to support the design of the future single digital account, envisioned to have additional functionalities and support options beyond the Trusted Helper Service.

For the purposes of this research, HMRC defined Unpaid Agents broadly to cover multiple different types of helpers who do not receive payment for the help they provide to others in dealing with HMRC related tasks. Unpaid Agents therefore include friends and relatives, individuals working in voluntary and community sector organisations, employers, managers, colleagues, capacitors (including those holding Power of Attorney and executors), and pro bono professionals specialising in finance or tax matters, including agents, accountants and lawyers.

The research aimed to provide HMRC with insight into the size and nature of the Unpaid Agent population. It also helped to inform design decisions related to the single digital account and the services that Unpaid Agents will be able to access and help with. Specifically, the research aimed to:

• develop an understanding of the scale of the Unpaid Agent population, and the types of helper relationships
• understand the types of help offered, and different services helped with
• develop an understanding of those who use Unpaid Agents
• develop an understanding of how HMRC can motivate Unpaid Agents and the people they help to register their helping relationships with HMRC

To meet these research objectives, a mixed method approach was taken. An initial quantitative online survey was undertaken between June and July 2022 and sought to understand the number of Unpaid Agents in the UK population and profile their basic characteristics. As there was little existing information on the population of interest (Unpaid Agents), a pragmatic research solution had to be used to first define the population. This was done by inviting members of an existing online research panel to participate, rather than the more traditional random probability approach used for understanding known populations of interest. The approach used, whilst pragmatic, does not have the same level of statistical validity as a random probability survey and therefore a certain degree of caution should be used when interpreting quantitative findings.

A representative sample of 3150 people from the UK population responded to the survey invite and were asked questions to determine whether they helped someone else in their dealings with HMRC or were helped themselves. Overall, 520 people reported helping someone else free of charge in the past two years and were therefore defined as Unpaid Agents. 589 people reported being helped by others. Those defined as Unpaid Agents were then asked further questions about who they helped, how they helped, and why they helped.

The qualitative research was undertaken between September and November 2022 and expanded on the quantitative findings with a focus on the motivations and barriers of acting as an Unpaid Agent, as well as for those receiving help. Qualitative interviews were undertaken with 33 Unpaid Agents and a further 5 interviews with those helped by an Unpaid Agent. As part of the interviews, feedback on the Trusted Helper Service was collected and 4 propositions were tested by presenting them to respondents for comment. These propositions described potential services aimed at motivating Unpaid Agents and the people who use them to register their relationship with HMRC. This report summarises the combined findings from the quantitative and qualitative research.

1.2 Key findings

Understanding Unpaid Agents in the UK population

The quantitative survey asked respondents a range of questions to identify whether they were an Unpaid Agent, or helped by an Unpaid Agent, the capacity they helped in, why they were selected to help, and their reasons for helping. The qualitative interviews expanded on this, focusing on the reasons Unpaid Agents decided to help.

Findings from the quantitative survey show that:

• 19% of the UK population had helped someone with their taxes free of charge in the past 2 years
• 22% of the UK population had been helped by an Unpaid Agent in their dealings with HMRC in the past 2 years
• Unpaid Agents most commonly helped as a friend or relative (67%), whilst 41% helped as a ‘semi professional’ Unpaid Agent, that is in a more formal capacity
• Unpaid Agents who were helping friends or relatives were significantly more likely to only help just one person
• in comparison, those Unpaid Agents working in semi professional capacities were significantly more likely to be helping more people
• Unpaid Agents were selected by those requiring help mainly because they were trusted (35%), or were considered to be good with numbers (40%)
• the main reasons Unpaid Agents provided help was because the person who receives help did not know what to do (37%) or did not understand tax issues (31%)

The qualitative interviews uncovered additional reasons why Unpaid Agents helped others. Key reasons for providing help included care and altruism, the sharing of household tasks, or as part of wider support they provided to their relatives.

Experiences of providing and receiving help

The quantitative survey asked respondents questions to identify the types of tax and specific tasks they helped people with, as well as how easy it was to deal with HMRC when providing help. The qualitative interviews expanded on the types of tasks Unpaid Agents helped with, to further understand the involvement of Unpaid Agents.

Findings from the quantitative survey show that:

• taxes most commonly assisted with included Income Tax from Self Assessment (39%) and PAYE (38%)
• benefits most commonly assisted with included Working Tax Credit (27%), Child Benefit (24%) and Child Tax Credit (20%)
• specific tasks assisted with included registering for a Personal Tax Account (30%), calculating tax, preparing accounts or filing returns (29%), registering for taxes or benefits (27%), as well as general guidance (27%)

The quantitative analysis identified varying levels of help given by Unpaid Agents, ranging from gathering paperwork, which only involved indirect contact with HMRC, through to directly acting on someone else’s behalf in interactions with HMRC. The qualitative analysis further refined these findings and identified low, medium and high levels of involvement between Unpaid Agents and those they help:

• low involvement included indirect dealings with HMRC, such as providing verbal guidance or advice
• medium involvement help included activities such as helping with filling out paperwork, but without speaking to HMRC on behalf of the person who receives help
• high involvement help included direct dealings with HMRC, such as calling HMRC on behalf of the person who receives help

In the quantitative survey, 56% of Unpaid Agents reported that it was easy to deal with HMRC when acting on behalf of someone else. In addition, the qualitative interviews found that pain points in dealing with HMRC on someone’s behalf were typically experienced by Unpaid Agents with higher levels of involvement. This included for example needing to be physically present to complete HMRC tasks.

Awareness of registering someone to help when dealing with HMRC

The quantitative survey asked respondents whether they were aware that the person they have been helping can register online to allow them to act on their behalf when dealing with HMRC. It also went on to explore whether respondents had already been registered in this way or if they would be interested in registration in the future. No specific service or scheme was mentioned, only the general concept of registering to help when dealing with HMRC.

The qualitative interviews focused on what made someone willing to register. As described above, they also collected feedback on the Trusted Helper Service specifically, as well as 4 propositions testing potential future services.

Findings from the quantitative survey show that:

• 63% of Unpaid Agents reported that they were aware that the person they were helping could register online to allow someone else to act on their behalf in their dealings with HMRC
• 70% of those Unpaid Agents who were aware reported that the person they were helping had already registered their relationship with HMRC
• relatively high levels of awareness of registration among Unpaid Agents might, to some extent, be attributed to the broad and hypothetical nature of the questions asked (see Section 5 for more detail)

The qualitative interviews revealed 3 types of Unpaid Agents: Educational, Permanent and Uncomfortable. These are important categories for analysis when further examining willingness to register. Individuals falling under the 3 types differed in their tax expertise and intentions to provide long term help. The 3 types can be found across the helping relationships of semi professionals, friends or relatives.

These different types of Unpaid Agents identified similar drivers to using the Trusted Helper Service, such as accountability, control and independence. They also identified barriers to the Trusted Helper Service, such as the limit of helping 5 people, or the limited tax areas currently covered in the service.

Preferences for specific propositions were based on the Unpaid Agent type as well as other factors. Efficiency, vulnerability of the person who receives help, geographical distance and lower levels of trust between the person who receives help and the Unpaid Agent, increased the willingness to register and an interest in the propositions.

In contrast, trust between the person who receives help and the Unpaid Agent, providing support without contacting HMRC directly (low and medium involvement) as well as formality, proportionality and complexity of the propositions had a negative impact on motivations to register.

1.3 Key messages

The findings clearly indicate that there is not one type of Unpaid Agent, and not one type of help provided. Instead, Unpaid Agents are a diverse group who help people with a range of HMRC tasks. Furthermore, there are multiple factors that influence an Unpaid Agent’s motivation to register with HMRC. The conclusions drawn from mixed method findings of this research can be categorised into three main considerations and recommendations for HMRC:

• focusing efforts on the Unpaid Agents most likely to register with HMRC
• providing more detail about the registration process to motivate Unpaid Agents and those they help to register
• focusing on the different needs of Unpaid Agents and the people they help to provide targeted support

2. Introduction

2.1 Background

HMRC and HMT’s 10 year Tax Administration Strategy sets out the vision of a single digital account, bringing together data across different taxes and different data sources to provide personalised services for taxpayers. The single digital account aims to also improve parallel services for agents and representatives, working towards HMRC’s vision of services designed for all agent types from the outset, with agents able to see and do what their clients can.

As part of the process of designing the single digital account, HMRC is looking at how best to serve the needs of different types of helpers or representatives. For example, those needing assistance can formally appoint someone to help them with HMRC related tasks by using paid services provided by organisations such as accountants and known as ‘Paid Agents’. Another option would be to seek help from an unpaid source, for example a friend or family member who might help free of charge and would be known as an ‘Unpaid Agent’.

For the purposes of this research, HMRC defined Unpaid Agents broadly to cover multiple different types of helpers who do not receive payment for the help they provide to others in dealing with HMRC related tasks. Unpaid Agents therefore include friends and relatives, individuals working in voluntary and community sector organisations, employers, managers, colleagues, capacitors (including those holding Power of Attorney and executors), and pro bono professionals specialising in finance or tax matters, including agents, accountants and lawyers.

Prior to this research, HMRC had limited evidence on the scale and scope of people acting as Unpaid Agents. Existing insight on this population includes findings from the Individuals Strand of HMRC’s Individuals, Small Business and Agent Survey 2020 which show that 19% of individuals aged 16 and above reported receiving help in their dealings with HMRC, from someone other than a professional advisor. This included 11% receiving help from a friend, family member or colleague.

Unpaid Agents and the people they help can register their helping relationship through HMRC’s Trusted Helper Service, which enables a friend or family member to perform a small number of tasks on behalf of the customer. Historically, the Trusted Helper Service has had low take up, although recent changes which allow customers to access the system through Government Gateway have seen an uptick in registrations. The Trusted Helper service differs from the single digital account offer in that it already exists and is limited in the areas of taxes it includes. The single digital account is an envisioned future service and is intended to have additional functionalities.

Qualitative research into the Trusted Helper Population 2019 identified two different types of helper: those with ‘strong family links’ and those with ‘loose community links’. Those with strong family links tended to help a small number of relatives with a broad range of social issues, whereas those with loose community links tended to help a larger number of people from their community with specific tax issues.

This research also indicated that legitimacy and transparency were key drivers for registering for the Trusted Helper Service. Helpers viewed tax as a serious matter and wished to be open to HMRC, whilst also feeling it was ‘the right thing to do’. Barriers to registering to the Trusted Helper Service included a reluctance to take on responsibility, and a lack of perceived benefits, as they were already informally sharing Personal Tax Account login details. Other barriers to registering included a lack of awareness of the service, as well as some people feeling some of the advice or help they received may not be correct or up to date.

2.2 Research aims

The current research aimed to provide HMRC with more in depth insight into the size and nature of the Unpaid Agent population to inform design decisions related to the single digital account and the services that Unpaid Agents will be able to access and help with. This research also aimed to provide HMRC with insight into the behaviours and attitudes of Unpaid Agents and people they help.

The research objectives were to:

• develop an understanding of the scale of the Unpaid Agent population, and the types of helper relationships
• understand the types of help offered, and different services helped with
• develop an understanding of those who use Unpaid Agents
• develop an understanding of how HMRC can motivate Unpaid Agents and the people they help to register their helping relationships with HMRC

2.3 Methodology

An initial quantitative online survey sought to understand approximately how many Unpaid Agents there are in the UK population and profile their basic characteristics.

As there was little existing information on the population of interest (Unpaid Agents), a pragmatic research solution had to be used to first define the population. This was done by inviting members of an existing online research panel to participate, rather than the more traditional random probability approach used for understanding known populations of interest.

A representative sample of 3150 people aged 16 and above responded to the survey invite and were asked questions to determine whether they helped someone else in their dealings with HMRC or were helped themselves.

Overall, 520 respondents reported helping someone else free of charge in the past two years and were therefore defined as Unpaid Agents. 589 reported being helped by someone else. Those defined as Unpaid Agents were asked further questions about who they helped, how they helped, and why they helped. The online survey took place between 18 June and 1 July 2022.

The qualitative research was informed by findings from the quantitative phase and comprised 38 interviews. The qualitative interviews explored respondent responses in more depth and developed greater insight into their motivations and concerns. The interviews took place between 26 September and 18 November 2022.

A purposive sample frame was developed using criteria identified in the quantitative phase. Areas of interest were sorted into primary, secondary and tertiary criteria, and the frame was monitored throughout the recruitment and fieldwork phases.

The sample comprised 20 Unpaid Agents who helped as a friend or relative, 13 Unpaid Agents who help in a semi professional capacity, 3 people who were helped by a friend or relative and 2 people who were helped by a semi professional.

Respondents were recruited from two sources: from respondents of the quantitative survey who had given permission to be re-contacted, and through free-find methods. All respondents were screened using a questionnaire to understand their helping needs and behaviours.

For further details on the survey design and execution, as well as qualitative design and research materials, please see the methodology section of the Technical Annex.

2.4 Interpretation of findings

This report outlines the findings from the quantitative and qualitative research. Quantitative survey data is weighted. Subgroup analysis has been conducted, but differences between subgroups are only reported when they are both statistically significant and relevant to the survey objectives. Additional analytical conventions include:

• results with a base size of fewer than 50 respondents are not included
• percentages for single response questions may not always add up to exactly 100% because of rounding
• where respondents have given multiple responses to a question, the sum of the individual responses may be greater than 100%
• statistical significance is judged at the 95% confidence level and therefore, questions asked of all Unpaid Agents will have a maximum confidence interval of +/- 5%
• there will be wider confidence intervals for any subgroups, for example, a subgroup of 50 respondents will have a maximum confidence interval of +/- 14% and a subgroup of 100 respondents will have a maximum confidence interval of +/- 10%

Qualitative research is not representative of the prevalence of perspectives and experience in the wider population. Rather, using a purposive sampling approach allows qualitative research to focus on the drivers of perceptions and processes.

Throughout the report, verbatim respondent quotes are used to illustrate findings. To provide additional detail, respondent quotes are labelled using information about the capacity and type of Unpaid Agent. For example:

“Quote.” Friend or Relative Unpaid Agent

2.5 Limitations and caveats

The quantitative research was designed with the aim of understanding the size of the population of Unpaid Agents in the UK, as well some of their key characteristics and who, how and why they help.

The approach used, whilst pragmatic, does not have the same level of statistical validity as a random probability survey among a known population and therefore a certain degree of caution should be used when interpreting findings.

The online panel used is a general purpose panel, however, panel members are internet users so there is a bias towards those who are more digitally capable.

Analysis of the survey data revealed certain subgroups of interest, however, it should be noted that there are limitations to how much detail can be provided by the survey as the overall samples are relatively small.

3. Understanding Unpaid Agents in the UK population

This section explores who makes up the population of Unpaid Agents in the UK. Through both quantitative and qualitative findings, this section also provides insight into the types of relationships between Unpaid Agents and the people they help, the reasons Unpaid Agents help, and the reasons people require help.

3.1 Prevalence of Unpaid Agents in the UK population

Just under one in 5 of the UK population reported having helped someone in any of their dealings with HMRC free of charge within the last 2 years (19%). A further one in 20 reported that whilst they had not helped someone, they wanted to do so (6%). Men were significantly more likely to have reported having helped (23%) compared with women (17%). Younger respondents were also more likely to report having helped, with 31% of those aged between 16 and 24 reporting helping others in any of their dealings with HMRC, compared with 10% of those aged over 65.

Slightly more than one in 5 of the UK population reported having received unpaid help in their dealings with HMRC during the last 2 years (22%). This compares similarly with figures reported in the Individuals Strand of HMRC’s Individuals, Small Business and Agent Survey 2020, where 19% of individuals aged 16 and above said they received help in their dealings with HMRC from someone other than a professional advisor.

Among the 6% who reported wanting to help but not having done so, there were a variety of reasons for not helping (see Figure 1). The most common reason for not helping was not knowing enough about taxes (38%). A further 25% reported not having helped due to not feeling comfortable acting on behalf of someone else.

Figure 1: Reasons for not being able to help to someone in their dealings with HMRC

Reasons for not helping	% of total
I do not know enough about taxes	38%
I do not feel comfortable acting on behalf of someone else	25%
I did not know that I could contact HMRC on someone else’s behalf	16%
I do not know how to register online to act on someone else’s behalf	14%
I do not live close enough to the person/people I want to help	12%
I do not know how to contact HMRC	9%
No one has asked me to help	7%
It was more work than I expected	6%
I did not have time to help	4%
Other	2%
Don’t know	11%
Prefer not to say	2%

Source: Why have you not been able to provide help to someone else in their dealings with HMRC? Base (unweighted): All respondents who answered that they hadn’t helped anyone else in any of their dealings with HMRC free of charge in the last two years but wanted to (176). Total does not sum to 100% due to multicoding.

3.2 Characteristics of Unpaid Agents in the UK populations

Respondents identified by the survey as Unpaid Agents (meaning they reported having helped someone else in any of their dealings with HMRC free of charge in the last 2 years) were asked more detailed questions about who they helped and how. For these questions, respondents were asked to focus on the person they helped most frequently.

Unpaid Agents helped others in their dealings with HMRC in a variety of different capacities, most commonly as a friend or relative (67%).

Figure 2: In what capacity Unpaid Agents offered help

Capacity in which offered help	% of total
As a friend or relative	67%
As a manager or colleague of another employee at your work	18%
As a volunteer in a voluntary or community sector organisation, for example Citizens Advice	14%
Helping vulnerable people as part of my job	13%
In a formal role as an unpaid or pro bono capacity (such as professional tax agent or lawyer)	11%
As Power of Attorney, Executor or Trustee	5%
Any other	2%
NET: Not as a friend or relative	41%
Don’t know	1%
Prefer not to say	1%

Source: For the next few questions please think about the person you have helped most frequently. In what role did you help this person? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

The capacities in which Unpaid Agents helped broadly split into two larger categories: friends or relatives and those not helping as a friend or relative. Those not helping as a friend or relative comprised those helping as the manager or colleague of another employee at work, as a volunteer in a voluntary or community sector organisation, those helping vulnerable people as part of their job, or in a formal capacity as an unpaid or pro bono tax agent or lawyer. It also included capacitors, such as people with power of attorney, executors or trustees. 41% helped in these more formal capacities overall.

People with certain demographic characteristics were more likely to help in certain capacities. For example, older people were more likely to help in the capacity of a friend or relative (96% aged 65 or above compared with 56% aged between 16 and 24). More than three quarters of women (76%) helped in the capacity of a friend or relative compared with 61% of men.

Younger people were more likely to help in other capacities, for example, the capacity of a manager or colleague of another employee at work, or as a voluntary worker. As such, 2 in 3 of those aged between 16 and 24 (65%) helped in another capacity (that was not that of a friend/relative) compared with 7% of those aged 65 or older.

When asked what person specifically they were helping, almost half of Unpaid Agents reported helping a family member (49%) in their dealings with HMRC. A further one in three helped a friend or colleague (34%) and one in 10 helped an employee (10%). A small proportion helped ‘other’ people (1%) including clients, customers or neighbours. A small proportion reported helping someone they did not know previously (4%). The status of these unknown relationships, that is whether they are part of a working relationship or not, is not clear.

Figure 3: Person helped in their dealings with HMRC by an Unpaid Agent

Person helped by an Unpaid Agent	% of total
A family member	49%
A friend or colleague	34%
An employee	10%
Someone you did not know previously	4%
Other	1%
Don’t know	1%
Prefer not to say	1%

Source: For the next few questions please think about the person you have helped most frequently. Who was it that you helped in their dealings with HMRC? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520).

Older people were more likely to help a family member with two thirds of those aged 65 or older (68%) helping a family member compared with just over one third of those aged between 16 and 24 (39%). Furthermore, more than half of women helped a family member (55%) compared with fewer than half of men (44%).

Those helping family members were most likely to help a spouse or partner (41%). One in 4 helped a relative aged between 31 and 64 (24%) with smaller proportions helping younger or older relatives.

The qualitative research confirmed and clarified the types of family relationships between Unpaid Agents and those who received help. The qualitative interviews covered examples of partners helping their partner, younger Unpaid Agents helping older family members, including parents and grandparents, and older family members helping younger family members including their children, nieces, and nephews. Furthermore, the qualitative interviews also included examples of siblings helping one another, and vulnerable people being helped by family members of varying ages.

The qualitative interviews also explored the role of semi professionals, with examples of colleagues helping colleagues in a range of sectors. These interviews also involved professionals helping clients, including social workers giving tax advice or guidance as part of their daily job, lawyers helping their clients with tax affairs, as well as volunteers helping people through voluntary and community sector organisations.

The quantitative research indicated that 43% of respondents reported helping only one person, whilst 36% helped two to five people and 11% helped 6 to 10 people. Smaller proportions reported helping more than 10 people.

Overall, Unpaid Agents who helped only one person tended to be older. 68% of those aged 65 or older helped one person but only 23% of those aged between 16 and 24 helped one person. Conversely, 35% of those aged between 16 and 24 reported helping 6 or more people compared with 3% of those aged 65 or older.

Most commonly, Unpaid Agents helped someone in their dealings with HMRC a few times a year (34%). 15% helped monthly, more than a quarter helped once a year (28%) and 15% reported only ever having helped once (see Figure 4).

Figure 4: Frequency with which an Unpaid Agent helped someone in their dealings with HMRC

Frequency with which an Unpaid Agent helped	% of total
At least once a month	15%
A few times a year	34%
Once a year	28%
Less often than once a year	6%
Only helped once	15%
Don’t know	2%
Prefer not to say	1%

Source: For the next few questions please think about the person you have helped most frequently. How often have you helped this person in their dealings with HMRC in the last two years? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to rounding.

Younger people were more likely to report helping at least once a month. More than one in 4 (27%) aged between 16 and 24 helped at least once a month compared with only 2% among those aged 65 or older. In addition, Unpaid Agents who reported only ever helping once were more likely to be older. Almost one in 4 (24%) helped only once among those aged 65 or older compared with only 6% among those aged between 16 and 24.

Those helping in a semi professional capacity were significantly more likely to help at least monthly. For example, those helping in the capacity of an employee’s manager or colleague were most likely to offer help at least once a month (31%) compared with those helping in the capacity of a friend or relative (12%). Those reporting helping only once were also more likely to be helping in the capacity of a friend or relative (18%) compared with those helping in the capacity of an employee’s manager or colleague (7%).

Overall, 6 in 10 respondents expected to help in a similar capacity again (60%). Men were more likely to say they expected to help again (64%) compared with women (55%).

3.3 Reasons for requiring help and selecting an Unpaid Agent

Trust or a perceived competence with numbers were key reasons for choosing a particular Unpaid Agent to help. Experience in dealing with HMRC or experience in working in finance or accounting were also important reasons. Unpaid Agents were usually selected for more than one reason (on average 2.3 reasons given per respondent).

Figure 5: Reasons for which an Unpaid Agent was selected to help

Reasons for which an Unpaid Agent was selected to help	% of total
I am good at numbers generally	40%
They trust me	35%
I have experience dealing with HMRC	32%
My work gives me experience with finances or accounts	28%
I am good at dealing with things online	25%
I have some financial qualifications	21%
It was convenient for me to help	20%
I am a qualified accountant or tax professional	14%
They could not afford to pay a professional	5%
Something else	2%
Don’t know	2%
Prefer not to say	1%

Source: What were the main reasons that this person asked you to help in their dealings with HMRC? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

The capacity in which an Unpaid Agent helped played a part in the reasons for selection. Friends or relatives were more likely to have been selected to help because they were trusted or because it was convenient. For example, among those helping as a friend or relative, 45% were selected to help because they were trusted. This compared with only 21% selected for this reason when helping as a manager or colleague. Similarly, among those helping as a friend or relative, 24% were selected to help because it was convenient. Being selected due to convenience was lower when helping as a manager or colleague (15%).

Those helping in a semi professional capacity were more likely to have qualifications or financial experience. For example, among those helping in the capacity of a manager or colleague, 39% were selected because they had some financial qualifications compared with only 17% selected on these grounds helping in the capacity of a friend or relative. Among those helping in the capacity of a manager or colleague, 22% were selected because they were qualified accountants or tax professionals compared with only 8% selected on these grounds helping in the capacity of a friend or relative.

3.4 Understanding why Unpaid Agents agree to help

The qualitative research helped to clarify the reasons Unpaid Agents agreed to help people. For friends and relatives, there were 5 main reasons: their expertise and knowledge of tax affairs, as part of the sharing of tasks in a household or as part of more general help they offer their loved ones, out of a general sense of care and altruism, or in some cases, simply because the person needed help, whether they felt comfortable providing it or not.

Overall, the qualitative findings demonstrated that partners helping partners were motivated by care and altruism, helping as part of their household task sharing and because of their knowledge of tax affairs. Young people helping older people were similar, except instead of task sharing, they were accustomed to providing help as part of the more general life support they provide for their older relatives. Vulnerable people tended to need the help more than others, and Unpaid Agents helped whether they felt like tax experts or not. For these Unpaid Agents, providing help gave reassurance that their relative would not be susceptible to the negative consequences of making mistakes with their tax affairs.

“I don’t want to see her get into trouble by missing something or submitting false information by accident. It also helps me to see that she’s got everything straight.” Friend or Relative Unpaid Agent

3.5 Reasons why an Unpaid Agent’s help was required

Based on the quantitative survey, the most common reason for requiring help was not knowing where to look or what to do (37%) or a lack of understanding of tax issues (31%). Other commonly reported barriers included disabilities (20%) or language (18%).

Figure 6: Reasons why an Unpaid Agent’s help was required

Reasons why an Unpaid Agent’s help was required	% of total
They did not know where to look or what to do	37%
They do not understand tax issues	31%
Disabilities made it difficult for them	20%
English is a foreign language to them	18%
Mental health problems made it difficult for them	17%
They are not confident using online systems	16%
They were not confident using HMRC online systems	15%
They were concerned about making errors	15%
They do not have access to a computer or internet	14%
They have difficulty reading or writing in English	14%
They find it overwhelming or scary dealing with HMRC	13%
Quicker or easier for me and the person needing help	9%
They were unable to sign up for a HMRC account	6%
Change in circumstances meant required more support	6%
They could not afford to pay a professional	5%
Something else	1%
Don’t know	2%
Prefer not to say	1%

Source: What were the main reasons that this person needed help? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

Older people were more likely to report being selected to help because the person they were helping did not know where to look or what to do. 48% aged 65 or above were selected for this reason compared with 33% aged between 25 and 44. As well as a lack of expertise, the qualitative research also highlighted vulnerabilities that may impede a person’s ability to manage their tax affairs, such as mental health. Unpaid Agents helping vulnerable relatives described the importance of being with the person as they completed HMRC tasks, providing a level of help simply by being present.

“People with mental health issues I think that’s the very, very common thing that they will need help because when you see too many things on the page, your brain just can’t cope, so for somebody to just sit there with you and just sort of you know put it into more layman terms that they can understand”. Friend or Relative Unpaid Agent

These cases extend the quantitative findings by demonstrating how connected the reasons for needing help are. In the example provided, the vulnerable person being helped by the interviewee experienced a range of barriers to completing HMRC tasks on their own. This included their mental health, their confidence with online systems, a concern about making errors and feeling overwhelmed dealing with HMRC. This example also establishes the link between the specific tasks that require help from Unpaid Agents and the type and intensity of help given by Unpaid Agents, which is the focus of the next section.

4. Experiences of giving and receiving help

Drawing on both the quantitative and qualitative insights, this section details the specific HMRC functions Unpaid Agents provide help with. This includes tasks Unpaid Agents help with, as well as the specific tax and benefits most likely to require an Unpaid Agent’s help. The section also outlines differing levels of involvement in dealing with HMRC and how easy Unpaid Agents found their experiences.

4.1 HMRC tasks that Unpaid Agents provide help with

Unpaid Agents were most likely to help fill out online forms (51%). Just over one in 3 (36%) provided someone with information or guidance and a further one in 3 helped someone get their paperwork together (33%).

Further analysis of the quantitative data suggested that the tasks Unpaid Agents provide help with could be divided into two groups based on the type of interaction (direct or indirect) with HMRC. Direct contact is defined as helping someone in any of their dealings with HMRC and in doing so, interacting directly with someone at HMRC or with a HMRC system (for example, speaking to HMRC on someone’s behalf over telephone or via email). Indirect contact is defined as helping someone in any of their dealings with HMRC but not interacting directly with someone at HMRC or with a HMRC system (see Glossary).

Where quantitative findings in this report use the terms direct and indirect contact, this is in accordance with how these terms are defined here.

The qualitative interviews with Unpaid Agents and people who received help developed this theme further. Based on findings from the qualitative interviews with Unpaid Agents and people who received help, tasks were further organised by 3 different levels of involvement. The direct and indirect definitions were used to explore contact with HMRC in the qualitative interviews, however the findings demonstrated that there were 2 layers of indirect contact that were helpful to define. The qualitative findings in this report therefore use the terms low, medium and high levels of involvement as this was found to be more meaningful. These categorisations became important distinctions when it came to understanding Unpaid Agents’ motivation to register with HMRC (see Section 5: Willingness to register with HMRC).

Low involvement: Unpaid Agent help provided verbal advice and guidance on tax affairs, such as directing a person to the HMRC website or alerting them to a form they may need to fill out. Examples of this type of help might include social workers helping in a semi professional capacity by directing the people they help to the HMRC website and would be considered ‘indirect’ in nature.

Medium involvement: Unpaid Agent help involved providing help with tasks beyond those of low involvement help, such as sitting next to the person who receives help while they fill out a Self Assessment form. Tasks at this level remained indirect in nature, for example:

“They’re usually physically next to me when I file the return. I gather all the information, go through it with them to make sure it’s all correct. We log on to the system, enter the information.” Friend or Relative Unpaid Agent.

Other examples of Unpaid Agents with medium levels of involvement include helping vulnerable family members. For one Unpaid Agent, it was important to provide help alongside their family member, but without taking over, so that the person who receives help maintains a sense of independence.

“Giving her independence because you are helping her to still do things, she’s actually completing with you. She can do things with you that maybe she couldn’t do on her own. With her I try and support her rather than do everything.” Friend or Relative and Semi Professional Unpaid Agent.

High involvement: Unpaid Agent help involved acting on a person’s behalf, for example by logging in and navigating their Personal Tax Account or directly contacting HMRC on a person’s behalf. This was considered the most resource intensive help offered by Unpaid Agents. Examples of Unpaid Agents with high levels of involvement include partners who are in charge of all financial and tax affairs in their household, meaning that they have access to their partner’s login details and complete Self Assessments on their behalf.

Figure 7.1: Types of help provided by Unpaid Agents: High involvement

Types of help provided	% of total
I spoke to HMRC on their behalf over the telephone	22%
I emailed HMRC on their behalf	14%
I wrote or sent letters to HMRC on their behalf	5%

Source: What type of unpaid help have you provided? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

Figure 7.2: Types of help provided by Unpaid Agents: Medium involvement

Types of help provided	% of total
I helped them fill out online forms	51%
I helped them get their paperwork together	33%
I helped them fill out paper forms	28%
I signed into their HMRC online account with the details the gave me	12%

Source: What type of unpaid help have you provided? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

Figure 7.3: Types of help provided by Unpaid Agents: Low involvement

Types of help provided	% of total
I provide them with information or guidance	36%

Source: What type of unpaid help have you provided? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

Figure 7.4: Types of help provided by Unpaid Agents: Other responses

Types of help provided	% of total
Any other unpaid help	1%
None of the above	3%
Don’t know	2%
Prefer not to say	0%

Source: What type of unpaid help have you provided? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

4.2 Taxes that Unpaid Agents helped people with

The quantitative research found that the taxes most commonly helped with included Income Tax through Self Assessment (39%) or Income Tax from wages (PAYE) (38%). This was supported by the qualitative research. Only a small proportion did not help with any kind of tax at all (6% indicated none of the above). Where Unpaid Agents reported that they helped with ‘Other tax’, verbatim comments indicated these included references to Capital Gains, Inheritance Tax or Tax Credits.

Figure 8: Taxes that Unpaid Agents helped people with

Taxes that Unpaid Agents helped with	% of total
Income Tax through Self Assessment	39%
Income Tax from wages (PAYE)	38%
Income Tax from a pension	23%
National Insurance	21%
VAT	11%
Company Tax	6%
Other tax	3%
None of the above	6%
Don’t know	2%
Prefer not to say	2%

Source: Which of the following taxes have you helped this person with? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

The frequency with which help was given varied by the type of tax helped with. Among those helping once a year, 51% helped with Self Assessment compared with 36% among those helping several times a year. Conversely, among those helping monthly or more often, 54% helped with income tax from wages compared with 26% who helped less often than once a year.

In addition, the capacity in which an Unpaid Agent helped made a difference to the types of taxes helped with. Those helping as a friend or relative were less likely than other groups to help with things like pensions, National Insurance, VAT or Company Tax. Among those helping as a friend or relative, 22% helped with Income Tax from a pension compared with 35% helping with this task when helping as a manager or colleague. Among those helping as a friend or relative, 17% helped with National Insurance compared with 30% helping with this task when helping as a manager or colleague.

4.3 HMRC benefits or allowances that Unpaid Agents helped people with

HMRC benefits or allowances Unpaid Agents helped people with most included Working Tax Credit (27%), Child Benefit (24%) or Child Tax Credit (20%).

Figure 9: HMRC benefits or allowances that Unpaid Agents helped people with

Benefits or allowances helped with	% of total
Working Tax Credit	27%
Child Benefit	24%
Child Tax Credit	20%
Tax credit but not sure which	14%
Statutory payments such as maternity pay or sickness benefit	13%
Marriage Allowance	10%
Student loan repayment	5%
Tax-Free Childcare	5%
Other	3%
None of the above	30%
Don’t know	2%
Prefer not to say	2%

Source: Which of the following other HMRC benefits or allowances have you helped this person with? Base (unweighted): All respondents who helped anyone else in any of their dealings with HMRC free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

The frequency and capacity in which an Unpaid Agent helped was associated with the types of benefits and allowances helped with. Those helping in the capacity of a friend or relative were less likely to help with benefit related tasks than those helping in other capacities. Younger Unpaid Agents, helping more people, more frequently and in direct contact with HMRC, were more likely to be helping with benefits and allowances. Among those aged 25 to 44, 35% reported helping with Working Tax Credit compared with 12% among those aged 65 or older.

Among those helping monthly, 42% reported helping with Working Tax Credit compared with 10% helping less often than once a year. Furthermore, among those helping between 2 and 5 people, 38% helped with Working Tax Credit compared with 16% among those helping one person. Additionally, the research indicated that among those making direct contact with HMRC, 35% of respondents reported helping with Working Tax Credit compared with 22% making indirect contact. For Child Benefit a similar pattern is evident.

4.4 Specific tasks Unpaid Agents helped with

Specific tasks helped with included registering for a Personal Tax Account (30%), calculating tax (29%), registering for taxes or benefits (27%) as well as general guidance on tax issues (27%). On average, Unpaid Agents helped with 2 tasks.

Figure 10: Specific tasks Unpaid Agents helped with

Specific tasks helped with	% of total
Registering for an HMRC Personal Tax Account	30%
Calculating tax, preparing accounts or filing returns	29%
Providing general guidance on tax issues	27%
Registering for taxes, Tax Credits or Child Benefit	27%
Contacting HMRC to query or resolve a mistake	18%
Registering as an Authorised Helper using Trusted Helper Service	16%
Registering for Tax Free Childcare	15%
Contacting HMRC about an under or overpayment of tax	14%
Contacting HMRC to make a complaint	5%
Contacting HMRC to discuss a debt	4%
Offering advice that as a qualified professional you would usually charge for	0%
Other	3%
Don’t know	2%
Prefer not to say	2%

Source: Which of the following types of help have you provided this person? Base (unweighted): All respondents who helped anyone else in any of their dealings with HM Revenue and Customs (HMRC) free of charge in the last two years (520). Total does not sum to 100% due to multicoding.

Those helping in the capacity of a friend or relative were more likely than other groups to provide general guidance and less likely to help with more formal types of help such as registering for an HMRC Personal Tax Account. For example, 27% helping in the capacity of a friend or relative reported helping someone to register for a Personal Tax Account. This compares with 58% helping in the capacity of a manager or colleague or 56% helping in the capacity of a volunteer in the voluntary or community sector. Similarly, 9% helping in the capacity of a friend or relative reported registering as a Trusted Helper compared with 29% helping in the capacity of a manager or colleague of an employee or 49% helping in the capacity of a volunteer.

4.5 Channels used by Unpaid Agents to deal with HMRC on behalf of those they help

Unpaid Agents were most likely to be in contact with HMRC via searching on the internet (46%) or using a Personal Tax Account (38%). Friends and relatives were less likely than other groups to work through formal or digital structures such as the Business Tax Account, webinars or webchats. For example, 12% helping in the capacity of a friend or relative reported using webinars or webchats and 10% reported using the Business Tax Account. This compares with 31% helping in the capacity of a manager or colleague of an employee using webinars or webchats and 35% using the Business Tax Account.

4.6 Unpaid Agents contact with HMRC and satisfaction with experience

In the quantitative survey, email was reported by Unpaid Agents as the most commonly used method of contacting HMRC on behalf of someone else (59%). This was followed by telephone (45%). Digital methods such as webchat (21%) or the app (13%) were used less commonly.

More than half of Unpaid Agents reported it was easy to deal with HMRC on someone else’s behalf (56% reported that it was very or fairly easy). Just over one in 10 (13%) found it difficult.

Furthermore, in qualitative interviews, Unpaid Agents noted that HMRC’s systems had become easier to navigate in recent years. Unpaid Agents highlighted an improvement in the availability and usability of online services but described the challenges of contacting HMRC over the phone, in particular the long wait times. Unpaid Agents also noted some difficulties understanding terminology and described finding the website confusing, often struggling to find the answers to queries.

“If something doesn’t work and you need to get in touch with HMRC, phoning them is a nightmare. You’ll be on the phone for at least an hour”. Friend or Relative Unpaid Agent

Unpaid Agents with more intense involvement (medium and high involvement level), found communicating with HMRC on behalf of someone to be a particular point of friction. Existing services proved problematic for Unpaid Agents when they needed to be physically present with the person they were helping in order to carry out tasks, for example, speaking on behalf of someone on the phone. This caused issues due to difficulties finding the time to help and the need to travel large distances to provide help in person.

In the quantitative survey, the majority of Unpaid Agents reported being able to help with everything they were asked to do (86%). 7% were not able to do everything they were asked to.

5. Willingness to register with HMRC

This section focuses on the willingness of Unpaid Agents and the people they help to register their helping relationships online with HMRC. This section covers both awareness of the possibility to register as an Unpaid Agent, and general willingness to do so. It also details the drivers of and barriers to registering. Furthermore, it explores propositions designed to increase willingness to register for both Unpaid Agents and the people they help.

HMRC currently offers the Trusted Helper Service where an individual can register online with HMRC to help a friend or relative. This service allows Unpaid Agents to complete a number of tasks on behalf of someone else.

In the quantitative survey, Unpaid Agents and people who received help were asked whether they were aware of the possibility to register their helping relationships with HMRC and, if they had not yet done so, whether they would be willing to register in the future. The survey questions did not make any references to the Trusted Helper Service. This means that the quantitative findings reflect respondents more general perspective on registering their helping relationship with HMRC.

Building on from the quantitative survey, the qualitative research gained more in depth views on the Trusted Helper Service and registration in general. In addition, it explored 4 propositions to better understand the motivations and barriers of Unpaid Agents to register with HMRC. Each proposition approached the topic of registration from a different perspective and included various service design options. The propositions were presented and discussed during the qualitative interviews.

5.1 Perceptions of registration: Trusted Helper Service and beyond

5.1.1 Awareness of registration and willingness to register

The quantitative research sought general information and opinions on registering helping relationships with HMRC. Questions in the quantitative survey asking about awareness of registration and willingness to register were deliberately worded broadly. The aim of wording questions broadly was to capture respondents’ general opinion on registration and to cover the different ways in which Unpaid Agents can register their helping relationship with HMRC. Examples of ways helping relationships can be registered with HMRC are the Trusted Helper Service, Power of Attorney, 64-8 form and paid agent digital services. Because of the broad wording of the questions, reported awareness of registration may be higher than expected, had the questions solely focused on the Trusted Helper Service.

In the quantitative survey, almost 2 in 3 Unpaid Agents reported being aware that the person they were helping could register them to act on their behalf when dealing with HMRC (63%). Unpaid Agents aware of this typically helped more people, more frequently and in a semi professional capacity. For example, awareness was at 55% among those helping one person but increased to 71% among those helping between 2 and 5 people and 72% if helping more than 6 people. Among those who helped less often than once a year, awareness was at 52% but increased to 83% among those helping monthly. Additionally, awareness was 60% among those helping in the capacity of a friend or relative but increased to 79% among those helping as a manager or colleague and to 92% among those helping the vulnerable as part of their job.

7 in 10 Unpaid Agents (70%) who were aware that they could be registered to act on behalf of someone else reported having been registered to do this. Around half of those who were either not aware or aware but not registered, were willing to register in future (53%).

Unpaid Agents who reported that they were not registered were more likely to help fewer people, less often, with tasks that did not involve direct communications with HMRC. For example, among those helping one person, 46% reported that they had not been registered to act on behalf of others. In contrast only 19% of those helping 6 or more people reported that they had not registered. Similarly, among those helping less often than once a year, 42% reported that they had not been registered. In comparison, among those helping at least once a month only 15% reported that they had not been registered.

Looking at the wider UK population, among those who had been helped in their dealings with HMRC, 63% were aware that someone else could be registered to help on their behalf. Furthermore, almost 2 in 3 who had been helped (64%) would be interested in registering someone to help on their behalf. A further 3 in 10 (30%) were not interested and one in 20 of those who had been helped (6%) reported that someone was already registered to act on their behalf.

As stated above, the quantitative survey referred to registration in a more general sense, and responses were recorded without an explanation of registration being provided. In contrast, the qualitative interviews provided a more thorough explanation of the registration process, and this additional information led to participants changing whether they were willing to register or not. Some Unpaid Agents, both semi professional and friend or relatives, who perceived registration to be too much effort in relation to the effort required to complete the helping task, changed their minds as they learned of potential improvements in efficiency and verification, leading them to be more willing to register than before.

In contrast, some initially willing Unpaid Agents, both semi professional and friend or relatives, stated that they were less willing to register after receiving more information. These Unpaid Agents saw the benefits of reducing the burden to HMRC call centres through improved access, however, once hearing about the process, they became more concerned about the amount of access Unpaid Agents may have, leading to feelings of discomfort.

5.1.2 Willingness to register and intention to provide future help

From the qualitative interviews, 3 different Unpaid Agent types emerged: Educational, Permanent or Uncomfortable Unpaid Agents. These 3 types were found across friend and relative as well as semi professional relationships.

Unpaid Agents falling under each of the 3 types differed in their tax expertise and intentions to provide help in the long run. The type of Unpaid Agent had important implications for the interest in registering with HMRC.

Educational Unpaid Agents: This type of Unpaid Agent possessed relevant expertise in tax affairs. They had an interest in educating the person who received help to become more independent in dealing with their tax affairs. This type of Unpaid Agent therefore had no interest in providing help in the long run. With regards to registration, Educational Unpaid Agents were interested in effective solutions without a long term commitment. An example of an Educational Unpaid Agent would be someone who helped their colleague on a short term basis. Another example would be an Unpaid Agent who helped their partner but wanted them to become more independent in dealing with HMRC tasks in the future.

Permanent Unpaid Agents: Similar to Educational Unpaid Agents, Permanent Unpaid Agents had relevant expertise in tax affairs. However, they were more comfortable than Educational Unpaid Agents in providing long term help. Permanent Unpaid Agents who helped as friends and relatives tended to help the same person, while those helping in a semi professional relationship helped different people over time. In addition, Permanent Unpaid Agents who helped friends or relatives tended to provide help on the basis of mutual trust. Those helping in a semi professional relationship tended to provide help based on their professional capacity rather than mutual trust. An example of Permanent Unpaid Agents would be someone who took over all financial responsibilities within a relationship with a partner. Another example would be an Unpaid Agent who works as a lawyer or social worker who provided HMRC advice as part of their regular client engagement.

Uncomfortable Unpaid Agents: In contrast to Educational and Permanent Unpaid Agents, the Uncomfortable Unpaid Agent had limited expertise relevant to HMRC tasks. They provided help mainly because the particular person they were helping needed support, not because they had any relevant expertise. At the same time, Uncomfortable Unpaid Agents were open to providing help both in the short and long run. An example would be an Unpaid Agent who helped their vulnerable family member or a young Unpaid Agent who helped their parents.

The qualitative research also found that Unpaid Agents could present as different types within different relationships. They could be an Educational Unpaid Agents when they helped their colleague but a Permanent Unpaid Agent when they helped their partner. However, Uncomfortable Unpaid Agents did not act as Educational or Permanent Unpaid Agents due to their limited expertise relevant to HMRC tasks.

The qualitative interviews showed that an Unpaid Agent’s general willingness to register is mainly dependent on their type. This was more important than their status as semi professional and friend or relative for how Unpaid Agents considered the propositions explored in section 5.2.

5.1.3 Barriers and drivers of general registration

The qualitative interviews revealed different barriers and drivers when discussing registration as an Unpaid Agent with HMRC in general. General registration was discussed before sharing more details on the Trusted Helper Service and the propositions.

Across all types of Unpaid Agents, a sense of increased responsibility and accountability as a result of registration was identified as a barrier to registration. In addition, several concerns were raised around:

• the independence of the person who receives help, in particular if they were vulnerable or a client within a semi professional relationship
• would registration create additional workload, both for the Unpaid Agent and the person who receives help
• the need for specific qualifications to be eligible to register as an Unpaid Agent
• a general lack of confidence in HMRC managing additional services such as Unpaid Agent registration

No obvious benefit of registration was identified for Unpaid Agents who provided help such as verbal guidance or form filling (low or medium involvement). Similarly, registration did not appear to offer benefits for occasional help. From the perspective of people that receive help, Unpaid Agents can often already help them without registration, particularly for Unpaid Agents who live close to the person who receives help.

However, the qualitative research also revealed drivers which increased the willingness to register. The drivers were less generalisable and more dependent on the type of Unpaid Agent. For Permanent and Educational Unpaid Agents, registration appeared to offer improved efficiency in two ways. Firstly, the person who receives help does not need to be present when helping, this was particularly relevant for high involvement. Secondly, Permanent and Educational Unpaid Agents expected tailored information access such as access to relevant parts of the online tax account of the person who receives help without having to rely on this person to provide it.

For Uncomfortable Unpaid Agents, registration allowed them to be more confident in providing help by operating within an official HMRC approved framework.

Another driver was identified by Permanent Unpaid Agents who helped in a semi professional relationship. Such relationships can be more transactional and less built on mutual trust. In this case, registration increased the legitimacy of the Unpaid Agent by operating within an official framework which could reduce concerns around potential abuse.

5.1.4 Barriers and drivers of registering with the Trusted Helper Service

Following the discussion of general registration, respondents were presented information on HMRC’s Trusted Helper Service and asked questions about their perceptions of this service. The information on the Trusted Helper Service set out that Unpaid Agents who are registered with the service can perform several tasks, such as checking whether the person who receives help is paying the right amount of income tax, checking or updating that person’s personal tax account, and checking or updating that person’s company car tax. Respondents were further informed that the person who receives help must accept the registration of the Unpaid Agent and is still legally responsible for their own tax. Finally, respondents were informed that Unpaid Agents can register as a Trusted Helper for up to 5 people.

Feedback from the respondents revealed various drivers of and barriers to registering with the Trusted Helper Service in its current set up. Across all types of Unpaid Agents (Educational, Permanent or Uncomfortable), respondents identified the limit of helping 5 people in the current service as either too high or too low. For example, for Unpaid Agents with large families or networks, helping only 5 people would be too low. For other respondents, helping up to 5 people seemed unmanageable and raised concerns to what extent proper help can be provided. Furthermore, across all types of Unpaid Agents, respondents indicated that the tax areas currently covered by the Trusted Helper Service do not align with the taxes and benefits they provide help with. For example, an Unpaid Agent who helped with Self Assessment would currently not be covered by the service.

No particular benefit of registration was identified for Unpaid Agents who provided help categorised as low or medium, such as verbal guidance or help with form filling. Having login details for the HMRC online account of the person who receives help available, also often reduced the perceived benefits of the Trusted Helper Service. It should be noted that HMRC strongly advises against sharing login details with others and that customers doing so may be non-compliant.

Important drivers which increased an interest in registering with the Trusted Helper Service, across all respondents, were:

• the accountability of the person who receives help for the decisions taken by the Unpaid Agent which reduced concerns about the liability of the Unpaid Agent
• the approval requirement in the Trusted Helper Service which ensured control and independence of the person who receives help while preventing abuse
• reducing the need for a vulnerable person with communication difficulties to communicate directly with HMRC

In addition to these drivers, Permanent and Educational Unpaid Agents expected the Trusted Helper Service to enable more efficiency by reducing the need to involve the person who receives help. They also saw value in the service for allowing remote help and information access.

5.3 Propositions: What would motivate registration?

As part of the qualitative research, 4 propositions were developed to better understand the motivations of Unpaid Agents to register with HMRC. It is important to note that each proposition represents a hypothetical approach towards registration and HMRC is not committed to implementing any of the discussed features. Each proposition approached the topic of registration from a different perspective. The propositions were presented and discussed during the qualitative interviews with respondents.

5.2.1 Proposition 1: Helping people to stay in control

The first proposition was about helping people who receive help from Unpaid Agents to stay in control. It was described as: “A way of giving help that allows the Account Holder control to give as much or as little access to their account, for as short or as long a time as they like.” The proposition envisioned a dashboard in which the account holder could immediately turn permissions on or off. Permissions would affect different aspects of their account and for different lengths of time. In addition, HMRC would periodically check whether both sides are happy to proceed with the scheme. If permission would not be renewed after a certain period of time, the account access for the Unpaid Agent would be automatically ceased.

5.2.2 Proposition 2: Being clear with each other before you start

This proposition was about both the Unpaid Agent and the person who receives help being clear with each other before they started. It was described as: “This will give both parties the tools to think through, discuss and agree on levels of sharing, decision making power, and for how long.” The idea of this proposition was to bring clarity into the process by HMRC giving help by:

• explaining how helping relationships work
• taking both sides through what agreements could be made
• taking the person who receives help through how to make decisions that work for them
• covering aspects such as who is responsible for mistakes and what is the cooling off period if someone made a mistake within which the account holder can reverse the action

In addition, the activities of both the person who receives help and the Unpaid Agent would be tracked to highlight who has done what which would provide clarity on both sides. If the person who receives help were to request this, HMRC could also provide them with a summary of the activities on the account and who has done what.

5.2.3 Proposition 3: Being seen to do the right thing

This proposition was about allowing the Unpaid Agent to be seen to be doing the right thing. It was presented as “By making the helping relationship formal and visible it will reassure everyone, including HMRC, that the Account Holder’s best interests are being considered. It also gives you protection”.

Once registered, HMRC would contact the account holder and the Unpaid Agent to confirm the relationship. Both parties would have documentation to show to anyone who has concerns about the helping relationship. For those helping in a semi professional capacity, they would have documents to keep in their records.

5.2.4 Proposition 4: Making helping easy

The last proposition was about making help easy. It was presented as: “HMRC will make providing help as easy as possible, while protecting the Account Holder’s interest”. Once registered and with the permission of the account holder, HMRC will work with the Unpaid Agent as if they were the account holder. HMRC would fast track authorisation for certain cases: people with Powers of Attorney; people who are already registered with HMRC’s Agent Services Accounts; helpers registered with specific voluntary organisations; and other circumstances.

5.2.5 Feedback for each proposition including what works, what requires improvement and who it was more or less suitable for

Each proposition received detailed feedback regarding positive and negative aspects as well as improvement ideas. Each proposition was also analysed to understand who it was more or less suitable for.

Feedback on Proposition 1: Helping people to stay in control

This proposition was valued across all respondents for its focus on ensuring that the account holder maintains control. The periodical check by HMRC was also positively received to adapt to changes in the relationship between an Unpaid Agent and the person who receives help. However, periodical checks were less important for Permanent Unpaid Agents in friend and relative relationships due to the high levels of mutual trust and the permanent nature of these relationships. Educational and Permanent Unpaid Agents valued expected efficiency gains by tailored information access.

However, requirements for the account holder to review the access was considered as additional admin. Ideas of safeguarding the account holder’s control were often not relevant to relationships built on mutual trust, in particular Permanent Unpaid Agents who helped friends and relatives.

Improvement ideas suggested included automatic renewals instead of periodical reviews. Unless the account holder actively intervenes, the access would be automatically renewed. This is particularly important for Permanent Unpaid Agents helping friends and relatives for which no changes in the relationship are expected over time. In addition, temporary account access was suggested to enable ad hoc help.

Proposition 1 was more relevant for relationships involving a vulnerable person who receives help and a Permanent Unpaid Agent who wants to balance efficient help and the person’s independence. Permanent Unpaid Agents helping vulnerable people spoke of a concern that the balance of power would be uncomfortable if the person they were helping did not maintain control of their own account, and therefore this proposition was very appealing. Similarly, it was more relevant for any type of relationship characterised by lower levels of trust such as colleagues or professional to client relationships. It was also relevant for Permanent Unpaid Agents with high involvement but with geographical distance. In this case the proposition would allow efficient help without the person who receives help being present.

“That’s a perfect solution (…) it still gives them their independence.” Permanent, Friend or Relative Unpaid Agent

Feedback on Proposition 2: Being clear with each other before you start

The summary of decisions and the joint decision making were viewed positively by respondents. Automatic summaries would reduce the need to inform the person who receives help about the progress. This is particularly important for Educational Unpaid Agents. For this type of Unpaid Agent, the summary would allow for traceability which would help the Unpaid Agent to monitor the decision of the person who receives help.

Certain aspects of proposition 2 were viewed less positively, for example, the summary established a level of accountability and responsibility which was felt to go beyond helping relationships that are built on care and altruism. The proposition also appeared complex with high involvement requirements of the person who receives help. Overall, it was regarded as a highly formal framework.

It was suggested to use a more formal language instead of the casual language present in the proposition.

It was felt the proposition was less relevant for vulnerable people who received help as it could create a dependency on the Unpaid Agent due to the joint decision making. It would also be less relevant for Permanent or Educational Unpaid Agents helping friends and relatives as these relationships are felt to be built on trust and do not require a formal agreement. Proposition 2 was more relevant for Permanent Unpaid Agents as it enables a transparent process of help while informing the person who receives help about the progress. It was also more relevant for complex tax issues with high involvement over a longer period of time, as it allows both sides to monitor the process.

“If I make a mistake, would I receive any kind of fine? I would be concerned about that. Although I’d be registered as a helper, I don’t have any official qualifications or vetting to see how well I understand the area. It makes me wonder what the repercussions would be if I made a mistake.” Permanent, Friend or Relative Unpaid Agent

Feedback on Proposition 3: Being seen to do the right thing

All respondents highlighted that documentation increased trust and a sense of overall security. Unpaid Agents with low involvement preferred it for being less formal than proposition 2.

However, this proposition was still considered rather formal, especially for intimate permanent relationships. It was also unclear how much access the Unpaid Agent would have and what they would be able to do. Respondents suggested including a time limit, similar to Proposition 1, to make sure access is reviewed after a certain period and that the level of access the Unpaid Agent would have should also be clarified.

Proposition 3 was less relevant for Permanent Unpaid Agents in relationships with high levels of trust because it was perceived as too formal. It was more relevant for Permanent Unpaid Agents with medium to high involvement in professional relationships. In this case, the Unpaid Agent could provide the right documentation to prevent any accusation of abusing the tax account of the person who receives help.

“People are sceptical and do question things, so I must be clear that I am playing a very clear role with no benefit to me, otherwise your intent could be questioned, and my reputation is really important to me.” Educational, Friend or Relative and Semi Professional Unpaid Agent

Feedback on Proposition 4: Making helping easy

Across Unpaid Agents with medium to high involvement, the ease of access granted to the Unpaid Agent was seen as a positive as it would require less interaction with the person who receives help. This would allow for potential efficiency gains by enabling fast decision making and processes. The fast tracked options for sensitive cases such as with Power of Attorney, were also positively highlighted.

However, proposition 4 and the unrestricted access of the Unpaid Agent was criticised by people that receive help, who felt it reduced their control in the process. Similarly, Educational Unpaid Agents thought it would take away the agency of people that receive help. Suggested improvements were to combine this proposition with the access restriction feature from proposition 1 and to add a time limit on the access.

Proposition 4 was less relevant for Permanent and Educational Unpaid Agents with high levels of trust and geographical closeness. It was more relevant for vulnerable people that receive help or Permanent Unpaid Agents with medium to high involvement. In both cases, proposition 4 would allow for easy and efficient help due to the high levels of control of the Unpaid Agent.

“I think this one is a no, there needs to be some formal restriction on what can be accessed by the helper.” Educational, Semi Professional Unpaid Agent.

5.2.6 Proposition features and general factors driving and lowering motivations to register

The analysis revealed features across all propositions as well as general factors which impacted the motivation of respondents to register.

Geographical distance between the person who receives help and the Unpaid Agent was a factor impacting the motivation to register. For example, some Unpaid Agents were driving long distances to provide help when a task required them to help in person. Being able to provide help remotely by accessing the account of the person who receives help could reduce the need to travel. Another factor driving motivation to register was vulnerability and independence of the person who receives help. Unpaid Agents who helped a vulnerable person often wanted to ensure the independence of the person who receives help while providing efficient help.

“I don’t want to rob him of his own initiative and independence. I want to build up his confidence to help him be as independent as possible.” Permanent, Friend or Relative Unpaid Agent

Lower levels of trust were another factor which positively impacted motivation to register. Registration offered the Unpaid Agent a way to demonstrate legitimacy while minimising concerns about potential abuse. This can be particularly relevant for semi professional relationships in which Unpaid Agents and the person who receives help can be more distant.

Efficiency was another positive factor for the motivation to register. For example, Unpaid Agents valued expected efficiency gains and a way of making decisions without the person who receives help being present. Overall, both Educational and Permanent Unpaid Agents were interested in efficient solutions.

However, the qualitative interviews also revealed general factors and proposition features which lowered the motivation regarding registration and the propositions.

Many relationships were built on mutual trust, particularly for Permanent Friend or Relative Unpaid Agents. In such relationships, issues such as accountability or information sharing were not important while login details of HMRC accounts could be shared. This diminished the potential benefits of a registration. Lower motivation regarding registration was also associated with low and medium involvement, in this case, registration did not offer any particular advantages

“When I’m helping family members, it’s an understanding that they need help and they trust that I will do it correctly, that it is on both of us if we make a mistake.” Permanent, Friend or Relative Unpaid Agent

Formality was a factor which raised concerns from two perspectives. First, Unpaid Agents were concerned that registration could be seen as entering a formal agreement which lowered their interest in registering because it would go beyond the altruistic help they provided. Second, particularly Unpaid Agents in helping relationships built on trust perceived the degree of formality that comes with registration as not necessary because the trust outweighed a need for a formal agreement.

Related to this, Unpaid Agents in helping relationships built on trust, as well as those providing less involved help, felt that the complexity and effort of registering and the propositions such as Proposition 3 were disproportionate given that they were currently able to provide the help required without being registered.

A final factor was accountability, this was particularly relevant for semi professional Unpaid Agents. Such Unpaid Agents were often concerned about being responsible for decisions or tax affairs of the person they helped.

The qualitative research also revealed four cases in which registration with HMRC is most relevant. For each case it was possible to identify the most suitable proposition.

1. High involvement: This level of involvement often involves calling HMRC on behalf of the person who receives help. If registration would eliminate the need for the person receiving help to be present and give verbal consent when calling HMRC, this would be an important motivation for registration. This is relevant across all Unpaid Agent types and helping relationships. This aspect was not covered by the propositions.

2. Permanent Unpaid Agents with high involvement who live far away from the person who receives help: Particularly for Permanent Unpaid Agents with high involvement, registration could reduce the need to have the person who receives help being present. Proposition 1 (helping people stay in control) as well as proposition 4 (making help easy) are particularly suitable because they would enable efficient help.

3. A Permanent Unpaid Agent who helps a vulnerable person: Permanent Unpaid Agents who helped a vulnerable person often wanted to ensure that this person stayed in control and maintained their independence. Registration could provide an official framework which allows the Unpaid Agent to provide effective help while ensuring that the vulnerable person maintains control over their tax account. Proposition 1 particularly speaks to these needs because it would allow the vulnerable person to turn access permissions for the Unpaid Agents on and off as they see fit.

4. A Permanent Unpaid Agent with lower levels of trust between them and the person they help: Examples of relationships with lower trust were those between lawyer and client, colleague to colleague, but this also occurred within families. In these relationships trust was lower because the people receiving help had concerns about the Unpaid Agents’ intentions including whether the Unpaid Agent may misuse use the access to their tax account. Registration was perceived as a good way to address these concerns because it provides an official framework approved by HMRC. This was particularly relevant where helping relationships were expected to be permanent. Proposition 1 (helping people stay in control) or 2 (being clear with each other before you start) would be especially suitable because they ensure the account holder’s control (Proposition 1) or a summary of decisions (Proposition 2).

6. Conclusions

Overall, the findings of this research confirm and extend existing research. The purpose of this research was to better understand Unpaid Agents and the people who use them, and to understand what would motivate Unpaid Agents to register with HMRC. These findings were linked as different types of Unpaid Agents had different motivations to register with HMRC. The findings clearly indicate there is not one type of Unpaid Agent, and not one type of help provided. Instead, Unpaid Agents are a diverse group who help people with a range of HMRC tasks. Furthermore, there are a range of factors that influence an Unpaid Agent’s motivation to register with HMRC, and different Unpaid Agents prefer different digital solutions. Specifically, the conclusions drawn from mixed method findings can be categorised into three main implications for HMRC:

• focusing efforts on the Unpaid Agents most likely to register with HMRC
• providing more detail about the registration process
• focusing on the different needs of Unpaid Agents and people helped

6.1 Focusing efforts on the Unpaid Agents most likely to register with HMRC

In the survey, around half of Unpaid Agents not aware of registration or not registered to act on behalf of another person indicated that they were willing to register in the future (53%). Unregistered Unpaid Agents were more likely to help fewer people and with lower involvement tasks that did not require communicating with HMRC. Frequency of help was also important, as for those helping only once a year, 42% were not registered, compared with the 15% of unregistered Unpaid Agents who helped at least once a month. Whilst again, for the quantitative research the concept of registration was left broad and therefore the exact proportions should be treated with caution, efforts to encourage registration should be focused on the unregistered Unpaid Agents who are willing to register.

The qualitative findings provided more clarity on which of the unregistered Unpaid Agents this is likely to be, highlighting their level of involvement with those they are helping and their intentions for helping in the future as indicators of willingness to register. Unpaid Agents, both semi professionals and friends or relatives, with medium and high levels of involvement were more willing to register with HMRC, even if they only helped one person once a year. For these Unpaid Agents, registration could provide them with greater efficiency in completing their tasks, and therefore, even with a lower frequency of help, the act of registering was viewed as proportionate. These tasks involved a need to contact HMRC and to act on behalf of someone else, or where Unpaid Agents were required to be physically present with the person they were helping in order to complete a task. In contrast, Unpaid Agents who helped with some low and medium involvement tasks including providing advice or helping people fill out forms, were less willing to register, even if they helped multiple times a year. In line with this, people who were helped with low and medium involvement tasks were less willing to register, as it was perceived as too effortful.

These findings suggest that focusing on the level of involvement may be more important than the frequency of help in understanding willingness to register, however further quantitative research into involvement would provide more confidence for HMRC to prioritise solutions.

How an Unpaid Agent intends to help in the future was also related to their willingness to register. Registration was seen as particularly beneficial by Permanent Unpaid Agents, either acting as a semi professional or as a friend or relative, in particular when helping vulnerable people, where low trust in the relationship was seen as an issue or where they lived far away from the person they were helping. Therefore, because of their intention to continue helping others with their tax affairs, Permanent Unpaid Agents would be a useful focus for increasing registration with HMRC. Educational Unpaid Agents had more mixed attitudes about registration due to their intention for the person they were helping to become responsible for managing their own tax affairs in the future. Uncomfortable Unpaid Agents seemed unlikely to become willing to register, despite seeing some benefits, which are detailed in the next section. Further research investigating the prevalence of the 3 Unpaid Agent types, Educational, Permanent and Uncomfortable, would provide useful insights to inform the prioritisation of HMRC efforts moving forward.

6.2 Providing more detail about the registration process

Understanding registration is an important step in motivating Unpaid Agents to register. This research quantified awareness, finding that 63% of Unpaid Agents were aware that the people they helped could register to allow the Unpaid Agent to act on their behalf. Awareness was higher amongst Unpaid Agents who had helped six or more people (72%). It was also higher amongst Unpaid Agents who had helped people at least once a month (83%). These figures compared to 55% of Unpaid Agents who helped only one person. A slightly lower figure of 52% was also recorded amongst Unpaid Agents who helped less than once a year.

Overall, of those Unpaid Agents aware of registration, 70% had been registered by someone to act on their behalf in the past. These findings were limited by the broad definition of registration used, and the follow up qualitative interviews revealed that there was mixed understanding of what registration entails.

Where the quantitative survey captured a single snapshot of willingness to register, the qualitative findings were able to demonstrate that willingness changes, either increasing or decreasing, when Unpaid Agents have more detailed and accurate information about registration. There were cases in the qualitative research where Unpaid Agents, both semi professional and friends or relative, initially saw registration as too burdensome in comparison to the task they were completing. Once hearing of the benefits, in particular the additional verification and efficiency of access, they became more willing to register with HMRC.

In contrast, there were also cases where learning more about registration decreased an Unpaid Agent’s willingness to register. Where an Unpaid Agent could initially see the benefit of having easier access to accounts and agreed with reducing the burden on HMRC call centres, once learning more about registration, they were concerned about the amount of access Unpaid Agents may have.

Based on these findings, there are two ways forward. Firstly, in future survey research about awareness and understandings of registration, an explanation of registration, the registration process and the benefits of registration should be provided alongside the question. Secondly, more transparency about the registration could be provided to unregistered Unpaid Agents, who even if aware of the concept of registration, may not fully understand the benefits such as improved efficiency to act on behalf of someone they help.

6.3 Focusing on the different needs of Unpaid Agents

The research clearly demonstrated that Unpaid Agents are a large and diverse group. Therefore, improving their willingness to register will require targeting different types of Unpaid Agents with different solutions that address the different needs of Unpaid Agents.

As mentioned above, there is a need for efficiency for Unpaid Agents helping with medium and high involvement tasks. Common examples of this were about improving digital access to accounts and reducing the need to be present with the person being helped, or to need to contact that person at the time of providing help. Therefore, digital solutions should focus on improving the efficiency for Unpaid Agents completing tasks on behalf of someone else.

The qualitative findings showed that all Unpaid Agents, whether semi professional, friends or relatives, Educational, Permanent or Uncomfortable, emphasised the importance of maintaining the independence of the person they helped. This was a particularly strong preference for Educational Unpaid Agents who desire the people they help to be able to manage their own affairs in the future, as well as Permanent Unpaid Agents helping vulnerable people, who felt uncomfortable with the power imbalance. Similarly, the people who were helped by Unpaid Agents were very interested in maintaining their own independence and taking part in the management of their own tax affairs.

As discussed throughout the report and in the above section, the interviews revealed three types of Unpaid Agents, that spanned across friends and relatives and semi professionals. Because the type of Unpaid Agent was closely aligned with the Unpaid Agents’ intentions to continue helping people in the future, messaging about registration could emphasise different benefits in the context of future help.

Educational Unpaid Agents were interested in helping people become more aware of tax affairs and to start taking on more and more responsibility in their own tax management. These Unpaid Agents were interested in any solutions that made the helping process easier but were not as interested in any solutions which formalised the relationship as a long term helping arrangement. In contrast, Permanent Unpaid Agents viewed helping behaviours as part of their role, either at work, volunteering or in their friendship or family relationships. Permanent Unpaid Agents also sought greater ease in HMRC processes and systems that enabled helping and were more interested in formalised arrangements that were transparent and clearly defined roles.

Uncomfortable Unpaid Agents helped people because the person needed help, and not because they had any particular expertise. Uncomfortable Unpaid Agents were interested in solutions that provided greater confidence in the help they were providing, but also seemed the most reluctant to register. This means that while they saw the benefits of helping within an HMRC approved framework and with support, they may still be hard to reach.

Registration could also be seen as more relevant by different types of Unpaid Agents if their specific relationships were represented or referred to in messaging about registration. The findings clarified the nature of relationships between relatives helping relatives. The quantitative findings demonstrated the prevalence of partners helping partners, and the qualitative findings demonstrated different dynamics for other types of family relationships such as grandchildren helping their grandparents, grandparents helping their grandchildren and siblings helping one another. The qualitative interviews also highlighted other important dynamics including relatives helping vulnerable relatives.

Similarly, both the quantitative and qualitative findings provided additional insight into what was termed ‘loose community ties’ in previous research, demonstrating the types of semi professional relationships that exist. These ranged from colleagues helping colleagues, managers helping employees, to volunteers providing help to people from the community. Importantly, different Unpaid Agents had different expertise, different reasons for helping and different helping behaviours in terms of the number of people they help and the frequency with which they help.

Tailored messaging using examples of more nuanced relationship profiles may improve the perceived relevance for Unpaid Agents helping in this capacity. The profiles could emphasise the benefits of registering with HMRC for specific relationship types, for example, being able to call on your partner’s behalf instead of needing them to be in the room. The relevance of different relationships was not comprehensively tested in this research, and so further research into whether representing different relationships affects the willingness of Unpaid Agents to register may help to determine whether this is a priority for registration communications.