Research and analysis

Enabling retirement savings for the self-employed: pensions and long-term savings trials

Updated 28 March 2019

Foreword by the Minister for Pensions and Financial Inclusion

I am proud of the government’s achievements in normalising workplace pension saving for millions of today’s workers. Our priority remains to build on this with a more inclusive, stronger savings culture for future generations. Automatic enrolment has transformed pension saving among today’s workers. Almost 10 million workers have been automatically enrolled into a workplace pension by their employer, and only 9% of those enrolled chose to opt-out. The greatest gains have been achieved among those younger workers who were historically least well served by the pensions system – in particular, many more women and the lowest-paid are building savings for their future.

Whilst automatic enrolment is successfully rebuilding the fabric of the UK’s pension system we also have to confront, like many other countries, long-term challenges and opportunities resulting from changes in the labour market and demographics. Self-employment has become an important feature of the UK labour market – with latest data suggesting around 15% of the workforce is self-employed. While individuals classified as self-employed will benefit from the new State Pension, they are much less likely than employed workers to accumulate private pension saving.

Finding effective, durable retirement saving solutions for self-employed individuals is a long-term challenge for our generation. Self-employment is highly diverse: incomes, assets and employment experiences vary widely. For many individuals the drivers of self-employment may be freedom; flexibility and control over their lives.^{[footnote 1]} Yet their aspirations; type of work; sector; and irregularity of income impacts on motivations and capacity for pension saving. We want to ensure in particular that younger generations, including those who may be in lower-paid and less secure self-employments, have sufficient opportunities to build and maintain saving habits – which may have been acquired through automatic enrolment during employment. Making it easy for individuals to save matters. The diversity of self-employment, however, demands a different kind of policy response.

As part of the 2017 Review of Automatic Enrolment^{[footnote 2]} the government committed to scope, develop and test targeted interventions aimed at establishing what works to increase retirement saving among the self-employed. Our approach provides an evidence-based step towards implementing the government manifesto commitment on the self-employed and retirement saving; this includes using insights from automatic enrolment where relevant. Over the past 12 months my Department has carried out considerable engagement with a broad range of sectors and organisations, including financial services; new and growing Fintechs; and service providers to the self-employed. This approach, supplemented by research with self-employed people themselves, is deepening our evidence-base to reveal opportunities for trials that may have greatest effect at scale.

I am delighted to confirm the Department is working with a broad range of delivery organisations and service providers to the self-employed – including Aviva plc; Barclays plc, Lloyds Banking Group plc, NEST Insight, Smart Pension and Aegon along with the Association of Independent Professional and the Self-Employed (IPSE). Working with these bodies we will commence a programme of research and trialling activity from 2018/2019, following preparatory work already begun. Our approach focuses on testing behaviourally inspired messages and tech tools, aimed at prompting self-employed individuals to save; the trials will also test how and where to deliver propositions. The trials focus on three areas:

• marketing interventions aimed at people who have previously saved – for example, after being automatically enrolled whilst employed – to encourage them to continue their saving behaviour
• marketing interventions using trusted third parties for the self-employed, such as trade bodies/unions etc. to promote the value of saving and provide an easy connection to an appropriate savings vehicle
• behavioural prompts – testing messages combined with prompts through invoicing services and/or the banking sector to seek to engage self-employed people to think about starting a regular saving habit at a point when they are receiving income

This report provides the government’s delivery plan and complements our agenda to empower and improve the consumer experience, in particular through Pensions Dashboards and the Single Financial Guidance Body. It is also a call to action. Developing and testing potential solutions relies on the development of new partnerships within and across sectors including payment systems; accounting technology; self-employed workspaces and growing Fintech firms providing new services. The government calls on organisations within these sectors to work with us to co-design and test interventions.

Our long term ambition is for future generations to have confidence in retirement saving – no matter what type of employment or self-employment they experience during their working lives – so they can prepare for greater security in later life.

Guy Opperman, MP
Minister for Pensions and Financial Inclusion

Executive summary

Introduction

1. The government is committed to enabling the self-employed to achieve greater financial security in later life. To assess the most effective approaches, we are continuing to develop our evidence on the depth and nature of pension saving amongst the self-employed, recognising this differs within and across cohorts, and given different patterns of transition between self-employment and employment.

2. The government’s 2017 manifesto included a commitment that:

We will continue to extend auto-enrolment to small employers and make it available to the self-employed.

In its 2017 Review of Automatic Enrolment^{[footnote 3]}, the government committed to carry out trialling activities on a number of different approaches aimed at improving retirement savings for self-employed people, drawing on the learnings from the success of using a default saving approach in automatic enrolment. The objective is to use these trialling activities to inform and develop the evidence base, in order to identify effective policy interventions which can then be tested at scale in future.

3. In developing our approach to scoping and developing trialling activities we have had regard to The Public Sector Equality Duty. Alongside this, the government’s approach is based on a set of core design principles, which includes using learning from the success of automatic enrolment where relevant to improve retirement saving coverage and savings levels among the self-employed.

4. This report describes the range of approaches that the government has prioritised to deliver a series of interventions that will help build the evidence base around what works to support the self-employed to save for retirement. Chapter 1 describes the key principles underpinning our approach in developing the framework for trialling and the main areas in which to focus this trialling activity. The key principles are:

• increase participation in retirement saving for the self-employed
• remain open-minded as to whether pensions alone are appropriate for self-employed retirement saving
• capitalise on existing points of contact which provide prompts into saving
• improve consumer outcomes, including through better use of financial technology
• support the development and accessibility of good quality products

5. In addition, the chapter outlines the programme of trialling activities which will aim to test a range of approaches in relation to joining a pension scheme or retirement savings vehicle, and facilitating and making regular saving into pensions or other savings vehicles. Trials will focus on the following three key areas:

• movements between employment and self-employment – specifically, this includes encouraging individuals who move from employment to self-employment to continue to save, building on the foundation they have built through automatic enrolment
• points of contact – maximising the existing communications and systems used by and surrounding self-employed people to facilitate or create nudges towards saving (for example, this may include accounting systems, banking apps)
• accessibility of savings – assessing self-employed behaviours when the option of greater access to liquid savings is available, alongside a pension or long-term savings product

6. The department will work with a range of delivery partners, including NEST Insight, IPSE, Smart Pension, Aegon, Aviva plc, Lloyds Banking Group plc and Barclays plc on a range of evidence gathering and trial activities. We are also seeking additional organisations to work with government in specific sectors (including payment systems, for example).

Self-employment in the UK – current experience and trends

7. Chapter 2 provides an overview of the current landscape of the self-employed across the UK and the evidence around modern working relationships. It describes the composition of the self-employed, including earnings, sectors, age, gender and educational attainment. It also discusses the government’s plans for improving the clarity of employment status.

8. The chapter highlights changes in the self-employed population over time, including: high levels of net growth in part-time working in the self-employed, particularly for males; a higher net growth rate of women entering self-employment in recent years; and higher levels of growth than previous years in more professional occupations (for example, advertising, IT and banking) and those educated to degree level.

Self-employment: savings behaviours and labour market experiences

9. Chapter 3 provides information about pensions and savings assets amongst the self-employed. While pension participation has fallen for the self-employed in recent years, total assets of the self-employed are broadly comparable to those held by employees. In addition, the percentage of people contributing to a pension broadly increases with age to age 50 to 59 before falling again for employees and the self-employed. Pension assets also broadly increase with age, but at a much greater rate for employees than for the self-employed. Similarly, while the likelihood of pension participation for employees is broadly similar for all employees earning over £35,000 per annum (around 85%), the likelihood of saving into a pension amongst the self-employed broadly increases with income – but at a much lower rate.

10. The chapter also highlights differences in attitudes to saving, including a perception that investing in property was the safest way to save for retirement and a preference for alternative forms of saving to a pension amongst the self-employed compared to employees. Flexibility in access to pension or savings products was highlighted as important in a number of studies. The report also summarises a number of examples of Fintech products that offer support to individuals to save by making saving easier.

11. In exploring labour market patterns, new evidence highlights that the majority of people (55%) enter self-employment for the first time from employment. A higher proportion of men entered self-employment as a first job (20% men; 5% women) and a higher proportion of women entered self-employment having had caring responsibilities as their main activity prior to that (15% women; 0% men).^{[footnote 4]}

12. The average number of employments for the self-employed in a lifetime was marginally higher than that of employees – the self-employed had a mean number of between seven and ten jobs over a lifetime compared to between six and nine jobs for employees. The mean number of jobs in a lifetime was also slightly higher for females compared to males: males with self-employment as their main activity in one year had, on average, between seven and nine jobs compared to self-employed females who had between seven and 11 jobs over a lifetime. Examples of possible points of contact that the self-employed use or interact with include: accounting or invoicing software packages, online platforms such as, invoicing systems or banking systems, the self-assessment system, digital trading platforms, the GOV.UK website and online sites for advertising their services.

A framework for a trialling and research programme

13. Chapter 4 describes the programme of trials and how they aim to test a range of approaches in relation to encouraging the self-employed to start saving, either through short-term savings vehicles, such as cash ISAs and / or into a pension scheme (or an alternative long-term savings vehicle), or to re-start contributions if these have lapsed. The trials will gather evidence around messaging and the impact of systems and default mechanisms, such as setting up a contribution from income on a pre-committed basis, to drive voluntary adoption of savings behaviours. The chapter provides details of delivery partners and timeframes for delivery of trials which begin from winter 2018.

Next Steps

14. Chapter 5 describes how the Department will set up, run and monitor a programme of trialling activities as detailed in chapter 4. It also includes a call for action for additional delivery partners from a range of sectors including, invoicing software providers accounting organisations and self-employed work spaces. The government will facilitate a design workshop, working with a range of organisations that provide services to the self-employed and financial services, in early 2019. Alongside this, we will continue to engage with interested parties during 2019/20 to share findings and seek input and to inform the programme of activity.

Chapter 1: Introduction

1. The UK self-employed population is a diverse group, encompassing an estimated 4.75m people (around 15% of the workforce)^{[footnote 5]}. Pensions and savings coverage among the self-employed population varies considerably. For some it is low – particularly, as was the case with employees before automatic enrolment, among low to moderate earners and people below age 40 years while amongst some – often older, and more affluent – self-employed people, retirement saving is more common.

2. However, overall pension participation among the self-employed continues to decline over time. Analysis from the Family Resources Survey^{[footnote 6]} suggests that 14% of the self-employed were saving into a pension in 2016 to 2017; this has decreased from 30% in 2007 to 2008. It is important to recognise, however, that the self-employed have comparable total assets to employees, but a greater proportion of self-employed savings are in property rather than pensions.

Government commitment to trialling interventions

3. The government’s 2017 manifesto included a commitment that:

We will continue to extend auto-enrolment to small employers and make it available to the self-employed.

As part of its 2017 Review of Automatic Enrolment the Department recognised that automatic enrolment could not therefore be straightforwardly extended to this group.

4. The government committed to use behavioural insights from automatic enrolment to help inform the development of policy approaches that can simplify and potentially automate saving for self-employed people. Given the significant diversity of self-employment and limited evidence base, the Review committed to work towards the manifesto commitment through carrying out trialling activities, working with a range of delivery partners, to establish what does or does not work, before testing proposals at scale. A similar approach was used to test the efficacy of different savings nudges prior to the introduction of automatic enrolment.

5.To assess the most effective approaches, we are continuing to develop our evidence on the savings behaviours of the self-employed, recognising that this differs within and across cohorts given the different and changing patterns of movement between self-employment and employment.

Principles underpinning our approach

6. This report describes the range of approaches that the government has prioritised, including how and why these have been identified and how they are being developed, having ongoing due regard to section 149 of the Equality Act^{[footnote 7]} (The Public Sector Equality Duty). Our approach to the development of the trials has been underpinned by core principles, informed by extensive discussions with stakeholders, in addition to the principles set out by the Pensions Commission of fairness, affordability and sustainability.

Core principles, which provide the foundation of our trial programme:

• increase participation in retirement saving for the self-employed and encourage saving at levels which are affordable, using learning from the success of automatic enrolment
• remain open-minded as to whether pensions alone are appropriate for self-employed retirement saving, in particular given the irregularity of earnings levels
• capitalise on existing points of contact which provide prompts into saving in a way that is easy to understand and is as simple as possible to implement
• improve consumer outcomes, including through better use of financial technology, to support the self-employed to overcome current barriers to saving
• support the development and accessibility of good quality products that suit the specific needs of the self-employed i.e. exploring the option of greater access to liquid savings, alongside a pension or long-term savings product

Framing trialling activity – building resilience for future generations

7. Our priority is to encourage pension saving amongst self-employed young people below age 35 years (Millennials) and those aged between 35 and 49 years (Generation X). While the evidence points to particular groups, our approaches will also apply to other age groups, as we would want to ensure that all those capable of saving for retirement are encouraged and supported to do so. Our aim is to support the self-employed to start saving from a younger age and to normalise this behaviour so that it is sustained. We recognise that a significant proportion of younger people may enter self-employment for the first time following a period of employment where they may have been automatically enrolled into a workplace pension. We want to help encourage good savings behaviours, so that individuals keep saving into a pension.

8. However, in order to ensure the widest possible learning, at this initial stage the trialling activities will not exclusively focus on any specific groups of self-employed individuals. We will however assess the efficacy of our trials on self-employed people with different characteristics and may consider more targeted trialling in due course. It will be important to ensure that any groups targeted through the interventions reflect the self-employed population and have an appropriate split of men and women and protected groups. The Department will be working with the cross government Trial Advice Panel^{[footnote 8]} and Ethics Advice Panel to ensure the robustness of the approach and that any intervention is ethically sound.

Programme of trialling activity – from 2018

9. The trialling activities will test a range of approaches in relation to joining a pension scheme or retirement savings vehicle, and facilitating and making regular saving into pensions or other savings vehicles. They will also gather evidence about how messaging around pension savings is perceived and understood by the self-employed as well as the measurable impact of different approaches. There have been limited systematic or formally evaluated studies seeking to test the effectiveness of communication interventions to increase voluntary long term saving amongst the self-employed. In developing the trialling activities there is an opportunity to test more traditional marketing interventions and to test variants that rationalise the reason to save for the future, such as focusing on family rather than on the individual.

10. Beyond testing communication prompts we are keen to understand what steps lead to regular saving for self-employed people. We want to test the effectiveness of a pre-commitment to save at certain points in the management of day-to finances, ideally when income is received, similar to the way that the process of pension saving is embedded in payroll for automatic enrolment. For the self-employed we are exploring using systems that will work towards achieving a similar outcome, for example methods of creating a pre-commitment to make a contribution to a pension, or other savings vehicle, via payment or accounting systems. In this way, the behavioral learnings from automatic enrolment can be applied to and tested against the different financial journeys that self-employed people undertake.

11. The main trial areas and delivery organisations are detailed below:

National Employment Savings Trust (NEST) Insight

The Department for Work and Pensions (DWP) and NEST Insight will commence a research and trialling programme to deliver a range of trialling activities commencing late 2018 and during 2019 to 2020. The initial trials will focus on testing whether certain types of messaging or marketing interventions can increase the propensity of the self-employed to save into a pension. Later trials will build on the findings and test the scope to make it easier to prompt and/or facilitate contribution through existing systems which self-employed people use, for example through invoicing services or accounting software.

Aviva plc

Following development work in 2018, from early 2019, Aviva will run a series of pilots promoting pension saving to the self-employed. The pilots will test, across a number of different intervention groups, how best to engage the self-employed with the benefits of pension saving. These campaigns will take place across the first half of 2019 with some starting from February 2019. The outcomes will be monitored on an ongoing basis throughout 2019. The results and learnings of the pilots are expected to be available from the second half of 2019.

Lloyds Banking Group plc

Lloyds Banking Group are working with the Department to carry out primary research in relation to self-employed people in their customer base to test reactions to potential policy interventions. During the first half of 2019 the Group will carry out qualitative and quantitative research to test reactions to approaches aimed at making it easier to save – for example, facilitating saving from a bank account into a pension scheme; and through a facility on the banking app that makes it easy for savers to make one-off payments. Lloyds Banking Group are working closely with the Department on the design and framing of this activity. This will then help to inform concept testing through a further series of focus groups with some of the Group’s self-employed customers, scheduled for later in 2019.

Barclays plc

Barclays plc is carrying out research in relation to their self-employed customer base, working closely with the Department on how this is framed and evaluated, to explore savings behaviours. The anonymised research, which will be undertaken from early 2019, will focus in particular on self-employed customers across Barclays Business Bank and Personal Bank. The aim of this research is to enhance our understanding of savings, including retirement savings, habits and motivations among this population who access Barclays financial services. Barclays plc envisage using outputs to help inform and support the Department’s evidence-based trialling programme.

DWP is also working with The Association of Independent Professionals and the Self-Employed (IPSE)

IPSE, Smart Pension, Aegon and NEST Insight

IPSE (the Association of Independent Professionals and the Self-Employed), Smart Pension and Aegon are working with NEST Insight to develop a research trial that will find new ways to get more of the self-employed saving for retirement. Researchers will test a variety of communications with IPSE members. IPSE, Smart Pension and Aegon will work to identify what framing devices and messages most effectively encourage the self-employed to save for retirement. Researchers will then be able to track their user journeys (for example the time spent on pages, click-through rates and the saving products purchased). From this, IPSE, Smart Pension, Aegon, and NEST Insight researchers will get an insight into the best ways to encourage IPSE members and the self-employed to save for later life. In this trial, IPSE and Smart Pension are working together to develop and test a new, holistic saving solution which might complement traditional pensions. The new solution, which Smart Pension will develop, will allow self-employed people an additional means to save for short, medium and long-term goals like retirement.

We welcome opportunities to extend the reach of the trials, and we welcome other partners’ involvement in delivering the programme.

12. Learnings and insights from the trials will be identified and made available on a rolling basis. The Department is committed to making information publicly available, working with delivery organisations to ensure the monitoring and evaluation is robust.

Delivery partners – trialling and finding solutions

13. The successful development and delivery of automatic enrolment is grounded in the role played by a huge range of partners including employers, intermediaries and the pensions industry. A similar approach will be needed in relation to the self-employed.

14. The self-employed population is spread across a complex landscape involving a wide variety of sectors and points of contact from recruitment services, to accountants through to market places and shared work spaces. It is likely that a broad coalition of partners (including unions, trade associations, the wider financial services sector and others) will need to work together to develop and implement effective and durable solutions.

15. We have already carried out extensive engagement work with a broad range of organisations, including service providers to the self-employed and their representative bodies and financial services organisations. In addition, the Department has worked closely with key industry bodies, in particular the Association of British Insurers (ABI), the Pensions and Lifetime Savings Association (PLSA) and the Federation of Small Businesses (FSB). Our approach has also benefitted from discussions and input with think tanks, for example including Demos and the Resolution Foundation.

16. As part of this engagement work, DWP and HM Treasury delivered, in partnership with the Association of British Insurers (ABI), a TechSprint^{[footnote 9]}. This was a two-day technology-based idea creation event that looked at how technology could be used to make it easier for the self-employed to build up long-term savings. The event confirmed the challenges and opportunities of finding a pension saving solution for the self-employed. Many of the solutions proposed by participants were based on working within the current legislative/regulatory framework, and using or adapting existing income smoothing platforms. It also suggested the potential efficacy of highly targeted solutions applied through the systems (including financial services) that the self-employed themselves engage with; and the scope for technology based solutions to address the challenge in new and innovative ways.

Call to action

17. The trialling activities outlined in this document will begin soon. DWP will conduct internal monitoring of the NEST trials and review the outputs (for example, numbers of individuals prompted to access their accounts) and outcomes (such as changes to pension participation rates and amounts saved) from the trialling activities. We will also seek to build in opportunities for those involved in the trials to provide feedback on how they found the intervention. If the findings justify progression to full scale trials, we will seek to evaluate any subsequent trialling activity.

18. This report provides a framework for initial trialling activities (at Chapter 4). The Department is committed to continuing to engage, work with and where possible form delivery partnerships with organisations that can help increase retirement saving among the self-employed.

Call to action for trial partners:

• chapter 4 summarises the trial areas we are continuing to develop. This includes development of larger-scale trials executed through partnerships with existing commercial service providers to self-employed groups
• we would welcome the opportunity to connect with organisations from a range of sectors including: invoicing software providers; accounting organisations and self-employed work spaces. If you are interested, please email us: self-employment.deliveryplan@dwp.gov.uk

Chapter 2: Self-employment in the UK – current experiences and trends

This section gives a brief overview of the current landscape of the self-employed across the UK, and also considers the evidence around the gig economy^{[footnote 10]}. However, the transitional nature of this work, coupled with the significant number of gig workers who potentially already fall within the scope of automatic enrolment, suggests that the focus of the government-led trialling activities should be on the traditional self-employed population.

Self-employed population and developing trends

1. The self-employed are a highly diverse and evolving group which has seen significant growth. The number of self-employed increased from 3.3 million people (12.0% of the labour force) in 2001 to 4.75 million (14.7% of the labour force) in 2018^{[footnote 11]}.

2. As the 2017 Review noted there is no single agreed definition of self-employment. The three main approaches use the definitions of employment law, tax law and self-definition. There are three employment statuses within employment law – employee, worker, or self-employed. People whose employment law status is that of an employee or a worker should be auto-enrolled as long as they meet the other eligibility conditions, including those relating to age and earnings.

3. Generally, self-employment offers people more economic independence than those working under a contract of employment. When considering activity for trialling activity, we have focused on the more traditional forms of self-employment, where an individual or group of proprietors runs their own business. This will include businesses constituted as:

• sole proprietors – an individual who works on their own account (for example a plumber who works as a sole trader)
• partnerships – two or more people who own and run their business (for example, consultancies or accounting practices)
• limited company – a registered company, usually a private company limited by shares. Typically, a self-employed person will be the sole shareholder, or one of a limited number of shareholders, giving them significant control over the company; will be a director of the company; and will actively participate in the day-to-day running of its business

Self-employed – Key facts

Self-employed are a highly diverse and evolving group:

• 4.75 million individuals classify themselves as self-employed (14.7% of UK workforce)
• traditional occupations include building and construction related roles, hairdressing and caring and more recently we have seen growth particularly focused in professional occupations such as IT, banking and advertising
• earnings growth is lower among the self-employed than in the employed workforce
• self-employed people are typically older than members of the employed workforce
• there is higher net growth in women entering self-employment in recent years

4. In December last year the government published the Automatic Enrolment Review 2017: Maintaining the Momentum^{[footnote 12]} which summarised the evidence with regards to the self-employed population and their behaviours and attitudes towards saving for retirement. More detailed analysis, particularly with regard to transitions between self-employment and employment, was published in the Automatic Enrolment Review 2017: Analytical Report^{[footnote 13]}.

5. This document provides a further update on existing and new evidence made available since December 2017. It describes the way that the growing evidence base has informed the range of approaches and priorities which are shaping our plans. The Annex includes more detailed findings used to underpin the trialling framework.

Labour market sector analysis

6. The diverse nature of the self-employed is reflected across the sectors in which they work, their average age, earnings and educational attainment (See Table A1 of Annex A for recent trends in the growth in numbers of self-employed people). There has been a growth in levels of self-employed part-time work in recent years: the number of male part-time self-employed workers increased from 273,000 to 569,000 between 2001 and 2016, while female part-time self-employment increased from 439,000 to 812,000 over the same period. Evidence also suggests that some part-time self-employed workers, particularly men, are also in receipt of pension and retirement income – around a third of total income for male part-time self-employed workers and 10% of total income for female part-time self-employed workers came from pension and retirement income. A much smaller proportion of total income of part-time self-employed workers came from employment – 2% of total income for females and 1% of total income for males (See Table A2 of Annex A).^{[footnote 14]}

7. While there remains a core of traditional trades such as construction and plumbing, much of the recent growth in self-employment has been those entering self-employment in more professional occupations such as advertising, IT and banking. The self-employed person is typically older than the employed worker with the average age being 47 years old, which is seven years higher than that of the average employee. They are also generally older when they first become self-employed at an average age of 32.^{[footnote 15]} While the self-employed are still more likely to be male overall, there has been a higher net growth rate of women entering self-employment in recent years^{[footnote 16]}.

8. The most recent ONS data Trends in Self-Employment in the UK^{[footnote 17]}, provides further insight on the composition of the self-employed, in particular focusing on earnings and education. It found that earnings growth was, on the whole, greater for employees compared to the self-employed (See Annex A, Table A3). In addition, previous PPI analysis of the Wealth and Assets survey found that men who have low levels of educational attainment were more likely to enter self-employment at a younger age than those with degrees; in comparison, self-employed women were more likely to be educated to degree level than men.^{[footnote 18]}

Gig economy and employment status

Alongside the growth in self-employment, the UK, much like other countries, has seen the rise of the ‘gig economy’ and platform working. As set out in the 2017 Review of Automatic Enrolment, a large percentage of those working in the gig economy potentially already come within the scope of the automatic enrolment framework, with employers having legal duties to establish and contribute to a workplace pension for qualifying individuals.

The Chartered Institute of Personnel and Development (CIPD)^{[footnote 19]} estimates that in 2016 to 2017:

• approximately 1.3 million people (4% of all those in employment) were working in gig economy roles of these, approximately 200,000 (15%) identified themselves as self-employed
• over half of gig economy work is being undertaken by someone with an existing, full-time, job
• 34% of gig workers were aged 16 to 30 compared to 11% of self-employed workers or 26% of employees
• half of gig economy respondents had been active in their current role for a year or less
• a fifth of respondents had only been engaged for up to three months
• the length of time reported in gig work was similar across age groups and by gender

In addition, the government’s December 2018 Good Work Plan^{[footnote 20]} sets out the vision for the future of the labour market and ambitious plans for implementing the recommendations arising from the Taylor Review of Modern Working Practices. The government commits to legislate to improve the clarity of the employment status tests, reflecting the reality of modern working relationships.

We will ensure any changes are also considered in relation to automatic enrolment so that there is coherence and clarity for individuals and businesses about who is eligible for automatic enrolment.

Chapter 3: Self-employed: savings and labour market behaviours

We have used this analysis to help inform and construct a framework for the trialling activities. We will continue to develop the evidence base, working with partner organisations, including to refine our assessment of whether and how interventions or support can be suitably segmented to different groups within the self-employed population.

This section provides information about pensions and savings assets among self-employed people, including how asset accumulation varies for different cohorts. It also provides detailed information about labour market behaviours, with evidence and analysis concerning transitions into and the duration of periods spent in self-employment.

Drawing on the evidence base, our trialling programme is focussed on:

• testing engagement opportunities during transitions into self-employment both from employment and/or caring responsibilities
• maximising systems and contact points that self-employed people use – including payment systems; invoicing software and banking services
• product accessibility and suitability – given income volatility, testing the role of shorter term saving as a means of paving the way for transitioning to retirement savings

Savings behaviours among self-employed people

Total assets held by self-employed

1. Pensions and savings coverage among the self-employed varies considerably. While some, mainly older self-employed individuals have reasonably good provision for later life, pension participation among the self-employed population continues to decline over time.^{[footnote 21]} DWP replicated the analysis conducted by Pensions Policy Institute (PPI) as part of their ‘Policies for increasing the long-term saving of the self-employed’ publication using the latest Wealth and Assets survey data 2014 to 2016^{[footnote 22]}. Figure 3.1^{[footnote 23]} shows that the trends from earlier analysis continues: while the self-employed have comparable total assets to employees, a greater proportion of their savings are in property rather than pensions.

2. The updated analysis has confirmed that there is a ten-year lag in assets accumulation for self-employed compared to employees. While we know that assets are currently broadly comparable for self-employed and employees for each generation, we do not yet know how changes in property ownership rates may impact on overall assets for younger generations. A decline in property ownership would have an impact on overall assets for the self-employed, potentially making it more important to enable self-employed individuals to start saving from a younger age. Furthermore, given automatic enrolment is still relatively new (having been introduced in 2012), the data will not yet show the extent to which automatic enrolment has and can have an impact on the pension accumulation of young people. This could therefore result in a greater difference between total asset levels for employees compared to self-employed and also potentially makes it more important to enable self-employed individuals to start saving from a younger age.

Figure 3.1: Total assets by generation for self-employed and employees – 2014 to 2016

Generation	Property wealth	Pension wealth	Financial wealth	Physical wealth
Millennials (15 to 34) – self-employed	44%	10%	2%	44%
Millennials (15 to 34) – employee	37%	27%	1%	34%
Generation X (35 to 49) – self-employed	59%	14%	3%	25%
Generation X (35 to 49) – employee	41%	37%	2%	20%
Baby boomers (50 to 69) – self-employed	55%	26%	6%	13%
Baby boomers (50 to 69) – employee	36%	50%	4%	11%
Silent generation (70+) – self-employed	63%	13%	10%	14%
Silent generation (70+) – employee	54%	37%	2%	20%

Source: DWP analysis of the Wealth and Assets Survey 2014 to 2016

3. As automatic enrolment savings accumulate amongst the employed there is potential for a greater discrepancy between self-employed and employee pension asset levels in the future as more people are brought into automatic enrolment and as contribution rates are increased.

Pension participation

4. Now that participation in workplace pension saving has grown, largely as a result of the introduction of automatic enrolment, the self-employed are much less likely to be actively contributing to a private pension than employees, which is shown by overall pension participation statistics^{[footnote 24]}. The proportion of people saving into a pension was consistently lower amongst the self-employed for each age group when compared to employees (Figure 3.2) but the likelihood of pension participation^{[footnote 25]} increased with age up to age 60 for both groups. Pension participation was highest for those aged 40 to 49 and 50 to 59 in the self-employed (37% for both age groups) and the employed (79% for both age groups).^{[footnote 26]}

Figure 3.2: The percentage of people actively contributing to a private pension by age, 2014 to 2016

Age	Self employed (percentage)	Employees (percentage)
16 to 29	0%	42%
30 to 39	21%	73%
40 to 49	37%	79%
50 to 59	37%	79%
60 to 64	25%	63%
65 and over	8%	27%

Source: Wealth and Assets Survey, 2014 to 2016^{[footnote 27]}

5\ . However, as shown overall in Figure 3.3, the median amount of assets held in private pensions was markedly lower for the self-employed compared to employees, particularly in the 50 to 59 and the 65 and over age brackets.

Figure 3.3: Median amount of wealth held in pensions by age, 2014 to 2016

Age	Self employed	Employees
16 to 29	£0	£6,000
30 to 39	£10,000	£20,000
40 to 49	£24,000	£50,100
50 to 59	£27,000	£91,000
60 to 64	£50,000	£80,100
65 and over	£36,500	£108,000

Source: Wealth and Assets Survey, 2014 to 2016^{[footnote 28]}

6. Figure 3.3 shows that while the median pension assets of the self-employed reached a peak of £50,000 in those aged 60 to 64 before dropping again in the 65 and over group, median pension assets in the employed population peaked in the 65 and over group, with median pension assets of over double that of the self-employed peak median pension assets at £108,000.^{[footnote 29]} The reasons for the decrease in pension assets in the self-employed are not known: some individuals may be utilising pension assets during retirement and/or during periods spent in self-employment from age 60 or it could be due to a compositional effect, for example if people are entering self-employment at later stages in life with lower levels of saving and are using self-employment to add to their retirement income.

7. Overall, this shows that for the self-employed and employees, pension assets broadly increase with age but at a much greater rate for employees than for the self-employed. Median levels of pension asset growth between the ages of 40 and 59 is notably slower for the self-employed compared to employees: it is not clear what is driving this, although it is possible that the median does not increase markedly as, as already noted, people entering self-employment do so at later ages and may have existing lower levels of pension provision and assets. In addition to the likelihood of pension participation and median pension assets increasing with age, analysis of the Wealth & Assets survey also highlighted that the likelihood of saving into a pension broadly increases with income for the self-employed and for employees (Figure 3.4).^{[footnote 30]}

Figure 3.4: The percentage of people actively contributing to a private pension by income, 2015 to 2016

Income	Self employed (percentage)	Employees (percentage)
£0 but less than £10,000	27%	31%
£10,000 but less than £20,000	21%	58%
£20,000 but less than £35,000	25%	75%
£35,000 but less than £50,000	30%	85%
£50,000 but less than £65,000	34%	83%
£65,000 but less than £90,000	33%	86%
£90,000 or more	42%	85%

Source: Wealth and Assets Survey, 2015 to 2016^{[footnote 31]}

8. Figure 3.4 highlights that while the likelihood of pension participation is broadly similar for all employees earning over £35,000 per annum at around 85%, the likelihood of saving into a pension amongst the self-employed increases with income but at a much lower rate. The data show that pension participation and pension assets are associated with age and income in the self-employed – older people and those who earn more are more likely to save into a pension and are more likely to have higher levels of assets (although at a significantly lower level than their employed counterparts).

9. This, combined with earlier findings from the PPI^{[footnote 32]} that highlighted that the self-employed experience a ten-year lag in the accrual of their savings, and findings from Her Majesty’s Revenue and Customs (HMRC)^{[footnote 33]} that younger people with lower incomes and irregular cash flows were least likely to be saving into a pension (10% of those categorised as ‘Younger and Unprepared’ in their sample were saving into a pension) suggests that the most impactful interventions may be those targeted at younger people in self-employment (and/or those young people likely to enter self-employment) who are part of the ‘Younger and Capable’ cohort^{[footnote 34]}, those who have the capacity to save but are not currently preparing for retirement. A further target group may be those with lower earnings, to encourage saving from a younger age/lower income bracket to try and bring self-employed saving and pension assets more in line with that of employees.

Attitudes to saving

10. In addition to exploring the savings behaviour of the self-employed, we have also considered their attitudes to saving. The main reason, cited in the Wealth and Assets survey, that self-employed gave for not contributing to a pension was that they thought they could not afford to do so, (22%) while 46% thought that investing in property was the safest way to save for retirement. The self-employed were also more likely than employees to say that they preferred alternative forms of saving to pensions (14% self-employed compared to 6% employees)^{[footnote 35]}. Details of research carried out by MORI IPSOS on behalf of HMRC are in the Annex, ‘Saving for Retirement amongst the self-employed’.

11. A recent report by IPSE found that almost a third of those surveyed said that a flexible pension / liquid saving solution that was tailored to the self-employed would best encourage them to save. Qualitative research by NEST Insight also highlighted a preference for saving products that offered liquidity or flexibility. This was again supported by NEST Insight’s recently published report Working for Today, Preparing for Tomorrow^{[footnote 36]} which suggested that saving mechanisms for the self-employed needed to be easy to understand, with flexible access, enabling the self-employed to be in control of their savings.

12. The 2017 Automatic Enrolment Review highlighted that there are already a number of pension and long term savings products available with a reasonably high level of accessibility for those who would wish to save into them. Recognising the demand for alternative forms of saving, and the wish among some groups to be able to access savings when needed “for a rainy day”, there are already signs of innovation in the market, although not specifically targeting the self-employed, these would be available to them if they meet the relevant eligibility conditions.

Access to liquid funds

The government’s Help to Save scheme launched in September 2018. Help to Save supports working families on low incomes to build up a rainy-day fund by offering a 50% bonus on up to £50 of monthly savings. It has been designed to encourage a regular savings habit, helping people to build a financial buffer to protect against unexpected costs and income shocks. The first bonus will be paid after two years. Customers can continue saving for a further two years and receive a second bonus. This means that individuals can save up to £2,400 in the account and benefit from government bonuses worth up to £1,200. The scheme is open to adults in receipt of Working Tax Credit, or those in receipt of Universal Credit with minimum household earnings equivalent to 16 hours a week at the National Living Wage, currently £542.88 a month.

The Lifetime ISA may be appropriate for those wishing to save for the long-term, but for whom workplace pensions are unavailable, such as the self-employed. The Lifetime ISA is a long-term savings product intended to support younger people saving for their first home, or for later life. Individuals aged 18 to 40 can open an account and save up to £4,000 each year until they reach 50. The government provides a 25% bonus on all contributions within these limits. Savings invested in a Lifetime ISA can be withdrawn without charge from the age of 60, or for the purchase of the account holder’s first home (provided the individual takes out a mortgage and is buying a property worth up to a maximum of £450,000). Savings can also be withdrawn if the individual develops a terminal illness before age 60. All other withdrawals incur a 25% government charge. The charge is made to reflect that the account is intended to be used for the specific purposes outlined above and held over a long term. This means that the product can be used flexibly, with customers able to access the vast majority of their savings and growth when they need them. The charge recoups the government bonus and any interest or growth arising from it.

NEST Insight has been working with Harvard Kennedy School and the Money Advice Service to understand and solve the challenges facing people who have little or no liquid savings put aside for emergencies. NEST Insight’s ‘Sidecar Savings’ trial is an academic research project, looking at how people behave when given the option of having a liquid savings product alongside their workplace pension, and how this impacts their financial wellbeing. To use the sidecar savings model, employees must pay in an amount over and above the auto enrolment minimum contribution level. A worker’s normal pension contributions continue as usual, with the additional contributions initially flowing into the emergency savings account. If the emergency account hits a certain level, those extra contributions will then start ‘rolling’ into the pension. The idea is to help workers build up an emergency buffer of short-term savings, whilst also helping them to save more for later life if the savings cap is reached. The NEST Insight sidecar savings trial was launched on 12 November 2018. Timpson is the first employer to roll out the trial within their organisation of over 5,600 workers.

13. In addition to innovations in the market in relation to savings products, there is a growing market of Fintech organisations using app based approaches to enable individuals to invest through savings vehicles – a few organisations that we have engaged with include:

• Money Box^{[footnote 37]}
• Plum^{[footnote 38]}
• Chip^{[footnote 39]}
• Portify^{[footnote 40]}
• Trezeo^{[footnote 41]}

Many of these organisations allow for someone to pre-commit a proportion of their income or payments into savings vehicles. We are continuing to explore the opportunities provided by this type of technology to understand if and how it may help to improve access to saving among the self-employed.

14. In international terms^{[footnote 42]}, the UK ranks above the USA, Canada, Australia and Germany in terms of self-employment as a proportion of the overall workforce, though the UK rate is below the EU average of 16.1% (the UK rate is 15% of total workforce). Countries with a large self-employed population are: Colombia (51.3%), Greece (34.1%), Brazil (32.9%) and Turkey (32.4%). These countries use a mixture of Fintech, behavioural science, social security contributions and a wide range of pension products to encourage the self-employed to invest (emotionally and financially) in their future. In Australia there are approximately 13 million people (OECD, 2018) in employment, with the self-employed accounting for 10.1% of the Australian workforce. The gig economy in Australia represents roughly 100,000, (0.8%) of their workforce, with growth continuing to be strong. GigSuper^{[footnote 43]} is an example of an Australian private sector pension scheme offer for the self-employed. As part of our work on developing our understanding of the diversity of self-employment, we continue to explore international evidence, what pension products the self-employed have available and how other countries are tackling under saving for retirement among the self-employed.

15. Alongside market and product innovations, through our engagement with organisations, we recognise that there are opportunities for service providers of the self-employed such as trade bodies and trade unions to engage with their members in relation to pension savings. Some organisations, such as IPSE, already offer access to a pension for their members. Others, such as FSB, provide access to a pension scheme for their self-employed members where they act as an employer for automatic enrolment purposes, while some trade unions such as Unite provide fact sheets and signpost their members to appropriate guidance.

16. While there appears to be no standard approach, consistent with individual organisations’ decisions on appropriate offers for their members, we are continuing to work with relevant organisations to understand opportunities to share best practice – including learning from the trialling programme. Our aim is that this will then help inform and drive awareness and access to retirement saving among those who provide services to self-employed individuals.

Labour market patterns

17. Evidence presented in the 2017 Automatic Enrolment Review highlighted how the majority of the self-employed had previously had at least one year where employment was their main activity^{[footnote 44]}. We have expanded this evidence using Lifetime Labour Market Database (L2) data^{[footnote 45]}, to explore the previous employment status of those entering self-employment for the first time and the average number of jobs held by self-employed compared to employees over a lifetime.

This evidence has implications for our self-employed trials. Given that the majority of people move into self-employment from employment and that employed people have, on average between six and nine jobs in their lifetime or between seven and ten jobs for the self-employed^{[footnote 46]}, a growing number of people working in self-employment will, at some point in their career, have been automatically enrolled into a workplace pension while in employment (see Figure A1 of the Annex for more detail on time spent in self-employment). The question is, therefore, how we support those who have been employed to continue to save into a pension once they become self-employed. This will be explored through the trialling framework outlined in Chapter 4.

Movement between self-employment and employment

Figure 3.5: Previous employment status for people entering self-employment for the first time, 2016

Gender	Employed	Care	First job	Not active	Benefits	Other
Male	55%	0%	20%	5%	15%	5%
All	55%	5%	15%	5%	15%	5%
Female	55%	14%	5%	5%	15%	6%

Source: DWP analysis of the Lifetime Labour Market Database (L2)

18. Figure 3.5 shows that the majority (55%) of all first-time self-employed people have previously been in employment immediately prior to self-employment. While there was little difference between men and women who became self-employed following employment, there were other differences observed. Figure 3.5 highlights that a higher proportion of men entered self-employment as a first job (20% men compared to 5% women) and that 15% of women compared to 0% of men entered self-employment having had caring responsibilities as their main activity.^{[footnote 47]} Further analysis also highlighted that women were most likely to enter self-employment from caring activities between the ages of 30 to 39 or 40 to 49.

19. Women entering self-employment from caring responsibilities potentially has implications for our trials. For example, the GOV.UK website could be a potential touch-point for female carers entering self-employment when claiming caring benefits. In order to understand more about the self-employed population and transitions between employment and self-employment, analysis was conducted to explore the average number of jobs worked in a lifetime, split by whether somebody has been self-employed as their main activity at any point in their career. The average number of employments^{[footnote 48]} worked for all those included in the sample was between 6 and 9 jobs, compared to a slightly higher figure for the self-employed of between seven and ten jobs.

20. Women tended to have a higher mean number of employments than men: males with self-employment as their main activity in one year had, on average between seven and nine jobs compared to self-employed females who had between seven and eleven jobs over a lifetime.

Points of contact/touch points for the self-employed

21. It is important to understand how the self-employed interact with systems and institutions to help highlight where possible interventions could best be placed or targeted (including the benefits system, see Figures A2 and A3 of Annex A) We therefore would like to use the trials to test the efficacy of different channels and touch points. By touch points, we mean places, organisations, software etc. that the self-employed use or with which they interact. We have done work to identify the most potentially useful touch points and will be prioritising these through our trialling activity. Qualitative Research by NEST Insight identified a range of possible touch points^{[footnote 49]} for engaging with the self-employed that may provide an opportunity to prompt people into saving – for example by using:

• accounting or invoicing software packages
• online platforms such as digital platforms, invoicing services or banking systems
• tax self-assessment system^{[footnote 50]}

22. Other possible touch points, not explicitly referred to in the NEST research, include digital trading platforms such as Etsy or eBay, the GOV.UK website (for example, for advice on setting up a business), online sites for advertising their services (for example, Facebook, Upwork etc.) or payment systems using direct deductions at the point of purchase. It will be important to consider how interventions can build upon these existing touch points for the self-employed.

23. The high level of churn that we have found between employment and self-employment, suggests that many self-employed people will, at some stage in their working life, have an opportunity to contribute to a workplace pension through automatic enrolment. Therefore, we would expect a greater proportion of future self-employed individuals to have accumulated some private pension assets than in the past. The challenge, however, is in prompting the self-employed to remain saving when they move to self-employment. This is an area we will focus on when developing the trialling framework detailed in the next section.

24. However, we recognise that the point at which somebody moves from employment into self-employment may not be the optimal time to intervene and encourage continuation of existing pension saving (for example, given the possible uncertainty over future earnings, initial costs associated with setting up a business). We are therefore also keen to test/vary timings of interventions for those moving from self-employment to employment.

Chapter 4: A framework for a trialling and research programme

This section sets out the programme of trials. In developing our approach to testing, we have had regard to the principles and learning from automatic enrolment as well as the evidence base set out in Chapter 3 and will adopt a test and learn approach. The framework set out in this document aims to deliver a series of interventions through trialling activity that will help to build the evidence base around what works. The longer term objective is to identify scalable approaches that could be introduced for the wider self-employed population.

1. We will use trialling activities to test a range of approaches in relation to both joining a pension scheme and facilitating and making regular saving into pensions or other savings vehicles. The trialling activities will gather evidence about how messaging around pension savings is perceived and understood by the self-employed as well as the measurable impact of the different approaches on saving behaviours.

2. We will test the impact of systems and mechanisms to drive voluntary adoption of savings behaviours; testing the impact of interventions to prompt individuals to start saving and then to normalise regular contributions. In order to ensure the widest possible learning, the trialling activities will not target particular sectors but we have focused on the more common forms of self-employment (as detailed in Chapter 2) primarily sole traders, limited companies and partnerships.

3. There have been limited systematic or formally evaluated studies seeking to test the effectiveness of communication interventions to increase voluntary long term saving amongst the self-employed. In developing the trialling activities there is an opportunity to test more traditional marketing interventions and to test variants that rationalise the reason to save for the future, such as focusing on family rather than on the individual.

4. Beyond testing communication prompts we are keen to understand whether there are mechanisms that can help self-employed people make regular saving contributions within the day-to-day management of their finances. A hallmark of automatic enrolment is that the use of payroll enables automatic contributions to be made into a pension scheme for the individual worker. For the self-employed, we are exploring using systems that will work towards achieving a similar outcome, for example, a commitment to make a contribution to a pension via an invoicing or accounting systems.

Trialling activity and delivery partners

5. The department has entered into partnership with NEST Insight to deliver a programme of trialling activity. NEST Insight has a well-established research programme both internationally and in the UK. NEST accepts self-employed members and the scheme currently has over 7,000 members who joined as self-employed. This number is building steadily. NEST Insight is well placed to test how best to help improve retirement outcomes among the self-employed.

6. The department is also working on trials with organisations from the financial services sector, including Lloyds Banking Group, Barclays plc and Aviva plc; and trade bodies including IPSE.

Trialling activities starting from 2018

Primary research/insights: winter 2018 to summer 2019

Delivery organisation	Trialling activities
NEST insight	Primary research[footnote 51] Conduct literature review of existing evidence around communications findings to inform message testing. Undertake qualitative interviews with the self-employed to understand how messages are understood.
Lloyds Banking Group	Primary research Undertake research to test attitudes to possible interventions such as the facility to transfer money from a bank account into a pension. Carry out focus groups, followed by qualitative testing.
Barclays plc	Primary research Undertake research with their self-employed business account customers to enhance understanding of savings, including retirement savings, habits and motivations among this population who access Barclays financial services.

Trialling activities: winter 2019 to autumn 2019

Delivery organisation	Trialling activities
NEST insight	Testing messages on NEST members Use findings from qualitative research to develop proof of concepts with their self-employed members and members who have stopped contributing. Use email communications to test messages that can encourage members to (re)commence pension contributions
NEST insight, IPSE, Smart Pension and Aegon	Testing messaging and access to products on wider self-employed sample NEST Insight, working with IPSE, Smart Pension and Aegon will carry out a set of research trials. This will test message prompts with intervention and control groups. The intervention group will be offered a core pension product (supplied by Aegon) and a liquid ‘sidecar savings product’ (supplied by Smart Pension); and a long term savings vehicle (also supplied by Smart Pension). The latter will be available to individuals to save into, following a pre-set commitment saving level has been reached. The trial will aim to test behaviours including participation; savings levels and financial well-being.
Aviva plc	Testing messages on Aviva members using insurance channel Following primary research, the pilots will test, across a number of different intervention groups, how best to engage the self-employed on the benefits of pension saving. These campaigns will take place across the first half of 2019 with some starting from February 2019. The outcomes will be monitored on an ongoing basis throughout 2019. The results and learnings of the pilots will be available from the second half of 2019.

Trial areas we propose to develop from 2019

7. We would welcome the involvement of further partners to help deliver this programme of trialling activities in a range of capacities. Our intention is to run larger-scale trials based on the findings from the trial areas outlined above. We would expect to execute these trials via partnerships with commercial providers of services to self-employed groups potentially including aggregators, third parties who provide invoicing and payment services to contracted self-employed labour or invoicing software providers, or retail and business banking providers.

8. We are currently working with Mercer, who has also been researching various ways to improve access to pensions for the self-employed. Working with DWP, Mercer will be hosting a workshop, planned for March 2019, involving organisations engaged in the front end distribution of self-employed activities. This will complement the DWP/HMT and ABI sponsored tech sprint held with Fintech organisations in March 2018, which focused on the potential technical solutions to make it easier to make contributions into pensions. The workshop will bring together many organisations who provide services to, or represent the self-employed at the front-end of their day to day working environments, including payment systems, accounting firms, trade associations, market places and shared workspaces to help improve access to retirement saving.

Delivery organisation	Areas under development
NEST Insight and third party Partner to be identified: invoice system provider; software provider; pension provider	Testing messages combined with facilitative tool to make savings easier – after November 2019 To enable income to be pre-committed through third parties who provide invoicing and payment services to contracted self-employed labour or where banks automatically make deductions and this is diverted into a savings vehicle or pension.
NEST Insight and third party Partner to be identified: invoice system provider; software provider; pension provider	Testing the role of an intervention to facilitate savings via invoicing/payment systems – after November 2019 To enable income to be pre-committed through a percentage deduction of payments received on goods or services whereby income is diverted into either a pension product or alternative savings account i.e. liquid savings account / rainy day savings.

Chapter 5: Next Steps

1. The trialling framework set out in this document is not intended to be exhaustive – our approach will develop as our learning increases and additional partner organisations are identified.

2. We will continue to engage with a range of interested parties, including third party service providers to the self-employed and those that are involved in self-employed business and/or supply chains. This includes, for example:

• the department has engaged closely with the Federation of Small Businesses (FSB) for insight on how to encourage long term saving amongst the self-employed. We are continuing to engage with the Federation of Small Businesses (FSB); this included providing input into their planned primary research from Summer/ Autumn 2019 on pension savings in small businesses and by the self-employed
• following a roundtable discussion with Trade Unions, we are progressing activity with Community^{[footnote 52]} following their recent consultation with their members’ around attitudes to pension saving. The outcomes may identify opportunities to test messaging to encourage members to re-commence pension saving

3. We recognise a number of interested parties have suggested there may be a potential role for HMRC’s Making Tax Digital (MTD) in relation to self-employed retirement savings. This suggestion was also highlighted in a recent Treasury Select Committee report.^{[footnote 53]} The 2017 Automatic Enrolment Review concluded that the MTD agenda and continued software development concerning the use and integration of software that self-employed will use in relation to their tax returns provided useful opportunities to explore.

4. Since last year, there have been changes to the proposed time frame for the introduction of MTD for income tax self-assessment purposes. The introduction of mandatory digital VAT returns for companies with a taxable turnover of £85,000 per year will begin from April 2019. HMRC will adopt a test and learn approach and will use the learnings from the introduction of VAT Making Tax Digital before extending to the income tax self-assessment process which is not now expected to happen before April 2020 at the earliest.

5. Through MTD, customers will have a choice of third party software programmes to meet their obligations ranging from full accounting packages to more simple bridging solutions (e.g. spreadsheets or other API enabled add-ons). Although the implementation of MTD for income tax is unlikely to be in place before mid-2020, we are already engaging with software providers who will develop solutions for accounting and self-assessment purposes to explore the scope to also test behavioural prompts focussed on about pension savings.

6. In addition to exploring the contact points related to MTD, we also want to look at the potential that GOV.UK provides as an access point for information. This could, for example, focus on carers who enter self-employment from claiming caring benefits, potentially to provide prompts to access advice and guidance through the Single Financial Guidance Body. The trialling activities outlined in this document will begin soon. DWP will conduct internal monitoring of the NEST trials and review the outputs (for example, numbers of individuals prompted to access their accounts) and outcomes (such as changes to pension participation rates and amounts saved) from the trialling activities.

7. In addition, following a successful pilot, independently evaluated by the Centre for Ageing Better, the Pensions Advisory Service (TPAS) launched their ‘Midlife review for self-employed people’^{[footnote 54]} on 10th December 2018. The service encourages customers to think about four key areas to their lives: Money, Health, Work and Family and how they are interconnected and also looks at how saving into a pension could save tax and promote financial resilience to provide support in these areas. This provides a further area of learning to maximise as part of the research and trailing activity.

Call to action for trial partners:

• chapter 4 summarises the trial areas we are continuing to develop. This includes development of larger-scale trials executed through partnerships with existing commercial service providers to self-employed groups
• we would welcome the opportunity to connect with organisations from a range of sectors including; third parties who provide invoicing and payment services to contracted self-employed, invoicing software providers; accounting organisations and self-employed work spaces. If you are interested, please email us: self-employment.deliveryplan@dwp.gov.uk

Annex: Data that underpins delivery framework

Chapter 1

Table A1 shows the growth of the self-employed population relative to 2001. The self-employed population has seen the highest growth in part-time working, particularly for men, between 2001 and 2016. Growth in full-time working in self-employment has been higher for females than for males over this period.

Table A1: Index of growth of self-employed numbers, by gender and working pattern, 2007 to 2016 (Q4 of each year, 2001=100)

	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016
Male full-time (FT)	112.3	111.8	110.5	111.2	112.2	113.9	116.5	117.8	120.9	123.3
Male part-time (PT)	143.4	142.3	149.2	168	176.5	186	199.9	193.9	207.9	209.5
Female FT	115.6	117.7	127.7	129.4	134.8	138.1	157.8	156	161.8	177.3
Female PT	123.1	123.4	133.7	139.2	145.9	160.9	157	177.6	185	184.3

Source: ONS: Trends in Self-employment in the UK (Feb.2018)

For all groups in self-employment, the majority of the income of the self-employed comes from self-employment (Table A2). Part-time workers, have a higher proportion of their income from pension and retirement income when compared to full-time self-employed workers – 34% of male part-time workers’ and 11% of female part-time income comes from pension and retirement income. In comparison, pension and retirement income makes up under 4% of total income for male and female full-time self-employed workers. Part-time self-employed workers also saw a higher proportion of their income from ‘Other’ sources (14% male part-time; 23% female part-time) compared to full-time self-employed (6% male full-time and 10% female full-time).

Table A2: Sources of income of the self-employed (percentage, 2015 to 2016, UK)

	Self- employment	Pension and retirement income	Employment	Other
Male FT	89.8	3.9	0.4	5.9
Male PT	50.6	33.9	1.2	14.3
Female FT	83.7	3.8	2.3	10.2
Female PT	64.7	10.5	1.9	22.9

Source: ONS: Trends in Self-employment in the UK (Feb.2018)

Earnings of the self-employed

The ONS published their latest release on Trends in Self-Employment in the UK^{[footnote 55]}, reviewing changes to the composition of the self-employed over time, with a focus on earnings and education.

ONS (2018) examined changes over time in the median weekly earnings of men and women working full- and part-time as employees and self-employed. Earnings growth between 2000 to 2001 and 2015 to 2016 for the self-employed lagged that for employees in every case. Table A3 presents the median weekly earnings, the percentage growth over time and the ratio of earnings of the self-employed to their employed counterparts for 2015 to 2016.

Table A3: Median weekly earnings for self-employed men and women, 2000 to 2001 and 2015 to 2016

	2000 to 2001	2015 to 2016	Earnings increase over time	Ratio of self-employed to employee earnings 2015 to 2016
Male employees FT	£368.20	£533.00	45%	0.68
Male self-employed FT	£295.70	£363.00	23%	0.68
Male employees PT	£90.70	£153.00	69%	1.08
Male self-employed PT	£126.60	£166.00	31%	1.08
Female employees FT	280.1	£428.00	53%	0.57
Female self-employed FT	£197.70	£243.00	23%	0.57
Female employees PT	£99.40	£177.00	78%	0.68
Female self-employed PT	£71.90	£120.00	67%	0.68

Source: ONS Trends in Self-employment in the UK (Feb.2018)

Table A3 shows that, with the exception of men working part-time, employees earn more on average than their self-employed counterparts. While the biggest increases over time in median earnings were seen in females working part-time (78% increase for employees and 67% increase for self-employed), female self-employed earned, on average, 60 to 70% of the earnings of their employed counterparts. Similarly, self-employed men working full-time earned, on average, 32% less than their employed counterparts.

Educational attainment and self-employment

The ONS also explored changes in the composition of the self-employed with regard to their education levels. Between 2001 and 2016, the largest proportion of those entering self-employment have been those who have a degree (or equivalent). The share of the self-employed with a degree or equivalent increased by 14 percentage points from 19% in 2001 to 33% in 2016 as a share of total self-employed. When comparing the self-employed with a degree as a share of total employment (employees and the self-employed), the share of those with a degree increased from 2.3% to 4.9% over the same period; this highlights how more relatively highly-qualified individuals are opting to go into self-employment.

PPI analysed education patterns of the self-employed from the Wealth and Assets Survey 2014 to 2016.^{[footnote 56]} They found that men who had lower levels of educational attainment were more likely to become self-employed at a younger age than those with degree-level qualifications. Self-employed women were more likely to be educated to degree-level than men. However, PPI highlighted that this difference could be partially accounted for by the relatively low levels of women in manual industries, such as construction, where individuals are less likely to be educated to degree level.

Chapter 3: How do the self-employed fit into the changing labour market?

Saving for retirement amongst the self-employed

The drivers of saving behaviour for retirement among the self-employed – research carried out by IPSOS MORI on behalf of HMRC

Aims of the research: The research aimed to address the long-term saving behaviours of the self-employed and how these differ across different segments of the self-employed population. It looked at the key drivers and barriers to saving for retirement amongst the self-employed.

Methods: This mixed-methods research included a quantitative survey of just under 1,000 respondents to identify key savings behaviours and barriers for the self-employed population – this is an acceptable sample size for a survey of this kind and the researchers weighted the sample against key known characteristics of the self-employed population to ensure the findings were representative. The research was informed by a behavioural framework which provided a structure for exploring the extent to which a range of behavioural drivers and barriers influenced saving for retirement.

Main findings:

Pension saving: Overall, 32% of the sample were currently saving into a pension at the time of interview; 54% had not set aside any money for retirement in the last 12 months.

Self-employed population: The research identified five main segments of the self-employed population: 1) Younger and unprepared (18% of sample; 10% saving into pension); 2) Irregular earners (26% of sample; 12% saving into a pension); 3) Younger and capable (19% of sample; 23% saving into pension; 4) Property endorsers (15% of sample; 56% saving into pension); and 5) Pension endorsers (22% of sample; 68% saving into pension).

Property vs. pensions: In general, property was seen by respondents as the best way to save for retirement: 36% of respondents said the best way to save for retirement was to invest in property. 22% of respondents thought saving into a pension was the best way to save for retirement. The main reason for the preference for property was that it was seen to give the best return on the investment.

Affordability of retirement saving: Over half of all respondents stated that they could not afford to save for retirement – this was highest amongst the ‘Younger and unprepared’ (81%) and ‘Irregular earners’ (77%) segments and lowest amongst the ‘Property endorsers’ (32%) and ‘Pension endorsers’ (23%) segments. For those not currently saving into a pension, the main reason given was that they could not afford to do so (46%).

Number of years in self-employment

It is also important to consider the length of time people typically spend in self-employment when considering the extent to which the lower propensity to save when in self-employment relative to employment is a significant issue. Figure A1 presents the mean number of years’ men and women spend in self-employment by generation.

Figure A1: The mean number of years spent in self-employment^{[footnote 57]} by generation

Generation	Male	Female
Silent Generation	10.4	6.7
Baby Boomers	12.3	8.3
Generation X	8.2	6.2
Millennials	4.5	3.7

Source: DWP analysis of the Lifetime Labour Market database (L2)

Figure A1 shows that men typically spend longer in self-employment compared to women, with younger cohorts on average spending relatively more years in self-employment. However, these data suggest that, on average, people are typically spending more of their careers in other forms of employment or activity to self-employment. This therefore supports the previous findings that suggest that, for many, self-employment will not be their only, or even necessarily their main, employment activity throughout their career. However, younger cohorts will, understandably, be further away from the end of their working lives so may accumulate more years in self-employment than equivalent cohorts born previously, such as the Silent Generation.

This combined with the finding that the majority of people come from employment into self-employment (Chapter 3, Figure 3.5) provides further evidence that at least some of those in self-employment will have some retirement savings through previous employment (if eligible for automatic enrolment). It will therefore be important to understand more around what works to support these people to continue to save into existing pension pots.

Points of contact

In thinking about possible touch points for supporting the self-employed to save for retirement, analysis was conducted on whether the self-employed were likely to be claiming benefits while self-employed. Figure A2 shows the proportion of men and women claiming benefits when first becoming self-employed. Respondents who were claiming benefits could have claimed more than one of the categories (not working, low income, health, care and other).

Figure A2: The percentage of people claiming benefits when first becoming self-employed

Type of Benefit	Male	Female
Not working	17%	6%
Not claiming	74%	47%
Low income	3%	5%
Health	5%	5%
Care	2%	36%
Other	0%	1%

Source: DWP analysis of the Lifetime Labour Market database (L2)

Figure A2 shows that the majority of men and women entering self-employment did not claim benefits (74% men, 47% women). However, where benefits were claimed, there were differences in the types of benefits claimed by men and women when they first became self-employed: men were significantly more likely to claim out of work benefits (typically Jobseeker’s Allowance) when entering self-employment than women (17% compared to 6%)^{[footnote 58]} and women were significantly more likely to claim care benefits (child benefits or other caring benefits) than men (36% of women first becoming self-employed compared to 2% of men).^{[footnote 59]} This is also reflected in the finding that women were more likely to first enter self-employment from caring activities than men (see Chapter 3, Figure 3.5).

Figure A3 compares the proportion of self-employed claiming benefits to employees. As with the self-employed, the vast majority of employees did not claim benefits. However, employees were significantly more likely to not be claiming benefits than the self-employed.^{[footnote 60]} Furthermore, a higher proportion of those entering self-employment were claiming all benefit types compared to those entering employment. This could be reflective of the higher average age of the self-employed relative to the employed population or a difference in the composition of the self-employed relative to employees.

Figure A3: Proportion of self-employed and employees claiming benefits on first becoming self-employed or employed

Type of Benefit	Employed	Self-employed
Not working	9%	13%
Not claiming	79%	64%
Low income	1%	3%
Health	1%	5%
Care	9%	12%

Source: DWP analysis of the Lifetime Labour Market database (L2)

Further analysis highlighted gender differences between those claiming benefits as self-employed compared to those employed. While the proportions of men claiming care benefits were similar for the self-employed and for employees (2% self-employed; 1% employees) on first becoming self-employed or employed respectively, a significantly higher proportion of women entering self-employment claimed care benefits compared to women entering employment (18% employees; 36% self-employed). Furthermore, the same proportion (56%) of women aged 30 to 39 entering employment or self-employment for the first time do so from caring benefits.

1. Free Radicals, Alan Lockey. ↩

2. Automatic enrolment review 2017: maintaining the momentum. ↩

3. Automatic enrolment review 2017: maintaining the momentum. ↩

4. The analysis looks only at an individual’s main activity in a given year. Therefore, men may have also been claiming care benefits alongside employment, but this would not have been picked up in the main activity analysis. Furthermore, it is important to note that the main activity analysis will not pick up when an individual has combined self-employment with employment. ↩

5. UK labour market: November 2018. ↩

6. Workplace Pension Participation and Savings Trends of Eligible Employees Official Statistics: 2007 to 2017. ↩

7. Eliminate unlawful discrimination, harassment and victimisation and other conduct prohibited by the Act. Advance equality of opportunity between people who share a protected characteristic and those who do not foster good relations between people who share a protected characteristic and those who do not. ↩

8. The What Works Trial Advice Panel. ↩

9. ABI 2018 TechSprint. ↩

10. CIPD define gig economy as people who trade their time and skills through the Internet and online platforms, providing a service to a third party as a form of paid employment. ↩

11. UK labour market: November 2018. ↩

12. Automatic enrolment review 2017: maintaining the momentum. ↩

13. Automatic enrolment review 2017: analytical report. ↩

14. Trends in self-employment in the UK. ↩

15. Automatic enrolment review 2017: analytical report. ↩

16. While as noted, female part-time self-employed increased by 85% between 2001 and 2016, female full-time self-employment increased from 433,000 to 732,000, an increase of 69%.” Source: as above. ↩

17. Trends in self-employment in the UK. ↩

18. Policies for increasing long-term saving of the self-employed. ↩

19. To gig or not to gig? Stories from the modern economy. CIPD define gig economy as people who trade their time and skills through the Internet and online platforms, providing a service to a third party as a form of paid employment. ↩

20. Good work plan. ↩

21. Workplace pension participation and saving trends: 2007 to 2017. It is important to note that the fall in pension participation over time amongst the self-employed could be due to a compositional effect (i.e. due to changes in the types of people coming into self-employment) rather than necessarily a fall in the propensity to save amongst the self-employed. ↩

22. Thanks to Tim Pike at Pensions Policy Institute for sharing their code with analysts in the DWP to replicate the analysis with the latest data. ↩

23. The chart includes a breakdown of ‘physical wealth – this includes assets, other than property, such as cars, televisions, furniture etc. ↩

24. Workplace Pension Participation and Savings Trends of Eligible Employees Official Statistics: 2007 to 2017. ↩

25. This figure refers to participation in any type of pension (i.e. workplace pension, personal pension etc.). ↩

26. The sample size was too small to include the proportion of self-employed aged 16 to 29 saving into a pension. ↩

27. Comparison of attitudes and wealth between employees and the self-employed. ↩

28. Comparison of attitudes and wealth between employees and the self-employed. ↩

29. The sample size was too small to include the level of self-employed pension saving in those aged 16 to 29. ↩

30. Analysis of pension wealth has shown automatic enrolment has led to higher levels of pension participation across the board but particularly amongst lower earners (see, for example: Automatic enrolment and pension saving in the UK). Given the data presented here is from 2015 to 2016 and automatic enrolment was introduced in 2012, the effect of automatic enrolment on pension wealth accumulation will not be fully captured here. ↩

31. Comparison of attitudes and wealth between employees and the self-employed. ↩

32. Policies for increasing long-term saving of the self-employed. ↩

33. Drivers of savings behaviour for retirement among the self-employed. ↩

34. HMRC research, reference as above. ↩

35. Comparison of attitudes and wealth between employees and the self-employed. ↩

36. Working for Today, Preparing for Tomorrow. ↩

37. Money Box. ↩

38. Plum. ↩

39. Chip. ↩

40. Portify. ↩

41. Trezeo. ↩

42. OECD self-employment figures. ↩

43. GigSuper. ↩

44. Automatic enrolment review 2017: analytical report. ↩

45. See Annex 7 of the Automatic Enrolment Review 2017 analytical report for full explanation on the Lifetime Labour Market Database (L2) data source. ↩

46. DWP analysis of the Lifetime Labour Market Database (L2). ↩

47. The analysis looks only at an individual’s main activity in a given year. Therefore, men may have also been claiming care benefits alongside employment but this would not have been picked up in the main activity analysis. Furthermore, it is important to note that the main activity analysis will not pick up when an individual has combined self-employment with employment. ↩

48. Employments refers to jobs (where an individual is an employee) or a period of self-employment listed as the main activity. ↩

49. The touch points identified here were suggested by the self-employed during interviews with NEST as part of the research. ↩

50. Retirement saving and the self-employed. ↩

51. NEST Insight primary research will help to inform the subsequent NEST Insight message-prompt trialling; and also the related trials (including the IPSE, Smart Pension, Aegon, NEST Insight). The primary research (including for example research undertaken by Lloyds Banking Group and Barclays plc) will inform their respective research programmes. ↩

52. Community. ↩

53. Making Tax Digital. ↩

54. Midlife Review for Self-Employed People. ↩

55. Trends in Self-Employment in the UK. ↩

56. Policies for increasing long-term saving of the self-employed. ↩

57. Time spent in self-employment is captured as where self-employment is listed as the main activity in any given year for an individual. ↩

58. Statistically significant at less than 1%. ↩

59. Statistically significant at less than 1%. ↩

60. Statistically significant at less than 1% ↩