New data can help inform policies that promote adequate retirement saving

Helen Miller from the Institute for Fiscal Studies describes in this blog how a project linking tax records to data on private pension saving could benefit those preparing for retirement, particularly the self-employed.

How much people save for their retirement is extremely important. It matters for individuals' wellbeing in later life. It also matters for society in general because those who don’t have enough private savings are more reliant on government benefits. It is increasingly important because the population is aging and people are living longer.

There are very large tax subsidies given to pension savings. They have been tweaked many times in recent years. And, in 2012, the government introduced pension auto-enrolment. This policy requires employers to put most of their employees into a workplace pension scheme and contribute towards it. This has led to almost 90% of eligible employees saving in a pension.

Despite substantial policy action, there are ongoing concerns that some groups are not saving enough for their retirement.

The pension savings landscape has changed in ways that make it harder to save enough for retirement. Broadly speaking, older generations were much more likely to have access to ‘defined benefit’ pensions, in which employers bear the risk of ensuring that employees have access to a certain income stream in retirement. Younger generations are much more likely to be saving in personal (or ‘defined contribution’) pensions. Compared to defined benefit schemes, these are less generous and expose individuals to more of the risk associated with variation in returns. At the same time, a period of low earnings growth and very low interest since the mid-2000s has made saving for retirement more unaffordable.

The self-employed are a particular concern because they are much less likely to invest in a pension and, when they do, they save less than employees. There is no equivalent to auto-enrolment for the self-employed but governments are actively looking at policy options.

New data will provide the first complete picture of pension saving

To design effective policies, governments need to know how pension saving varies across different groups in society and the impact of various policy changes on how and how much people save.

Research to date has used survey data to study these issues. This data is valuable. But even the best available survey datasets are incomplete because they only cover employees (not the self-employed or those not in paid work) and they do not cover pension contributions made outside of a workplace. Surveys also have limited information on people over time and they record people's earnings but not their total income (which is particularly important for many of those on the highest incomes).

We will be creating the first dataset that provides a comprehensive picture of how much money people are putting into personal pensions.

Specifically, we will use comprehensive information on pension contributions held by HMRC, the UK tax authority. The data will contain all pension contributions made by employers or individuals (whether they are employees, self-employed or not working) to personal pensions. We will link this information to individuals' tax records, which will provide information on how much income a person has and how that income is generated (for example whether it comes from self-employment). We will be able to track people over time, thereby observing, for example, how pension saving changes as income or government policies change. All data will be de-identified meaning that researchers cannot identify specific individuals.

New research will investigate saving by the self-employed

The self-employed have been the fastest growing part of the UK labour force since the early 2000s. Around 15% of the UK workforce is now self-employed but there is much churn in this population - many enter and leave self-employment each year.

We will use the new de-identified data to document the pension savings of the self-employed, including how pension contributions change as people move between employment and self-employment. We will study whether auto-enrolment has changed the saving habits of people who were auto-enrolled as an employee and went on to be self-employed. This will provide important evidence on the channels through which auto-enrolment operates and will inform the design of policies to encourage saving among the self-employed.

With the new dataset established for research, there will be opportunities to address many other important policy questions including, for example, how pension saving responds to various aspects of tax policy.

Read more about the dynamics in private pension saving on the related project page.