Context

The introduction of automatic enrolment into workplace pensions in the UK from 2012 was intended to address shortfalls in retirement saving among the working age population. It has resulted in substantial increases pension participation rates, leading to higher average saving within the pension, with low opt-out and short-term cessation rates.

Existing evidence shows that the effect of automatic enrolment in generating new saving has partly been to reduce some spending and increase some borrowing. However, it is not yet clear if it is also offset by reduced levels of saving elsewhere within an individual’s finances. This report builds on these studies by asking whether people also drew on their savings to cover some of the cost of being enrolled.

The study

The study was undertaken by Nest Insight in collaboration with the University of Nottingham and Warwick Business School, with funding from the Nuffield Foundation. It used new analysis of the UK’s largest household panel survey (Understanding Society, 2009-2020), linked to automatic enrolment pension records from the National Employment Savings Trust (Nest).

Further to consent given in Wave 11 (2020) of Understanding Society, the initial available sample of all linked individuals was 1,672, reducing to 1,522 people who were automatically (rather than voluntarily) enrolled into Nest. Analysis of these linked cases, defined as the treatment group, enabled:

• examination of the broader financial position of automatically enrolled individuals and
• by exploiting the staggered roll-out of automatic enrolment to UK firms over the preceding years, estimation of the effect of automatic enrolment on active monthly non-pension saving and contributions to other pensions outside of the workplace pension using a staggered difference-in-differences econometric approach – over the 5 years before the firm’s on-boarding to Nest and 4 years post on-boarding , referred to as pre- and post-‘staging’.

Critically, comparisons of savings behaviours were made between the treatment group and a representative sample of the working-age population in employment who formed a control group who were ‘yet to be enrolled’ into Nest.

Key findings

• Compared with the in-employment working population as a whole, Nest pension savers typically had slightly lower incomes, lower education levels and were less likely to own a home compared with the in-employment working age population.
• 54% of Nest enrolees had no stock of non-pension financial savings, 56% were not contributing to any new non-pension financial saving, and these rates were similarly low among the enrolees' spouses or partners.
• No statistically significant effects of automatic enrolment on active monthly non-pension saving, or on the likelihood of contributing to a self-invested personal pension, were observed (estimated coefficients for each of the 10 years covering pre-staging and post-staging were centred around zero; p>.05).
• Further examination indicated that automatic enrolment did not substantially change non-pension savings behaviours (for each of the 10 years covering pre-staging and post-stating, 95% confidence intervals were wide and greatly overlapping).
• However, smaller-sized effects and differences between subsets of individuals could not be ruled out.
• Overall, the average active monthly saving among those who were saving was £338, and the analysis was able to rule out a decline in active monthly savings by 40% of more.

Points to consider

Methodological strengths or limitations:

• The authors present an analytical approach which has a sound theoretical basis. They note, however, that the accuracy of the survey data was dependent on self-reporting, the granularity of financial variables was lower than would be available in administrative data, and that they were constrained by small sample sizes.
• As such the results may include false negative results (i.e. it was not possible to detect meaningful differences where they existed).

Applicability:

• The findings should be of particular interest to policymakers, as well as employers who might also offer schemes to employees to encourage them to save into non-pension savings.

Relevance:

• The findings are highly relevant given the challenges to and importance of financial wellbeing in current economic landscapes, both for retirement and during working life
• The findings are also highly relevant given the – policy-defined – passive enrolment of employees into Nest workplace pension, which implies certain risks and responsibilities to policymakers, the scheme provider, as well as individuals.

Generalisability/transferability:

• There is likely to be some transferability of the learning from the study to other similar markets, where there is automatic enrolment into defined contributions workplace pensions and where the contexts of financial wellbeing among workers are similar.