Quality

Quality of the UK Funded Occupational Pension Schemes statistics

1.11 As a result of redevelopment work, users have reported a significant improvement and greater confidence in the quality of the statistics compared with the previous MQ5 PFS survey. ONS now uses TPR’s register of UK funded occupational pension schemes as the sampling frame for FSPS. This no longer requires the 22% uplift to account for under-coverage of the population that was present in the previous sampling frame, the Pension Funds Online list. From the sampling frame, schemes are now split into DBH or DC schemes. The sample design ensures that DBH schemes with more than 100,000 members and DC schemes with more than 25,000 members are fully covered and required to complete the survey every quarter. For the remaining smaller schemes, a sample is selected at random to represent the population. The sample is fixed for 18 months to ensure that the selected sample is representative of all schemes.

1.12 The pension statistics team has established good working relationships with pension fund managers and despite the challenge of survey response during the COVID-19 pandemic, the response rate remained high. As the smaller pension scheme sample is changed every 18 months, it is important that ONS supports the newly sampled schemes in understanding what information is required from them.

1.13 The survey questionnaire now includes automated error messages on mistakes such as missing data and errors in sums. This approach also improves efficiency as errors are detected instantly without the pensions team having to follow up. Other automated validation checks take place once respondents have completed the survey and any issues are queried. The ONS Secure Data Collection system allows the team to communicate easily with respondents in order to receive revised returns quickly, ensuring the timeliness of FOPS publication. Automated processes are also used in producing the statistics, for example, to convert raw data into a weighted sample and run disclosure checks, whilst thorough desk notes and well-documented code make the end-to-end process of production replicable.

1.14 ONS has published Quality and Methodology Information which explains to users the strengths and limitations of the data, the methods used, and the data uses and users. The publication allows users to better understand any limitations of the data and form conclusions on whether the information is appropriate for their specific use.