Dear Ed,

I am writing to seek your agreement to cancel the accreditation of the War Pension Scheme (WPS) publication.

The annual WPS publication was designated as National Statistic in 2010, with the status confirmed in 2012 following an independent review by the OSR.

The WPS publication provides statistics on claims, awards and recipients of pensions, allowances and other payments under the War Pension Scheme.

The publication relies on data extracted from the War Pension administrative system. In 2024 the War Pension administrative system transferred from the War Pension Computer System (WPCS) to Compendia Touch (CT).

Following the introduction of the CT data system, a number of data quality issues were identified in the first reporting period using this source. At that stage, it was assessed that these issues were primarily associated with the migration from WPCS to CT and would not persist beyond the initial period. These limitations were communicated through the inclusion of ‘unknown’ categories, explanatory notes in tables, and supporting detail within the Background Quality Report (BQR).

During the subsequent reporting cycle, it has become evident that some of these data quality issues remain, and that updates to system workflows are on-going. Further time is therefore needed to fully understand both the underlying data quality issues and the impact of evolving system processes on reporting outputs.

We are content to continue publication, as the key headline measures, including total claims registered, claims cleared and War Pensions in payment, remain unaffected. The impact is largely confined to more detailed breakdowns. We consider that these limitations are sufficiently transparent to users through the presentation of ‘unknown’ categories and accompanying commentary.

Engagement with your team has highlighted an opportunity to strengthen the BQR by providing greater transparency on workflow and methodology changes, and their implication for the statistics. This will support users in better understanding the data and its limitations.

A key ongoing risk is that the impact of system changes on reporting is not always identified in advance. While working relationships between the data supplier and reporting team remain strong, there is an opportunity to further strengthen shared understanding of system change processes and to more proactively anticipate impacts on statistical outputs.

Finally, claims clearance times data also contains some known inconsistencies, such as cases recorded as both registered and cleared on the same day. The application of standard benchmarking is not feasible due to the complexity of claims and comprehensive validation is not proportionate. The publication will continue to present robust summary measures (median, quartiles and interquartile range) and work will continue to identify opportunities to improve data quality.

As these issues are not expected to be resolved within the coming year, and in line with the emerging findings of your review, I am seeking your agreement to remove the accreditation of the WPS publication. As an Official Statistic, it will continue to comply with the Code.

Yours Sincerely

Tony O’Connor, Chief Statistician – MOD