Requirement 1: Substantial revisions and potential bias in the revisions reduce the usefulness of the monthly trade in services estimates for key users. To help the statistics better meet user needs.

ONS should:

• provide more information and analysis in the bulletin on the uncertainty of the estimates to reassure users and aid the interpretation of the estimates.
• contextualise the revisions relative to the size of change in the trade balance and with reference to earlier periods of economic turning points and volatile growth.
• publish separate revisions triangles for trade in services.
• investigate the extent of any bias and, if deemed necessary, publish a plan that sets out how it intends to address the bias. ONS should share the analysis with users.

Requirement 2: To ensure that users have all the necessary information to appropriately interpret the UK trade statistics, ONS should synthesise its previous analytical work on trade asymmetries. This should include the work ONS is doing to minimise these asymmetries and how these trade asymmetries impact on use of the data.

Requirement 3: To improve users’ understanding of the quality of the UK trade statistics and the methods used to produce them.

ONS should:

• update the Quality and Methodology Information report and the methodology document and ensure these are updated regularly
• consider keeping only one document that covers quality and methods or improve the signposting to both existing documents, being clear to users on the difference

Requirement 4: To ensure quality, reliability and efficiency in production, that will also enable ONS staff to carry out both production and continuous improvement.

ONS should:

• review the resources available to deliver the UK trade statistics, including identifying and mitigating the risks of single points of failure
• review its knowledge transfer process and support business continuity during staff changes
• improve and introduce efficiencies in its production processes
• review its current processing system to understand and reduce the level of risk of errors
• publish a delivery plan with milestones for implementing a more effective and efficient processing system