Supporting the appropriate use of welfare statistics in public debate 

Published:

Decorative image. Column chart

Summary 

Welfare statistics are central to public debate about public spending, living standards, labour market participation, and the sustainability of the benefits system. They cover a broad and complex set of sources, each designed to answer different questions, and are used by government, Parliament, journalists, researchers, charities and the public to inform policy, scrutiny and wider understanding of the welfare system.   

Public confidence in welfare statistics is supported when statements are based on data that are accessible, clearly defined, and accompanied by the contextual information needed for accurate interpretation. The Code of Practice for Statistics and the Standards for the Public Use of Statistics, Data and Wider Analysis provide the framework for meeting these expectations. 

When welfare statistics are used without sufficient accuracy, transparency, or supporting information, this creates risks to public understanding and can undermine informed debate on welfare issues that are of significant public interest. 

In line with OSR’s role in supporting the appropriate use of statistics in public debate, this statement sets out key considerations when using commonly cited Department for Work and Pensions (DWP) sources, including the Family Resources Survey (FRS), Universal Credit statistics, and the Benefit Expenditure and Caseload Tables. 

Key things to consider  

Ensure the evidence matches the claim 

Those using welfare statistics in public discourse should ensure the evidence they cite matches the claim they are making.  For example, “total welfare expenditure” refers to a broad aggregate measure, covering many different forms of state support. It includes Universal Credit, pension-age benefits such as the State Pension, Personal Independence Payment, housing support, and payments to families with children. While useful for understanding overall welfare spending, it should not be used to support conclusions about specific parts of the system unless the relevant components are clearly identified. 

For instance, commentary focused on working-age benefits should make clear if the cited figure also includes pension-age benefits and other forms of support. Without this breakdown, audiences may wrongly assume that a total welfare figure relates only to the part of the system being discussed. 

Be clear about definitions 

Clear definitions are essential to accurate interpretation. Many welfare statistics rely on technical concepts that are not always widely understood, and using these terms imprecisely can change how a statistic is interpreted. For example, in the Family Resources Survey, household-level figures may combine support received by more than one family living at the same address. 

Users should therefore explain key definitions and make clear what unit a statistic measures. Without this, audiences may wrongly interpret figures about ‘households’, ‘claimants’ or ‘benefit units’ as describing the same thing, which may lead to conclusions about the amount of state support being received by individual families that the data do not support. 

Present statistics with sufficient context  

Before drawing conclusions from welfare statistics, users should present enough context for audiences to understand what has changed, why it may have changed, and what conclusions the data can reasonably support. This is particularly important where statistics are used to imply a cause, trend, or policy effect. For example, changes in the Universal Credit caseload should not automatically be presented as direct evidence of changes in welfare expenditure, economic conditions, or the effect of a particular government policy without explaining the wider factors that may be relevant. 

In practice, changes in Universal Credit claimant numbers or expenditure are likely to reflect a combination of factors, including economic conditions and policy changes, but also structural changes to the benefits system, such as the continued transfer of claimants from legacy benefits to Universal Credit. Not providing this context creates a risk that audiences may wrongly infer that a change has a single cause, or that it is mainly attributable to recent policy decisions. 

Be transparent about methods and sources 

Public confidence is strengthened when statistics used in public commentary are clearly defined, traceable to their source, and accompanied by enough information for others to understand how they were produced and, where necessary, replicate the analysis. 

This is particularly important when using open-access tools such as DWP’s Stat-Xplore, which allows users to create bespoke outputs from a wide range of welfare datasets. These data are openly accessible, which presents a heightened risk that figures may be extracted, combined or interpreted in different ways without sufficient explanation.  

Without clear documentation of the choices made, users may find it difficult to distinguish between robust analysis and outputs that are technically reproducible but potentially misleading. Published analysis should identify the dataset used, relevant variables and filters, the time period covered, and any important methodological choices. This enables users to scrutinise the evidence, understand its limitations, and assess whether the conclusions drawn are supported by the data.   

Conclusion 

OSR expects all those using welfare statistics in public debate to ensure that the statistics are accurate, transparent, and clearly explained. This supports public understanding and confidence in the numbers and those using them. Therefore, those using the statistics should: 

  • explain what dataset is being used and what unit is being measured 
  • explain whether policy, administrative or methodological changes may affect comparisons over time 
  • ensure the selected measure matches the specific issue being discussed 
  • ensure any relevant context, limitations or caveats have been clearly communicated  

We will continue to engage with public bodies to support consistent application of these expectations and to promote the appropriate use of official statistics in public debate. 

Universal Credit statistics: background information and methodology – GOV.UK 

Universal Credit Official Statistics: Stat-Xplore user guide – GOV.UK 

Benefit expenditure and caseload tables: guidance and methodology – GOV.UK