UKGC Releases Talks on Gambling Data Bias

Article Content

Regulator's Data Accuracy Under Scrutiny

The UK Gambling Commission (UKGC) has released partially redacted correspondence with its official statistics partner, the National Centre for Social Research (NatCen), following a Freedom of Information (FOI) request. The documents confirm that discussions have taken place regarding potential biases in the data used to measure gambling participation and harm in Great Britain.

The request, dated 2 July 2024, specifically sought communications between the regulator and NatCen that mentioned terms such as 'selection bias', 'response bias', 'desirability bias', and 'biased estimate'.

Why This Data Matters

Official statistics on gambling, produced for the UKGC by NatCen, form the evidence base for the UK's entire regulatory framework. These figures on participation rates and levels of gambling harm are cited by the government and the Commission to justify policy decisions, including the ongoing implementation of controversial financial risk checks.

Concerns about the accuracy of this data are significant for consumers. If the foundational statistics are flawed, it raises questions about whether the regulations built upon them are proportionate and correctly targeted.

Details of the Disclosure

The FOI request was submitted by NatCen itself and Dr James Noyes of the Social Market Foundation think tank. It was made in two parts:

1. All correspondence between the UKGC and NatCen since 1 January 2020, containing keywords related to statistical bias.
2. All correspondence between the UKGC and Dr James Noyes since 1 January 2020.

The UKGC granted a "partial exemption" to the request. It released two redacted PDF documents but withheld certain information. The Commission stated it did not disclose draft documents or information that has already been published.

Furthermore, personal data such as names and email addresses were removed under Section 40(2) of the FOIA, which protects personal information where there is no overriding public interest in its disclosure.

The search terms used in the request point to specific methodological concerns:

• Selection Bias: The risk that the group of people surveyed is not representative of the wider population.
• Response Bias: When survey participants provide inaccurate information, which can include desirability bias—the tendency to give socially acceptable answers, such as under-reporting gambling losses.

Significance for the Industry

While the full content of the discussions remains partially obscured by redactions, the release itself is significant. It confirms that the methodological soundness of the UK's official gambling data has been a topic of conversation between the regulator and its data provider.

The partial nature of the disclosure means that the full context and conclusions of these discussions are not available for public scrutiny. For consumers, this raises important questions about the reliability of the evidence used to shape regulations that directly impact their gambling activities and personal finances.