The use of commercial data to assess the reliability of surveys in a post-Covid world

Session Title

Public Health: Monitoring & Data Innovation

Presentation Type

Paper Presentation

Start Date

27-5-2026 12:00 AM

Abstract

Confidence in official statistics on gambling, collected by population surveys, has been negatively impacted in recent years as a result of falling response rates as well as changes to data collection methods. A change in data collection from in-person interviewing to online self-completion has resulted in higher reported levels of gambling and gambling-related harms, creating considerable uncertainty about the ‘true’ extent of both. Because these data are relied upon by the Government to determine regulatory policy, there is a risk that the use of unreliable data may distort market regulation. In this study, administrative data obtained from licensed gambling businesses, are used to assess the accuracy of survey estimates of participation. It shows that, in relation to three specific activities (non-remote casino table games, betting exchange and the football pools) online survey data substantially overstate levels of participation in the population. This finding indicates that policy-makers should exercise caution in interpreting the results of survey data in relation to these specific activities; and that further work should be undertaken to understand the extent to which this over-reporting may be generalised across all gambling. The use of ‘real world’ administrative data to test self-report survey findings is likely to have benefits in relation to a far wider range of statistical datasets compiled by government agencies and used to inform public policy.

Author Bios

Dan Waugh is studying for a PhD in market regulation at the University of Liverpool; and is a partner at the specialist gambling research and advisory business, Regulus Partners. Between 2000 and 2014, Dan held a number of senior finance and strategy roles within Britain's leisure, hospitality and gaming industry - with Whitbread PLC and The Rank Group Plc. He was also founding chair of the harm prevention charity, YGAM.

Share

COinS
 
May 27th, 12:00 AM

The use of commercial data to assess the reliability of surveys in a post-Covid world

Confidence in official statistics on gambling, collected by population surveys, has been negatively impacted in recent years as a result of falling response rates as well as changes to data collection methods. A change in data collection from in-person interviewing to online self-completion has resulted in higher reported levels of gambling and gambling-related harms, creating considerable uncertainty about the ‘true’ extent of both. Because these data are relied upon by the Government to determine regulatory policy, there is a risk that the use of unreliable data may distort market regulation. In this study, administrative data obtained from licensed gambling businesses, are used to assess the accuracy of survey estimates of participation. It shows that, in relation to three specific activities (non-remote casino table games, betting exchange and the football pools) online survey data substantially overstate levels of participation in the population. This finding indicates that policy-makers should exercise caution in interpreting the results of survey data in relation to these specific activities; and that further work should be undertaken to understand the extent to which this over-reporting may be generalised across all gambling. The use of ‘real world’ administrative data to test self-report survey findings is likely to have benefits in relation to a far wider range of statistical datasets compiled by government agencies and used to inform public policy.