Measuring displacement effects across gaming products: a study of Australian gambling markets

Abstract

This article seeks to establish the extent of displacement effects across gaming products. This is a particularly difficult question to address with precision. To date, many studies have looked at policy changes such as the introduction of a new product into the market to assess the severity of displacement effects. However, simple before and after comparisons are invalid. One needs to know what the appropriate counterfactual would have been in the absence of the policy change  – which by its very nature is unobservable. The alternative is to look at identical populations exposed to different regimes. Australia represents such a natural experiment. In Australia, Betting and Gaming legislation is determined at the state level, giving rise to some interesting differentials across states within a single country. This article estimates a state level (fixed effects) panel data model, exploiting the intra-state differences in the portfolio of gaming products available, to estimate the extent of displacement effects across the gaming sector. The results are particularly relevant to the current UK policy debate, which is focused on the potential impact on the existing market following a forthcoming, radical deregulation of the industry (with further liberalization proposed).

Acknowledgements

David Forrest was supported in this research by a Fellowship from the Leverhulme Trust. We are also grateful to Martin Pierce for preliminary data investigations and to David Davies, Department of Justice, Victoria, for helpful information regarding gambling regulation and legislation within Australia.

Notes

¹ In many jurisdictions, the term ‘casino’ is used to describe any location the main function of which is the provision of machine or table gaming. However, in Australia it is reserved for large Las Vegas-style venues and not used when referring to even large machine gaming rooms in hotels and clubs.

² In Australia, bars and pubs are referred to as hotels and clubs are community based nonprofit organizations.

³ Tribes are not obliged to file turnover and other financial data and the number of slots therefore, had to serve as a proxy for dollar expenditure on slots.

⁴ For Missouri, Siegel and Anders (1999), using county-level data, found that expansion of riverboat gaming was associated with decline in consumer expenditure on other forms of entertainment.

⁵ Linking of machines within venues is permitted in Victoria, NSW, Queensland, Australian Capital Territory (ACT) and Northern Territory. Externally linked machines operate in NSW, Queensland and ACT.

⁶ We have estimated the model excluding the expenditure variables (tobacco, alcohol and café/restaurant/hotel expenditure) and the results are remarkably robust. These results are available from the authors on request. We choose to report the results with their inclusion as they suggest interesting avenues for further research.