Share
Most of the United States now has some kind of a regulated system for online gambling. In most of the country, that is a relatively recent development.
Because of the small amount of time that has passed and the non-uniform implementation of online gambling regulation across the US, there have been few attempts to study the impact of such changes on the population. That’s especially true from a perspective broader than people who live in the same jurisdiction or who all fit within another demographic.
That has begun to change with the publication of a paper by Daniel M. McCarthy, Wayne J. Taylor, and Kenneth C. Wilbur. The paper, published by the University of California San Diego where Wilbur works as professor of marketing and analytics for the Rady School of Management, sought to identify the impact of online gambling expansion in two key areas.
The authors’ main findings are that regulating online gambling increases the frequency of what they deem irresponsible gambling behavior and that the same also increases tax revenues for enacting jurisdictions. The data has many applications for legislative bodies considering enacting online casino regulation measures.
How the authors collected their data
In data gathering, authors split 14 states into five groups that enacted gaming expansion laws between 2020 and 2023. As a control group, the authors used 18 states that did not do so in that time period.
To measure and observe trends, the authors pulled data from several sources. Those were:
- Calls to National Council for Problem Gambling (NCPG) helplines
- Gambling licensee revenue and taxes paid from Legal Sports Report
- Suicide data from the US Centers for Disease Control and Prevention
The authors also compiled financial records. The financial dataset includes electronic payment records from several million anonymous US adults to 5,000 large merchants and “monthly pre-tax household income for some panelists who received direct deposits.” Income records were also available for 33.7% of a subset of “potential gamblers” during the study period.
The term “potential gamblers” applied to individuals in the study group who fit three criteria:
- Transferred money to or from one of the 41 gambling merchants in the data
- Used tracked cards regularly for primary expenses in every year from 2019 through 2023
- Whose first gambling transaction was a deposit
That subset consisted of 717,724 people. The authors used a more recently devised approach to their analysis of all this data.
How the authors reached their conclusions
To make sense of the data points and glean insights, the authors used a generalized synthetic control framework. That method works to combat the parallel trends assumption (assuming that because phenomena are moving in the same direction at a similar rate and point in time, they must have a causal relationship) by using machine learning to produce counterfactual estimates (what unobserved outcomes might have occurred without observation).
In this case, that is, “what gambling-related outcomes would have occurred in legalizing states after policy enactment if the legalizing state had not enacted online gambling” regulation. This accounting is crucial because it allows researchers to analyze data with a greater degree of objectivity.
In the discussion of their research, the authors do a stellar job of acknowledging the limitations of both their analysis and data.
The authors’ acknowledgements in the paper
As an example of the limitations of the data, financial records could not identify the purposes of payments made to third-party wallet services or specify which form of gambling payments made to online gambling companies were used for. Furthermore, the entire data set does not address the frequency with which study subjects gambled, merely their transactions with identified gambling operators.
The authors also acknowledge that regarding sums of gambling helpline calls, “their normative interpretation requires further research.” Put more simply, the research doesn’t indicate the content of calls placed to the NCPG’s helpline.
The authors rightfully take care to not leap to place a clinical diagnosis on any individuals. The paper defines the gambling, not the gambler, as irresponsible if that gambling includes spending more than 1% of a gamblers’ average monthly income.
While these are apt and necessary acknowledgments, there are additional points to make.
Analysis of the paper
This is undoubtedly the largest-scale research project regarding the effects of online gambling regulation on US jurisdictions and populations to date. While scale alone is not a guarantee of accuracy or usefulness, there is little to criticize in the authors’ analysis and methodology.
Perhaps the greatest point of criticism is the categorization of individual US states in the study group to measure potential deviations in the process of gambling expansion. One of those five categories was states that implemented in-person sports betting prior to doing the same with online sports betting.
While accurate, this stratification may be over-simplified. To say that the environment for regulated sports betting in Arkansas is identical to that of Illinois just because both states implemented regulated online wagering after regulated in-person betting went live is reductive.
Legal online sports betting in Arkansas, compared to Illinois, consists of a fraction of the choices for consumers and none of the powerhouse national brands. For that reason, it’s fair to question whether Arkansans whose financial records were part of the data were as exposed to gambling adoption and marketing materials for sports wagering as Illinoisans were during the study period.
That leads to another criticism. Mere legalization is not tantamount to regulation. The authors accurately state, “legalization can … enable regulation,” but great deviation in regulated frameworks for online gambling exists from one US jurisdiction to the next, even within similar gambling types.
While pulling data from a broad number of states can account for the regulatory differences between jurisdictions, these deviations might create room for a deeper or separate generalized synthetic control analysis to address the potential influence of rules that are less or more restrictive than others. An example of these differences include regulations that ban gambling advertisements near college campuses that exist in many jurisdictions but are not universal.
Regardless of these criticisms, there are important takeaways for legislators in jurisdictions considering further expansion of regulated online gambling. The findings stress several points in such undertakings.
Insights for interested parties
The authors propose that policies to regulate previously black-market online gambling increase tax revenue as well as gambling helpline calls and irresponsible gambling frequency more than policies that do not regulate black-market online gambling. Furthermore, they state they find that gamblers with lower incomes are more likely to increase their frequency of irresponsible gambling after policy changes in favor of legalization.
To begin, it’s important to note that the study does not show that instances of gambling addiction increase with online gambling legalization. Focusing on NCPG helpline calls lasting longer than a minute as an example, the data show that in New York there was a “four-fold spike at the time of OSB (Online Sports Betting) legalization, but the other six states do not show a strong correspondence of calls to OSB policy timing.”
Concurrently, this paper emphasizes the need for public education about responsible gambling practices in step with the expansion of regulated online gambling. Compulsory, deliberate and well-funded modules in existing public education structures would be a great inclusion in such legislation.
Finally, this paper stresses that expanding the regulated structure for online gambling without codifying and funding resources for citizens who may gamble in an unsafe manner is truly irresponsible. Those resources should also be available regardless of a person’s ability to pay for access, as the paper shows that lower incomes correlate to greater irresponsible gambling measurements.
Hopefully, this paper will be but the start of a cavalcade of research on this subject of a similar quality. The more scholarship on the subject that exists, the better the policies that jurisdictions can enact as a result.
