Gambling Research Index

The Weekly Digest No. 1

Legal online betting and outpatient mental-health care, helplines that answer but do not refer, and borrowing as the symptom that leads

October 19, 2026. By Chris Beerman, LCSW, CADC.

Online sports betting legalization tracks with more depression and anxiety care, a third of US helpline calls end usefully, and Norwegian panels point to borrowing as the symptom that leads.

The week's strongest paper uses insurance claims for almost 62 million enrollee-years to put a figure on what legal online sports betting costs outpatient mental-health care. The rest of the list is smaller in scale but practical: an audit of helplines in 48 states, a symptom network that singles out borrowing, and the CFTC's formal move to treat event contracts as swaps.

Studies worth reading

Online Sports Betting Legalization and Behavioral Health Care Use

JAMA Psychiatry, United States.

A staggered difference-in-differences study of employer-sponsored insurance claims from 2016 to 2023, comparing 22 states that legalized online sports betting between 2018 and 2022 with 13 that had not, across 61,951,305 enrollee-years of adults aged 18 to 64. After legalization, the share of enrollees filling antidepressant or anti-anxiety prescriptions and making outpatient visits for depression or anxiety rose, by roughly 5 to 7 percent relative to the sample averages, with tighter estimates for men than for women. Substance-use treatment did not change, and neither did emergency or inpatient care. Claims do not show who bets, so this is a state-level association rather than a link between any individual's betting and their treatment, and the uninsured and publicly insured are outside the sample. Even so, I read it as the best population-level estimate yet of the care burden that legalization hands to health systems.

Invisible Obstacles: Audit of Consumer Experience When Calling Gambling Helplines in the United States

Journal of Gambling Studies, United States.

The paper has two parts. In the first, 179 adults who had gambled in the past year said they wanted a knowledgeable, non-judgmental person on the line rather than an automated system, and that one bad call would put them off calling again. In the second, trained callers posing as people looking for treatment made 96 calls, two to each of 48 states: 91 percent were answered, but only 66 percent of answered calls produced a treatment referral, and 31 percent met the authors' combined standard of adequate length, a concrete referral and a helpful experience. The audit is two calls per state, so single-state results are thin. It was funded by the Tennessee Department of Mental Health and Substance Abuse Services, and two authors hold editorial posts at the journal, its editor-in-chief and an associate editor.

Undirected and directed networks of problem gambling: Follow the money trail

European Psychiatry, Norway.

Four representative Norwegian population surveys (2013, 2015, 2019 and 2022; 3,232 to 6,034 respondents aged 16 to 74 each) and two linked panels (3,043 people followed 2013 to 2015 and 2,749 followed 2019 to 2022) were used to model how the Problem Gambling Severity Index items relate to one another. In the cross-sectional networks financial problems were the most central symptom; in the panels, borrowing money most strongly predicted later symptoms, and guilt was the symptom most predicted by the others. Whether the two panels' directed networks agreed was inconclusive, and the measure is a self-report screen spaced two or three years apart. For screening and for anyone designing financial safeguards, borrowing looks like the item to ask about first. Funded by the Norwegian Competence Center for Gambling and Gaming Research, using surveys funded by the Norwegian Gaming Authority.

Frontiers in Public Health, Serbia.

A cross-sectional survey of 2,379 secondary school students in Novi Sad, screened with the Gambling Disorder Identification Test: 13.2 percent gambled recreationally and 4.0 percent screened positive for gambling-related problems. Students who gambled were more likely to play sport, but once sex, mental-health history and substance use were accounted for, sport participation was not independently associated with problems (adjusted odds ratio 1.33, confidence interval crossing 1). Male sex, previous psychiatric treatment, family psychiatric history and substance use carried the association. Four schools chosen purposively, no response rate, a yes-or-no measure of sport and an adult screening tool limit how far this travels, but it is a useful corrective to treating young athletes as a risk group in themselves rather than an audience for prevention.

Forecasting 2026 Congressional Control: A Comparison of IEM and Kalshi Prediction Markets

PS: Political Science & Politics, United States.

The authors compared daily prices on the academic Iowa Electronic Markets, which caps stakes at $500 and charges no fees, with Kalshi's prices for control of the House, the Senate and both, from market opening to September 19. The two broadly agreed on the House but disagreed on the Senate, and their day-to-day movements were weakly correlated or not at all. Kalshi's prices showed significant short-run momentum in three of the series and the IEM's did not, which the authors attribute to Kalshi's larger scale and to the hedgers, speculators and momentum traders it draws. This is a forecasting paper, not a harm study, but it is some of the first peer-reviewed evidence on how the commercial election markets behave. The authors' school, the University of Iowa's Tippie College of Business, runs the IEM.

Policy and public health news

Also in the index

Sixty-two studies entered the index this week, among them a study of gambling and discrimination among substance-use treatment patients, with free full text, and a protocol for a ten-year clinical and register study of patients with gambling disorder. All of them, with filters, at gambling-research.com/research; this week's legislative and court activity is at gambling-research.com/bills and gambling-research.com/courts.