Towards more reliable gambling cost estimates, a population-based study with register-linkage

Methods commonly used in gambling cost studies yield higher estimates of gambling costs when arbitrary causality adjustment methods are used. Bayes theorem allows leveraging data on temporal patterns of gambling problems to estimate the plausible proportion where gambling is the precipitating factor for the experienced harm, rather than the other way around. Additionally, costs could be presented as Int$ per adult using the PPP exchange rate to facilitate the comparison of gambling costs between countries.