Gambling and its digital distribution, 2005-2026

The internet changed how gambling is delivered and how gambling-related material circulates. These changes need to be measured separately. Operator revenue measures market activity, while advertising expenditure measures commercial inputs. Neither measures how many people encountered content or suffered harm. This report brings together a structured interdisciplinary evidence review, nine jurisdictional cases across four World Bank income groups, and legal and platform-policy records. It also includes an original analysis of 162 months of Spanish regulatory data. The report preserves a nine-country development panel and demographic evidence without pooling incompatible surveys.

The strongest findings concern expanded access, diverse promotional pathways, substantial exposure in surveyed populations, and concentrated financial vulnerability. There is credible evidence that advertising influences some betting decisions. How much it contributes to population harm remains uncertain. A randomized direct-marketing intervention produced mixed outcomes. US policy studies offer stronger evidence about expanded gambling access than about advertising alone. In Spain, marketing and gross gaming revenue correlate strongly in levels, but weakly in monthly changes. Conditional estimates remain uncertain. Official enforcement records document serious misconduct in licensed and unlicensed settings, but do not support a global content-to-crime elasticity. Historical comparisons show migration online without uniform growth in total participation or typical individual betting. Longitudinal evidence from low-income countries and comparable measures of illegal markets remain particularly scarce.

Author: Emil Valiyev, Independent Researcher. Written 28 September 2026; file assembly completed 29 September 2026. This is a preprint, not a peer-reviewed journal publication.

Original research text: CC BY 4.0. Original software/code: Apache License 2.0. Third-party datasets and cited materials retain their own rights and licenses; see LICENSE_SCOPE.md. Extract the reproducibility ZIP beside the manuscript to restore the documented paths.

AI was used for file creation and document design. The research, analysis, conclusions, and interpretations were conducted entirely by me. When AI was used to create or format files, I explicitly instructed it not to add, remove, alter, or reinterpret any part of the existing research. Tool: Codex (model named by the author as GPT Astra6).