Data collection and sharing for research on games of chance

At present, licensed gambling providers are legally obligated to share anonymised data with researchers upon bilateral request for the purpose of studying gambling addiction (Gambling Act (Wok) art. 31m). However, this provision has been underutilised and faces several challenges:

1. The application process is suboptimally designed;
2. Due to the requirement for anonymisation, data cannot be linked at player level across providers;
3. Background characteristics cannot be linked to the data due to anonymisation;
4. The Gambling Authority (Ksa) is unable to conduct all desired research themselves;
5. Access under the law is limited to research for preventing gambling addiction.

The Ksa does collect structured data for supervision in a control database but cannot readily share this data with researchers or use it for all research purposes.

Legal Obstacles

There are three legal obstacles hindering further data sharing:

1. Legality of data sharing. GDPR prohibits processing of special personal data (which is likely involved here). To address this, data needs to be anonymised (possibly after linking) or an exception for scientific research (AVG art. 24) needs to be applied. Additionally, gambling legislation may limit sharing non-anonymised data for purposes other than gambling addiction prevention.
2. Securing cooperation from providers. Providers cannot be assumed to voluntarily share data beyond their existing obligations. Possible solutions include (1) self-regulation, (2) amending gambling legislation, or (3) instructing the CBS by ministerial order to gather the data.
3. Mandate of the Ksa. The Ksa can only conduct research within its legal mandate. However, due to current tasks, the Ksa cannot perform all desired research. To address this, the mandate of the Ksa under article 33b Wok needs adjustment.

Solutions

Four solutions have been identified for enhanced data sharing to address the above challenges (within legal frameworks):

1. Enhancing the current bilateral approach. By increasing the role of the Gambling Authority, the process can be smoother, although it does not improve data linkage.
2. Establishing a decentralised data sharing platform. Data access and traceability can be technically ensured. This solution offers high security but is relatively complex and requires investments from providers.
3. Appointing a central intermediary to facilitate data sharing and provide linkable pseudonyms for players. This intermediary could be a third party or the Gambling Authority (if mandate is revised). This solution relies more on organisational measures.
4. The CBS can gather and provide the data via CBS Microdata. This utilises CBS's authority to collect data and existing facilities for working with microdata.

Aligning with the data model used for Ksa supervision in all cases (CDB/GDT) is advisable. The solutions differ in legal approach, required investments, implementation period, and data protection risks. While the cost-benefit analysis was not part of this study, it is pertinent. In policy development, it may be decided to partially implement a solution or stagger its implementation over time. For example, short-term solutions to enhance the process could be employed first, with data linkage to background information achieved in the longer term. The choice of solution may influence how data sharing can be expanded in the future.