Understanding beneficial ownership data use

Summary

Measuring the impact of beneficial ownership transparency (BOT) reforms is an emerging area of research. This requires assessing whether the reforms are able to address users’ needs and to ensure beneficial ownership (BO) information from central registers is effectively used. The policy objectives of BOT reforms are wide-ranging and can include ensuring the proper functioning of legal entities and arrangements – collectively known as legal vehicles – and preventing their abuse in enabling corruption, tax evasion, and money laundering.

To better understand how reforms can be best designed for impact, Open Ownership conducted primary qualitative research with a range of users of BO information, interviewing over thirty participants from a range of sectors and jurisdictions. The research leveraged Open Ownership’s global network to gather user insights, understand the range of ways in which BO information can be used, and draw lessons for effective, user-centred implementation of BOT reforms. The intention was to explore how the use of BO information can inform access regimes and ensure these are in line with privacy and data protection laws – meaning that the infringement caused is necessary and proportional, and that users only have access to the information they need to achieve their purposes.

The research found that many experiences and needs are shared across different use types (that is, ways of using BO information). These were mostly expressed as challenges shared among the large majority of users. Currently, many users cannot access and process information in ways that allow them to answer their questions. Whether they work in law enforcement, for financial intelligence units (FIUs), tax authorities, procurement agencies, anti-corruption bodies, financial institutions, the media, academic institutions, or non-profit organisations, users require:

• Effective access to usable BO information.

This is often still a challenge, especially for data on foreign legal vehicles and individuals.

• Ways to easily retrieve relevant information by being able to flexibly process the data.

Limited options to process BO data and limited search functionalities on BO registers make it difficult, and in some cases impossible, to analyse the information in specific ways.

• Ways to understand relationships between subjects within and across information sources.

This requires sufficient information for users to be able to easily determine whether records about individuals, legal vehicles, and assets refer to the same or different subjects, and to accurately identify whether relationships between these subjects exist within and across datasets. The lack of common identifiers can mean this is often a very resource-intensive process. Data-service providers are critical intermediaries that address these barriers to effective BO data use, but can be unaffordable for some users.

• A minimum level of accuracy to be able to draw conclusions with confidence.

Users require up-to-date information and to be able to understand changes over time. Many users compare records with other information sources to check for discrepancies, which is also a common verification approach for registrars. For this, they need to be able to easily identify inaccurate or missing data which can indicate red flags. This is harder to do if accidental errors in registers’ data are ubiquitous.

Despite many similarities, the research identified a number of different needs in terms of how to access and process BO data. This is largely determined by the type of question the user seeks to answer. Whether individuals use BO information to help manage business risks, investigate tax evasion, or improve public procurement, the specific characteristics of the questions they ask constitute specific use types and user needs. These insights provide the foundation for a conceptual framework which identifies specific user needs based on the following elements:

• The nature of the question users are seeking to answer.

For example, questions can be quantitative or qualitative in nature, which has implications for whether or not users may be able to answer their questions with pseudonymised data.

• The scale of processing required to answer the question. This relates to the number of subjects (e.g. individuals or legal vehicles) users are interested in.

For example, analysis that involves the processing of large quantities of information requires ways to facilitate this. For example, this may require application programming interfaces (APIs) or access to the information in bulk.

• The scope of processing required to answer the question. This relates to the number and variety of connections or relationships between different subjects needed to be identified for users to reach their conclusion.

A small number of use types require a limited scope of data processing. The majority need to identify relationships between individuals, legal vehicles, and assets. This may mean information from a single BO register, multiple BO registers, or additional non-BO datasets is needed. The more extensive the scope of processing required, the greater the need for mechanisms to facilitate entity resolution – the process of establishing whether multiple records about individuals or legal vehicles are referring to the same or different subjects.

• The frequency of processing required to answer the question.

Users who need an up-to-date picture of BO information may require ways to process it on an ongoing basis. This may mean specific register features such as automated alerts, streaming APIs, or the ability to download up-to-date information in bulk are necessary.

The research found that different use types can cut across different user profiles and groups. For example, researchers, tax investigators, and law enforcement officers can all use BO information in the same way, depending on the questions they are trying to answer. In addition, as initial queries may generate new, unforeseen questions and use types, predicting how information will be used and the associated needs can be difficult. Therefore, it does not seem practicable to associate specific use types with specific user profiles and groups in most cases. The findings warrant reframing the narrative around access to BO information from who should have access, to include what this access should look like in order to enable the effective use of BO information. The research suggests that a large group of users should be able to access the information as structured data and use this flexibly.

These findings generate recommendations to help address user needs across the widest range of use types. Many of these echo the Open Ownership Principles for effective beneficial ownership disclosure. [2] Policy makers and agencies implementing BOT reforms can reduce current obstacles, frictions, and resource costs of BO data use by:

• Providing access via well-designed APIs and bulk downloads of up-to-date information to enable use types that require processing data at scale, allow intermediary users to provide services that support end users, and enable information to be flexibly used.
• Expanding the search functionality of BO registers to enable users to find the information they seek in a targeted way. Basic search portals with limited functionality can satisfy some simple user questions, but are unlikely to enable a wider range of use types.
• Enabling users to establish whether multiple records about individuals (or legal vehicles) are referring to the same or different subjects in BO registers by using or assigning reliable identifiers, such as the Legal Entity Identifier (LEI) or register-specific persistent identifiers for individuals.
• Verifying information at the point of submission of BO declarations to guarantee a baseline level of accuracy, such as preventing accidental errors. Capturing information in well-designed digital forms will also help prevent accidental errors and improve accuracy.
• Designing access provisions and data-use policies that allow all users that have a legitimate role to play in achieving impact, and ensuring they can access the information they need and use it in a flexible way. Some jurisdictions may need to put safeguards in place, but where these are excessive or poorly designed these can undermine the impact of BOT reforms.
• Adapting these findings to local contexts and adopting an agile, iterative, and user-centred approach.

Next page: Introduction