How policy informs technology, and technology informs policy: building a data standard for beneficial ownership transparency
In this blog post, we reflect on our work on the Beneficial Ownership Data Standard, and share what we’ve learned about how technology and policy can work together. You can also read Thom Townsend, Executive Director of Open Ownership, reflect on our work together on the Open Ownership blog.
In 2016, shortly after the ICIJ published the Panama Papers, Open Data Services began working with Open Ownership to build the Beneficial Ownership Data Standard (BODS). BODS is an open standard for publishing high quality data about how companies (and other legal entities) are owned and controlled.
When we started working with Open Ownership, there was no systematic way of collecting data about beneficial owners. To build BODS, we’ve worked across the full data pipeline, from policy and regulation, through to data standard design and publisher support. As a result, our work on the standard has fundamentally shaped how we understand good beneficial ownership policy, and our work on policy has fundamentally shaped how we build the standard.
In this blog, we reflect on our work with Open Ownership and share some of our thoughts about how technology and policy can work together to create a data standard that achieves public policy goals.
Putting usability at the centre of policy
Open Ownership had thought about collecting beneficial ownership data by using a data standard from the start. As there were very few regulatory frameworks in place for collecting this information in 2016, we were able to concentrate on how people needed to use the data at the same time as many disclosure mechanisms were being developed.
For example, the legal enforcement of beneficial ownership transparency turns on the concept of declarations — someone stating something about a company structure at a specific time. If the declaration can be shown to be false, then the person or company stating the falsehood is legally liable. We knew we needed to make it clear in the data model that ownership statements represent information that has been declared to a registry, not necessarily the truth. As a result, the core data objects in BODS don’t represent companies and people — they represent statements about companies and statements about people.
In order for BODS data to be useful, people need to be able to connect disclosures from different jurisdictions to reveal global ownership patterns. This meant that it was crucial to ensure the standard could handle a range of diverse ownership structures and legal declarations.
By taking this user-first approach, we were able to help countries develop legislation that encouraged the creation of useful, usable data. In Armenia, for example, we gave public officials advice about improving declaration forms. These changes made the forms easier to fill out, and helped to collect better information about company structures.
Using early analysis as a powerful advocacy tool
Once there was some data to work with in the form of the UK Persons of Significant Control Register, we found that early analysis can be a powerful advocacy tool. The joint report by Open Ownership and Global Witness Learning the Lessons from the UK’s Public Beneficial Ownership Register proved that access to open, structured beneficial ownership data can help us understand the pitfalls of data collection, and work to improve those mechanisms.
For example, the declaration form used to collect data required people to type their nationality into a free text box, the report identified over 500 different spellings of the term British, with 10 beneficial owners listing their nationality as Cornish. As a result of the analysis, we could recommend registries use a drop down list of pre-populated countries, rather than a free text field for users
In turn, these findings fed back into the development of the standard. The analysis of the Persons of Significant Control Register helped us to understand that we need to reduce scope for making honest mistakes, which can obscure disinformation in the register.
Developing resources and publisher guidance
Creating a data standard is as much about solving coordination problems as it is technical problems. A key part of making BODS useful was understanding the support implementers need to publish data.
In part, this means understanding the stages a jurisdiction goes through on the way to implementing Beneficial Ownership Transparency. After a jurisdiction has agreed to publish standardised beneficial ownership data, we support them in understanding how this work gets done — for example, ensuring publishers understand why it is important to use common identifiers for foreign companies.
To support implementers, we’ve developed a number of resources, including a disclosure scope workbook to help implementers think through ownership and control structures and guidance and templates to help systems designers create effective beneficial ownership declaration forms. The aim of all this work is to knit policy and technology together, providing the support people need to get publishing.
What comes next?
Good data standards are a collaboration between policy and technology. As Open Ownership thought about this dynamic from the start, we’ve been able to work to build a data standard that gives the best chance of collecting good quality, interoperable beneficial ownership data across jurisdictions.
We’ll be continuing our work with Open Ownership to support the development of BODS Version 1.0, build their internal technical capacity, and respond to the needs of people who use BODS.