Beneficial ownership data in procurement

Operationalising the use of beneficial ownership data in procurement

Operationalising the use of BO data in procurement requires making decisions about a number of things, including where to source BO data from, how and when to collect it, and how to verify it and whether it is published. These will all have an impact on data quality and useability. How and in what format data is collected will also affect its ability to be linked to other datasets, which may allow for specific types of analysis. The following section outlines some key considerations for implementers.

Data collection and coverage

Broadly, governments implementing the use of BO data in procurement can either buy commercially available BO data or collect it themselves in a centralised register. The latter is good practice, as this means that different user groups (including procurement authorities), can access the information through one central location in a standardised format. This removes some of the practical and cost barriers to accessing and analysing BO information. [79] If governments collect the data themselves in a central register to aid procurement, there are two main options:

1. governments can collect data during the procurement process and hold this in a central procurement-specific register;
2. governments can pull data from a central BO register that covers all sectors of the economy into procurement processes.

In both cases, there are practical and technical considerations to maximise the effectiveness of disclosures. Open Ownership (OO) has developed a set of principles for effective disclosure (see Box 5) for governments aiming to implement central registers, which also apply to procurement-specific registers.

Commercially available BO data

BO data from private third-parties have inherent limitations. Private sector BO providers building proprietary datasets often have limited visibility on (or rights to distribute) full ownership data, and often rely on inferring from partial information. They cannot coordinate the knowledge held by beneficial owners, shareholders, and companies or guarantee the coverage in the way that a government with the right powers and mandate can. [80] For example, the EU’s ARACHNE system uses Bureau van Dijk’s ORBIS data [81] – a commercial dataset which includes BO data – to detect potential conflicts of interests. However, there are debates about the quality of ORBIS data, and some data users have pointed to gaps in its coverage of certain countries (Germany, for instance). Whilst ORBIS covers a huge number of companies, the varying coverage per country highlights that BO data from private suppliers is no substitute for governments legislating for companies to disclose their BO. [82]

Central registers

A number of governments collect BO data themselves specifically for procurement (see, for example, Box 6). In this case, BO information can be collected at different stages of the contracting process (see Figure 1). Governments must ensure their disclosure regime comprehensively covers the legal entities to fulfil their policy aims. How and when data is collected comes with different advantages and disadvantages relating to the accuracy and quality of the data, as well what the data can be used for.

To illustrate this, if jurisdictions do not have a central BO register but want to use BO data to prevent fraud and corruption, they must collect the BO information of all entities submitting bids, in addition to conducting BO checks on the contract awardee, otherwise they will not be able to identify bid rigging. The UK’s proposed register of beneficial owners of overseas companies and other legal entities intends to collect BO data on foreign entities seeking to bid for UK government contracts over a certain threshold. However, the proposal is to only collect this data for contract awardees, making it impossible to detect bid rigging by cartels. [83] Collecting BO information for the sole purpose of procurement may also restrict governments from using BO data in other policy areas.

Governments can also collect information in a central register on all legal entities in a jurisdiction as part of a BOT regime and subsequently use this data in procurement. In this case, BO information can be collected and checked at different company lifecycle and procurement stages. Economy-wide BOT can allow for better visibility of full ownership structures when sufficient detail is collected. This can reveal other legal entities in an ownership structure that would not be revealed in procurement-specific BO data collection.

A key consideration for implementers is whether the disclosure regime comprehensively covers all companies relevant to procurement, including, for instance, foreign entities. Using data from other jurisdictions, where available, may not be feasible in the near future, as aspects like legal definitions and standards on verification differ. The UK, for instance, has proposed a separate register for foreign entities wishing to do business with the government, in addition to its central register for the BO of UK legal entities. [84] Implementers may face additional challenges when verifying the BO of foreign entities, such as verifying the identities of foreign nationals.

Collecting BO data on all legal entities also allows governments to use BO data in a number of other policy areas (for instance, AML), where some countries may already be under international obligations. In the Czech Republic, for instance, contract awardees were required to submit BO information as prerequisites for contracting. Following the implementation of a centralised BO register, as required by the EU Fourth and Fifth Anti-Money Laundering Directives, the Czech Republic removed its requirement to submit BO data in procurement. Instead of the supplier needing to provide this information as part of the procurement process for every bid, contracting authorities search the relevant information on the Register of Beneficial Owners, maintained by registry courts. [85]

Currently, not many governments appear to be using BO data in procurement. Where this is done, it is not done systematically, and appears to be done in an ad hoc fashion and not broadly publicised. The World Bank’s 2020 Good Practices Guide for Preventing and Managing Conflicts of Interest in the Public Sector, for instance, only mentions Chile, Indonesia, Moldova, and Ukraine as collecting this data as standard procedure. [86] In contrast, of the few other countries that do have central and public BO registers, most do not appear to be using them systematically in procurement. How Ukraine collects BO data on all legal entities as part of its BOT regime, and subsequently uses this in its procurement processes, is detailed in Box 6.

Public access

Publishing BO data – for procurement or otherwise – stimulates broader data use and scrutiny that is likely to drive up data quality. The publication of data can also have a deterrent effect. Therefore, in order to get the maximum benefit and full utility from BO data in procurement, including its systemic and indirect effects, governments should collect, verify, and publish BO data centrally. This also allows governments to use BO data in other policy areas.

Whilst there is a case for making data public to maximise effectiveness of BO data in public procurement, there are legitimate privacy concerns over publicising certain characteristics of individuals and the unequal impacts these may have. As a principle, governments and companies should not collect and disclose data beyond the minimum that is necessary to achieve their policy aims, or data that poses a significant risk of harm. [90] Governments should be mindful about what data they collect and the fields they publish, and consider exceptions on a case-by-case basis where credible threats to an individual emerge from the publication of data. [F]

Structured and interoperable beneficial ownership and procurement data

Internationally, there has been a growing trend towards e-government and e-procurement. [91] Operationalising the use of BO data in procurement is more easily done through integrated digital technologies rather than through paper based systems. In order to maximise the potential benefit of using BO data in procurement, it should be collected and stored as structured, interoperable and machine-readable data, which can be analysed easily and cheaply. When BO data is combined with other open and structured datasets, such as open contracting and spending data, analysis can provide powerful insights into procurement practices, consumption models, and supplier transactions.

A number of data standards have been developed that can be used off the shelf for implementing governments, which facilitate easy exchange of data between systems. The Open Contracting Data Standard (OCDS) was developed by the Open Contracting Partnership for this purpose, and is implemented around the world. It is built on the idea that it should be possible to follow a contracting process from planning to implementation through a unique ID. [G] For BO data, OO has developed the Beneficial Ownership Data Standard (BODS), which provides a structured template for describing BO as machine-readable data, laying out key data points for implementers to collect. [H] OCDS and BODS data can be easily combined.

When combining datasets, it is key to be able to identify individuals and companies across them. Matching people across datasets by using identifiers such as names is unreliable and labour intensive, and not viable for larger datasets. [92] A better approach is to use unique identifiers. Combining datasets allows for analysis that is otherwise not possible. For instance, without linking procurement data to spending data, it is not possible to comprehensively analyse a procurement system.

By using national identification numbers in BO disclosures as unique identifiers, Global Witness was able to link jade mining concessions to government officials in Myanmar. [93] Governments implementing BOT, like Nigeria and Jordan, are developing unique identifiers for their own registers.

When these different datasets are expressed in common formats, they can be incorporated into standard workflows, and the analysis can be automated. This is demonstrated through the Bluetail prototype, which connects contracting and BO data for multiple bidders as well as identifying red flags (such as the same beneficial owner appearing in multiple bids). The Sinar Project in Malaysia, a civi-tech organisation that uses open data sources to hold its government accountable, developed Telus. Telus is a web service application that imports and joins up open data sources for PEPs (using the Popit API), BO (using BODS data from the Global Register), and procurement contracts (using OCDS) to expose conflicts of interests in procurement. [94]

Figure 4. Bluetail: Prototyping the use of structured beneficial ownership data in procurement

Open Ownership has built on a prototype developed by mySociety and Spend Network called Bluetail, which shows how BODS and OCDS can be used to automatically raise red flags for corruption and collusion risks when procurement officers screen tenders. For more information see: www.openownership.org/blogs/tps-prototyping. Source: bluetail.herokuapp.com/tenders.

Verification

To maximise the impact of BO data, it is important that data users and authorities can trust that the data contained in a register broadly reflects the true and up-to-date reality of who owns or controls a particular company. This can be done through verification [I] (defined as the combination of checks and processes that a particular disclosure regime opts for in order to ensure the BO data is of high quality, meaning it is accurate and complete at a given point in time). This can include checking data conforms to known and expected patterns, cross-checking data against other government-held datasets, and frequently checking if data is correct.

Data should be verified on submission and updated – or confirmed that it still holds true – on a regular basis. Procurement officers want to be confident that the data is correct at the time of decision making, so data should at a minimum be (re)verified at that point in time. There will be more opportunities to verify data during a company life cycle in a central database than when data is collected and held for procurement alone. [95] Making the data public drives up data use and allows for additional verification from civil society, the private sector, and the general public.

Sanctions and enforcement

When governments use BO data in procurement, they can drive up compliance to a BOT regime by imposing sanctions relating specifically to procurement. A number of countries have implemented sanctions for the failure to provide or providing incorrect BO data (see, for example, Box 7). These sanctions range from preventing companies and their beneficial owners from signing contracts, or debarring them from being involved in procurement for a specific period of time. Sanctions are not a substitute for proper due diligence in advance. Debarring certain companies may lead to procurement shortfalls, but the presence of sanctions will help drive up compliance. Full BOT allows governments to more effectively enforce sanctions, such as debarment from procurement, as it becomes more difficult for individuals to hide behind complex ownership structures.

Footnotes

[79] “Principles”, Open Ownership, last updated November 2020, https://www.openownership.org/principles/central-register/.

[80] Jack Lord, “Beneficial Ownership Registers: who knows what?” Notes on open data and transparency, 8 November 2018, https://www.opendatanotes.com/post/beneficial-ownership-registers-who-knows-what-when/.

[81] “Arachne: Frequently asked questions”, European Commission, 7 August 2017, http://ec.europa.eu/social/BlobServlet?docId=15097&langId=en.

[82] Che Sidanius, “Why firms want more company ownership data”, LinkedIn, 13 May 2020, https://www.linkedin.com/pulse/why-firms-want-more-company-ownership-data-che-sidanius.

[83] “A Register of Beneficial Owners of Overseas Companies and Other Legal Entities: The Government Response To The Call For Evidence”, BEIS, March 2018, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/681844/ROEBO_Gov_Response_to_Call_for_Evidence.pdf.

[84] Paul Scully, “Register of Beneficial Owners of Overseas Entities Update”, UK Parliament, 21 July 2020, https://questions-statements.parliament.uk/written-statements/detail/2020-07-21/HCWS413.

[85] Kristína Bartošková, “Amendment to Regulation of Beneficial Owner Identification in Czech Public Procurement”, Baker McKenzie, n.d., https://publicprocurement.bakermckenzie.com/pages/czech-republic-beneficial-owner-identification/.

[86] “Good Practices Guide: Preventing and Managing Conflicts of Interest in the Public Sector”, World Bank, July 2020, 50, http://documents1.worldbank.org/curated/en/950091599837673013/pdf/Preventing-and-Managing-Conflicts-of-Interest-in-the-Public-Sector-Good-Practices-Guide.pdf.

[87] “Where do we go from here to stop the pandemic?”, Transparency International, 29 April 2020, https://www.transparency.org/en/news/where-do-we-go-from-here-to-stop-the-pandemic.

[88] “Guide to Beneficial Ownership Information in Ukraine: Legal Entities and Legal Arrangements”, Global Forum on Asset Recovery (GFAR), December 2017, https://star.worldbank.org/sites/star/files/bo_country_guide_ukraine_english_final.pdf.

[89] David Marghania and Tato Urjumelashvili, “Open Contracting in Ukraine: a collaborative effort for procurement reform”, OCP, 2 June 2015, https://www.open-contracting.org/2015/06/02/open_contracting_in_ukraine_a_collaborative_effort_for_procurement_reform/.

[90] “Data Protection and Privacy in Beneficial Ownership Disclosure”, The Engine Room, Open Ownership, and The B Team, May 2019, https://www.openownership.org/uploads/oo-data-protection-and-privacy-188205.pdf.

[F] In the UK, for instance, individuals can apply for their information to be concealed from the public register, if personal characteristics of a person when associated with a company put a person “or any person living with them, at serious risk of violence or intimidation.” See: “Applying to protect your personal information on the Companies House register”, GOV.UK, 16 September 2020, https://www.gov.uk/guidance/applying-to-protect-your-personal-information-on-the-companies-house-register.

[91] World Bank, “Benchmarking Public Procurement 2017: Assessing Public Procurement Regulatory Systems in 180 Economies”, 28.

[G] For more information on the Open Contracting Data Standard (OCDS), see: www.open-contracting.org/data-standard.

[H] The Beneficial Ownership Data Standard (BODS) has been developed to serve as a solid conceptual and practical framework for collecting and publishing BO data, and enabling the resulting data to be interoperable, more easily reused, and higher quality. BODS has been developed in collaboration with dozens of international experts in company data and in technical standard-setting, across civil society, business, and academia. A separate tool has been developed that allows the data to be visualised. See: standard.openownership.org.

[92] Alex Parsons, “Using private IDs in public: Options for improving deduplication in ownership data”, mySociety Research, July 2020, https://research.mysociety.org/media/outputs/Using_private_IDs_in_public__CwJQbXH.pdf.

[93] “Jade: A Global Witness investigation into Myanmar’s ‘Big State Secret’”, Global Witness, October 2015, https://www.globalwitness.org/jade-story/.

[94] “Telus”, Sinar Project, n.d., https://sinarproject.org/transparency/telus.

[I] For more information, please see Open Ownership’s policy briefing on the verification of BO data: Kiepe, “Verification of Beneficial Ownership Data”.

[95] Alex Parsons, “Collecting and making use of beneficial ownership data”, mySociety Research, July 2020, https://www.mysociety.org/2020/07/31/collecting-and-making-use-of-beneficial-ownership-data/.

[96] Kiepe et. al, “Early impacts of public registers of beneficial ownership: Slovakia”.

Next page: Conclusion