A review of the data collated so far from the UK PSC Register
UK companies with hidden owners lie behind some of the worst problems of our time. Among other things they have been used to breach UN sanctions, evade taxes and commit fraud. That is why Global Witness has been calling for an end to anonymously owned companies for years. Global Witness is a non-profit organisation that exposes the hidden links between demand for natural resources, corruption, armed conflict and environmental destruction.
In June 2016, the UK government started publishing the PSC or 'People with Significant Control' Register, the world's first open data register of the real owners and controllers of companies – known as 'beneficial owners'. It is a big step forward for transparency in this country and a potential treasure trove of information – but we need to make sense of it first.
That is why Global Witness has collaborated with DataKind UK, OpenCorporates, Spend Network and OCCRP to probe it in detail and give feedback on how useful it is. So on an autumn weekend in London, we brought together 30 volunteer data scientists, including the head of software for a Formula One team, an astrophysicist and a former consultant to the UK tax authority.
We had three aims: firstly, to get some basic insights into this data and what it could tell us about companies in the UK and secondly, to understand the flaws and gaps in the new register. Finally, we were curious to see whether the register could yield investigative leads. Here is a snapshot of what we found:
- Almost 3,000 companies listed their beneficial owner as a company with a tax haven address – something that is not allowed under the rules.
- There are problems with how the data has been inputted. For example, you can write anything in the nationality field and we found over 500 ways of putting 'British', including ten people who wrote 'Cornish'.
- 76 beneficial owners share the same name and birthday as someone on the US sanctions list. We have to do more digging to find out whether these are the same people, but it is an insight into what is possible with this new information.
- Most beneficial owners are from the UK, followed by a number from other European countries and India and China.
The data enables us to begin to understand the complex ecosystem of companies and their owners in the UK – where they are from and what other companies they own. In theory, the ability of law enforcement agencies, journalists and civil society to map this out will make it much harder for those with something to hide to do so behind layers of secret companies.
Under the new system, companies must provide ownership information at least once a year when they send in their annual return or confirmation statement. Out of 3.5 million companies, 1.3 million have already filed beneficial ownership information as of 17 November 2016.
Most beneficial owners (1.2 million) are British and unsurprisingly when you take the British out of the mix, Irish is the next biggest nationality, followed by a mix of European countries (Italy and Poland) and then China and India. On top of nationality, we also have information on where beneficial owners live. The vast majority are based in the UK and 183,000, or 13% of the total, live overseas.
One of the arguments made against creating these registers is that companies would find it difficult to identify their beneficial owners, but this does not appear to be the case. In only 2% of cases did companies say that they were struggling to identify a beneficial owner or collect the right information.
Just less than 10% of companies claimed to have no beneficial owner. This is possible under the legislation, because a key threshold for being considered its beneficial owner is whether you own at least 25% of a company. That is, however, quite a high threshold, which could be exploited by people looking to stay under the radar.
Approximately 30 beneficial owners have been successfully granted the right to keep their name off the register due to concerns about their security.
We were also able to build up a picture of company networks. These were only partial insights as we were working with just three months' worth of data, but it gave an idea of what was possible, and an understanding of complex corporate structures. Complexity does not necessarily indicate wrongdoing, however, this work will provide a much better map of how UK companies structure themselves and, in some cases, connections between companies may raise further questions about what they have been up to.
Inevitably with a bold experiment like this, there will be problems with the data, despite the excellent job that Companies House has done. The flaws that we identified split into two categories: messy data due to how it was collected and examples of what look like non-compliance with the law.
One of the big problems we found was that Companies House allows free text entry for a number of fields. An easy way to avoid this would be for Companies House to use drop down menus for fields such as nationality or title, with a potential 'other' box.
Another big challenge is the lack of unique identifiers for individuals and some companies. We were relying on using name and day and month of birth to try and identify individuals. This is fine if you are someone with an unusual name, but proves more difficult for someone with a common name, such as John Smith or Mohammed Iqbal. One solution could be to assign beneficial owners a unique number to allow better cross-matching.
There were 2,160 beneficial owners born in 2016 and also people who listed 9988 as their year of birth. This suggests that Companies House should consider doing some basic validation to improve the quality of the data.
On top of these data validation issues, we found a number of signs of non-compliance with the law. For example, 9,800 companies listed their beneficial owner as a foreign company. This is possible if the foreign company was listed on one of the stock exchanges deemed equivalent to the UK system (e.g. the US, EU and Japanese exchanges).
However, there were nearly 3,000 companies with tax haven addresses listed as beneficial owners, which is almost certainly not allowed under the rules. We will be handing this list of companies over to Companies House to investigate further.
With all of these examples it is not clear whether this is because the people filing the information did not understand the new system, or whether there is something more sinister going on.
One of the key arguments in favour of creating this register was to uncover wrongdoing. We did not expect to find many cases in just a weekend, but we did find some interesting leads – because the UK register is available as open data, we could compare it to other data sets.
Initial findings suggest 19 senior politicians (known as politically exposed persons), 76 people from the US sanctions list and 267 disqualified directors were listed as beneficial owners. However, these matches were based on name and month and year of birth. This means that it is hard to be certain that the matches are good ones. Any leads coming out of the new register will have to be followed up with more traditional investigative digging.
We also compared the beneficial ownership to government spending information. It was challenging given that a lot of the spending information does not have proper identification for companies (for example, it would be good to include company numbers), but we did find a number of recipients of government contracts who had beneficial owners based in tax havens.
The work was only possible because the UK Government made the decision to make the register free and accessible to all as open data, which is an important lesson for other countries that are thinking about creating similar registers.
Inevitably the data was messy, partly due to how the register is set up and partly due to confusion from those providing data. Global Witness is going to be working with Companies House to iron out some of these issues and hopefully as the register beds in, these kinks will resolved.
A big question that we were not able to answer was whether people had purposely or otherwise provided inaccurate information. This is going to be a serious challenge for the credibility of the register and the government will have to provide more reassurances on how it is going to ensure data accuracy.
There is clearly a demand for this sort of information from the business community. In September, Global Witness reported on a group of institutional investors that were calling for greater business transparency in order to manage risk better. It is also part of a broader trend for businesses to be more transparent about how they run themselves – something that we believe is good for consumers and companies.
Finally, we are hoping that other governments will see the potential of what the UK has done. The process will be difficult, but if our investigation has shown us anything, it is that this is incredibly useful information to have.
Robert Palmer is Campaign Leader, Anti-Money Laundering, at Global Witness
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.