‘One level up’ is EITI sweet spot for beneficial ownership
This is post is cross-posted at the EITI blog
As there was a lot of talk around beneficial ownership within the new EITI standard, we thought we’d try and model what it could look like. It’s one of those thorny issues. Civil society wants it, corporates often state their belief that it is a heavy reporting burden – but the arguments on both sides can tend to be at a theoretical level. So we thought we’d try and ground it a little (report here, data here), by using Norway as a proxy for what kinds of data that would yield, and to what end?
What we were seeking is to find a sweet spot which would impose very little reporting burden on companies and yet which could demonstrably yield value for the EITI process because it could answer questions. So we adopted what has been called the “one level up” principle – to look at what would happen if all EITI reporting companies included their own legal names and corporate identifiers, their shareholders and their corporate identifiers, and the jurisdiction of their parent companies. Any reasonably run company should be able to access this information in less than an hour!
The ‘first mile’ of corporate data – the Tanzanian company register in government offices in Dar es Salam
“Tullow” as it is referred to in media reports or ordinary conversation (and some current EITI reports!) but the full name of the particular corporate vehicle operating in that country (such as “Tullow Uganda Holdings B.V” in Uganda). And the identifier is a unique number that the company has within any company registry, much as individuals have unique tax and national insurance numbers. These are the bits of information which allow a particular corporate vehicle, the one which is legally responsible for signing the contract and observing its conditions, to be uniquely identified.
We restricted to one-level-up because it avoids the debate around finding ultimate ownership. Our friends at OpenCorporates.com actually claim that the concept of “ultimate ownership” is in any case misleading, since in many, perhaps most, cases we are not talking about a single dimension of ownership, travelling from local affiliate up to source like some explorer finding the source of a river. What happens if a company four steps up the chain has a cross-holding in another one three stages “higher”, for example?
We could model Norway because the EITI reports were in Excel format, the legal names were included, and the Norwegian company register can provide the corporate identifiers and shareholding structures of the parents. We took the companies in the 2011 EITI report and then matched them to these records and where possible, established the jurisdiction of the parent holdings. Finally, we related these company structures back to the size of payments reported within EITI.
The results are not particularly surprising in the case of Norway. The oil sector is dominated at both affiliate and parent level by companies which are incorporated in Norway itself – 64 percent, as you can see from the diagram.
Then again, many of the international companies had at least two levels of local vehicle: ExxonMobil Exploration & Production Company AS, for example, which is a Norwegian company, is owned by ExxonMobil Norway Upstream Holdings Inc, also a Norwegian-registered company with its own number in the Oslo registry. There might well be other Norway-based companies further up the chain. And yet one would imagine that this chain ultimately stretches back to ExxonMobil in the USA.
This illustrates a key point in the positioning of EITI. It’s the first mile of company information which EITI can most painlessly provide – and which is the most useful and needed. For instance, while the records in Norway are already online and available with a simple web scrape, the picture at the top of the blog shows you what we start wth – what the “first mile” of corporate ownership looks like in, for example, Tanzania, in the offices of the company register in Dar es Salam.
But once this level of information is there, it is likely to lead out into the global corporate web which is increasingly coming into the public domain. Even if a parent company is in a tax haven like the Cayman Islands, we at least know that it is there, and what its address is because the index to the Cayman Islands, along with around 70 other jurisdictions around the world, are digitised and online.
The world is full of “data wranglers” who can connect the first mile data to the rest of the growing repository of public domain data.
What immediate applications could the “one level up” approach be used for? One might be to indicate potential black spots in double taxation treaties. The graph from Norway shows 35% of sector activity controlled by foreign companies in some 16 other countries – what percentage of that is in juridictions that Norway has double taxation avoidance treaties? This may be a case which becomes more interesting in many EITI countries which may not have developed comprehensive treaties.
Another application is to be able to create a roster of parent companies – and their publicly listed company officers and shareholders – who represent the beneficial ownership of an extractives sector in a country as a whole. When available for public view, this can offer a powerful tool for civil society and local media to evaluate for possible connections with decision makers in government. Such data are also of great interest to new entrants in the private sector who are trying to understand the landscape for potential investment, one of the reasons the World Bank has sponsored digitisation and upgrading of company registers in over a dozen countries around the world to date.
We therefore believe “one level up” would position EITI well to contribute great value with a relatively small data set with minimal reporting burden.