CGD Europe recently published a (UK) election manifesto on development with proposals across 19 areas. One area that raised comments and feedback was the proposals on tax. In particular, Alex Cobham of the Tax Justice Network noted the “striking absence of public country-by-country reporting” and argued that this makes “the CGD position less progressive than any major party.”
The set of proposals are not CGD positions, but it is true that public country-by-country reporting is not amongst them. I can’t speak for others who contributed ideas to the manifesto, but the reason I did not suggest public CBCR is this: I’m not convinced by it.
‘CBCR’: What is that?
As part of the G20/OECD-led Base Erosion and Profit Shifting (BEPS) Action Plan, governments have begun requiring that very large multinational companies (those with turnover in excess of €750m) have to submit an annual document showing revenues, profits, assets, employees, and taxes broken down for each jurisdiction where they do business. The idea is to provide a risk assessment tool to revenue authorities to help them to direct their audits. These documents will be shared bilaterally between tax authorities under conditions of confidentiality. The European Commission has gone further and proposes that companies should be required to publish several of the headline indicators (which would cover 6,000 companies with European operations). The Economic and Monetary Affairs Committee of the European Parliament has gone further than the Commission and says that the requirement should extend to cover smaller multinationals with a turnover of more than €40m (around 42,000 companies in Europe). The Financial Transparency Coalition of NGOs also calls for this approach. Labour’s Shadow Chancellor John McDonnell goes a step further and says that the full tax returns of large companies should be made public.
More data, more progress?
Equating progress with disclosure creates a danger of jumping onto a treadmill of perpetually asking for more data without checking if it is working to deliver the hoped for outcomes.
There is a good general case for government data to be ‘open’ (i.e., more comprehensive, timely, and accessible) But there is a difference between government-statistics type information and requirements for private entities to collate and publish data. For the first category 'open by default' makes sense. But for the second category, considerations of privacy, confidentiality, and cost matter.
Nor can we assume that forcing any particular item of data disclosure from private entities is necessarily a progressive advance (think about Theresa May’s short-lived proposal that companies should publish details of how many foreign workers they employ, or the Daily Mail’s call for the government to publish names of benefit recipients for example). Questions of ‘what it’s for’ and ‘will it work’ matter.
Public country-by-country reporting: simple and profound?
Arguments for public country-by-country reporting tend to be broad and sweeping. The MP Caroline Flint says public country-by-country reporting would be “a simple yet profound way to tackle a huge problem of avoidance.” She argues that it would “help developing countries to help themselves” and that the lack of tax transparency is “one of the major stumbling blocks to their self-sufficiency.” However, this seems long on conviction and short on evidence. Her argument that “if developing countries got their fair share of tax, it would vastly outstrip what is currently available through aid” seems to reflect the often inflated perceptions of the scale of revenue at stake from taxing FDI more in poor countries. (Yes, estimates of the scale of revenues at stake from profit shifting are of the same order of magnitude as global aid flows. But no, if countries such as China, Brazil, or Mexico collected a fraction more tax, this would not be used to pay for public services in countries such as Ethiopia, Tanzania, and Bangladesh.)
The European Parliament Committee’s argument for rapidly and radically expanding public country-by-country reporting to some 40,000 companies is circular and incomprehensible:
Most countries use International Accounting Standards. Therefore, public CBCR will be a cost effective way to generate change in global corporate transparency for the benefit of our societies, including citizens, shareholders, tax authorities, investors, economists, and it will give them a means to hold governments and multinational companies to account.
The big question for public CBCR is whether publishing this data in this form will lead to better understanding and meaningful accountability. The risk is that if it doesn’t it will lead to confusion, undermining trust in tax systems which are largely dependent on voluntary compliance.
Transparency and accountability: try, learn, adapt
More general thinking about the case for transparency may be helpful.
Alan Hudson at Global Integrity in a really good post about the intrinsic and extrinsic case for transparency notes that relying on the argument that transparency must be good has led to a situation where the theory of change about the link between data and development remains unclear and under-examined. Hudson argues that unrealistic expectations make it harder to assess what works to inform more effective action. The point of transparency, he says, is that it enhances the ability of communities to try, learn, and, adapt their way towards better development outcomes, through a combination of evidence and politics.
I think it’s try-learn-adapt all the way down. We don't necessarily know what effective open governance (and associated data disclosures) look like, so that is part and parcel of the try-learn-adapt cycle of evidence and politics rather than a linear input to it.
There is a similar question about privacy and confidentiality to the one about openness—is it intrinsically good, or good because of what it enables? The extrinsic case for commercial confidentiality is that it enables entrepreneurs, companies, and investors to undertake their own try-learn-adapt cycles. A certain amount of confidentiality is necessary when making plans, and to protect expensive-to-obtain knowledge long enough to make it worth investing in, (and this is not just valuable to the investors, but because businesses grow by solving problems and meeting needs for consumers). However it is also the case that privacy and secrecy can be used to protect powerful vested interests against legitimate challenges; commercial confidentiality should not be an absolute.
Neither transparency nor privacy are unconditional principles or teams whose flags we need to fly. What is important is that institutions in society—commercial enterprises, investors, public organisations, academic and research institutions, NGOs, communities, and individuals—have the mix of information, privacy, and other enablers and incentives they need to learn and adapt and interact towards better outcomes.
Learning in practice
There is a well-worn path of trying, learning, and adapting around specific approaches to corporate disclosure which winds from pioneering leadership, to industry standard, to mandatory requirements. The idea is that you push and pull a few companies that are particularly vulnerable to reputational pressures, or that see an opportunity from taking early leadership, to recognise a business case for disclosure (on carbon emissions, human rights impacts, extractive industry payments to governments, marine stewardship, and so on). They develop and iterate tools for measurement, reporting, and assurance, which make it easier for others to follow. At the same time an ecology of data users develops; civil society, socially responsible investors, and regulators develop ways to analyse and use the disclosures, to rank and rate companies on their reporting, and eventually on their substantive performance. Both sets of actors learn and educate each other. Consumers, wider investors, and governments may begin to take notice. Companies start to work together, and with civil society organisations to codify expectations into voluntary standards. If it works, the first movers call for other companies to adopt the mechanism, and may support calls for legislation in order to level the playing field. In some cases official guidance, public procurement requirements, or mandatory disclosure legislation follows. This change process results in a rising tide of data disclosures, which might perhaps change behaviour, through internal processes or external accountability or market responses. It is a messy, uncertain, and often conflicted process rather than a simple and profound one.
Proponents of country-by-country reporting have made an admirable effort at trying this pathway. The Fairtax Mark is a voluntary standard which hinges on country-by-country reporting. It was launched in 2014 with the backing of the MP Margaret Hodge (then chair of the Public Accounts Committee) and has been endorsed by organisations including Christian Aid. It had a plan to accredit and license roughly 350 companies over three years. However three years later it has only attracted one FTSE 100 company (which operates only in the UK and Ireland), and only one company that has substantial international operations (Lush Cosmetics). In total it has accredited around 25 enterprises, mainly composed of UK-based co-ops and SMEs. The Fairtax Mark set out to show that a convincing business case could be made for multinational companies to undertake country-by-country reporting. It tried, and what we have learnt is that few are so far convinced. The results of this experiment are an opportunity for learning and adaptation (and does not need to be obscured by jubilantly counting retail branches and offices).
As part of a movement, the Fairtax Mark did contribute to a wider recognition by companies that they needed to start to say something about their tax strategy. Around two thirds of FTSE 100 companies now publish something on their approach to tax beyond statutory accounts, and this has been made into a mandatory requirement.
There is a try-learn-adapt cycle to be run here between business, civil society, regulators, and investors, but it won’t work if it is approached as if the correct outcome and template for reporting are already known, and the role of civil society is simply to herd companies with carrots, sticks, and leaks towards the pre-established goal of public CBCR.
Extractives are different
One reason that general country-by-country reporting seems like an obvious solution is because there is a strong case in the extractive industries. But it is worth considering why the different situation of these sectors.
The extractive sector is different because natural resources are part of a countries’ national wealth. The social benefit they generate, and a companies ‘social license to operate’ comes largely through fiscal contributions, made up of taxes and royalties. Access to mineral resources is generally through public contracts, and there are large risks of mismanagement and corruption at every stage. At the same time companies are faced with making large sunk investments in often politically unstable locations which are then subject to ‘obsolescing bargain’ problem that the government (or subsequent governments) may go back on the deal.
The campaign for revenue and contracts disclosure in extractives was driven by the idea that public transparency that would help all parties. If citizens had access to information about natural resource revenues and contracts they could ask better questions about the deals being done on their behalf, and how the revenues are spent. For businesses, if the fiscal regime, contracts, and revenue streams are subject to public scrutiny and social approval this will increase long term certainty by reducing the risk of social unrest, unmanageable expectations, renegotiation and fiscal hikes or expropriation and nationalization.
It is not obvious what the right amount of tax should be just by looking at the value of minerals exported, but extractives projects can be modeled using information from contracts, and knowledge of oil and mining value chains. Questions about whether it is better to collect more tax early in a project’s life, or later once it is profitable, and about how the fiscal system deals with commodity price volatility can and should be debated in public. Furthermore extractives revenues are large enough, and the companies few enough that this kind of public scrutiny is feasible (although not easy).
So, for companies in the extractive industries there is a strong case for demonstrating that they are contributing the revenues that their host societies expects of them, and country-by-country reporting may provide a means to encourage tax certainty, which benefits both businesses and society.
This argument does not transfer simply to other sectors. For companies whose business models are not based on public concessions and standardized exchange-traded products the headline indicators might not translate into reasonable expectations in the same way (for example see the case of Inditex). The amount that ‘society expects’ (however vague) is often not the same as the amount the law requires, but instead falls back on intuitions such as linking the location of consumers with the tax that should be payable.
In other words, for companies in other sectors it is not clear that CBCR would enable them to demonstrate that they are paying ‘what society expects them to,’ and the resulting controversies could increase reputational risk and tax uncertainty and undermine trust and tax morale.
If not country-by-country reporting, then what?
The try-learn-adapt approach suggest that we focus on objectives. It is worth unpacking the hopes attached to CBCR:
Government: Better tax policy and tax policy making
Revenue authority: Better administrative performance, trust in the revenue agency
Individual Companies: Taking a responsible approach to tax practice and risks, contributing openly to public debates on tax
Roles for taxpayer ombuds, supreme audit institutions, legislative scrutiny, citizens juries; publishing corporate tax strategies; and enhanced reporting about tax risks and reconciliation are just a few ideas for improving citizen engagement and trust. Abandoning the conviction that a rapid rollout of public CBCR is the only answer frees up space for debate, thinking, and experimentation about what kind of data and mechanisms (and perhaps more importantly what kinds of human and organisational interactions) could help to achieve these objectives—both in the UK and in developing countries.