Andrew Baker is a Faculty Professorial Fellow in SPERI, Sheffield Political Economy Research Institute. This blog follows his recent work with Richard Murphy to develop a new framework to assess tax spillovers.

There is growing interest among a range of development non-governmental organisations (NGOs) in something called ‘tax spillover’ analysis.  In the last two months, co-author Richard Murphy and I, have engaged in discussions with Oxfam, Action Aid and Christian Aid, as well as UK parliamentarians about spillover analysis.  There is a very clear growing demand for enhanced forms of tax spillover analysis.

The notion of tax spillovers originates in a piece of research conducted by the International Monetary Fund (IMF) in 2014.  Tax spillover refers to the impact of one country’s corporate taxation rules and practices on other countries.  The basic notion is that the structure and practices of one corporate tax regime can spill across national borders effecting macroeconomic performance elsewhere and the distribution of wealth across countries.  The IMF research noted that spillovers have a ‘significant and sizable’ impact in reducing corporate tax bases and rates, especially in ‘developing countries’ (2 or 3 times higher than in OECD countries) because they typically derive a greater proportion of their revenue from corporate tax.  In this sense, the whole concept of tax spillovers revolves around the idea that the practices and policies of advanced countries can be harmful for the revenue raising activities of developing governments and the overall performance and size of their economies.  How best to measure and document spillovers and the implications of this for global governance is what is now at stake.

Reliable ten-year data on 99 states worldwide shows that 65 have cut their headline rate of corporation tax by an average of seven per cent during that period (2006-2016), with the average headline rate falling from 27.45% to 20.73%.  Headline rates of corporation tax are however, a notoriously unreliable guide to the levels of corporate tax actually being paid in many jurisdictions.  Moves to reduce corporate taxes, through tax breaks, tax holidays, aggressive tax planning strategies and other loopholes, will in all likelihood result in real and effective rates far lower than this surface data can capture.  Headline rates are however an important signalling device.  This data is effectively telling us that a majority of the world’s states participate in a global competitive game by reducing their corporate taxes.  Because of this many NGOs believe spillover analysis can be a useful instrument for documenting and abating this race to the bottom.

The IMF spillover analysis was an impressively robust 86 page document and a useful first step forward in establishing the concept.  It also successfully established corporate tax spillovers as a real world phenomenon with concrete economic effects.  Yet the development NGOs we have spoken to remain sceptical about the IMF approach and are searching for alternatives.

Two factors are at the root of this scepticism.  First, is the question of whether econometric equations using panel data deployed by the Fund can ever fully capture spillovers.  In our revised spillover framework we propose evaluating the relationship between corporate taxes, a variety of other taxes and administrative procedures and practices that constitute tax jurisdictions in their entirety.  In contrast, the Fund analysis confines itself to two categories of spillover: base spillover – where taxable profits in a jurisdiction shrink through profit shifting activities and reduced investment; and strategic spillover, – the act of lowering corporation tax in response to developments elsewhere.  The IMF authors acknowledge that their empirical analysis has significant limitations due to patchy data on allowances, especially in developing countries, meaning average effective rates of corporate tax are often much lower than statutory rates (p.52).  The risk here is that the Fund’s approach significantly underestimates spillover effects and their costs to developing countries.  This was partly confirmed by Ireland’s spillover analysis, which claimed there were no identifiable spillover effects on developing countries from its corporate tax regime, despite a well-known aggressive re-routing avoidance scheme known as the ‘double Irish with a Dutch sandwich.’

The second factor is one of turf, or who gets to speak ‘authoritatively’ on spillover analysis.  If the Fund’s equation becomes the accepted and established method for calculating spillovers (and it does enjoy first mover advantage) that would also cast the Fund and its staff, given their skill sets, as primary interlocutor on tax spillover issues.  This in turn raises questions about whether the politics of the Fund and its organizational culture are development friendly.  At best the Fund’s record and reputation in developing countries could be said to be mixed and chequered.  It also raises questions about how the Fund itself might use spillover analysis and for what purpose.  A well known joke in Fund circles is that IMF stands for ‘It’s Mostly Fiscal’.  Crucially, the Fund provides a source of capital for countries experiencing fiscal distress and uses loan agreements to promote ‘sustainable public finances’.  In the hands of the Fund, there is a fear that spillover analysis could simply become an instrument for improving fiscal balances in developing countries, rather than enhancing the wealth and well-being of these countries.  The appropriate global governance of spillover analysis, requires a much fuller debate.

Our initial proposal for a revised form of spillover analysis, produced for the UK parliament’s All Party Parliamentary Group on Inclusive Growth is intended to inform debates on appropriate global tax governance.  Over ten years ago, my co-author Richard Murphy wrote one of the first templates for something called ‘Country by Country reporting’ (CBCR), which is designed to increase the obligations on multinational corporations to report what economic activity takes place where.  Beyond complex debates about the technicalities of CBCR, the process of translating it into domestic and international law through the OECD, the EU and the G20, does appear to be increasing transparency and changing appropriate behavioural norms.  For example, our conversations with senior figures in the corporate world does suggest reluctance to route profits through tax havens is growing due to the existence of CBCR.

When the G8 announced its support for country by country reporting at the Lough Erne Summit in 2013, there was a crucial inconsistency in the stance taken by David Cameron who chaired the Summit.  On the one hand, Cameron was launching a new international standard to make international corporate tax avoidance activities more difficult.  On the other hand, he was also lauding plans to reduce the UK’s corporate tax rates to the lowest in the G20. Scrutiny of corporations’ accounts was increasing, but many governments continued to facilitate a systematic lowering of corporate tax payments.

The new tax spillover framework we have proposed is intended to subject tax administration authorities and their procedures to closer scrutiny, as a counterpart to the emerging CBCR regime, addressing the inconsistency revealed at Lough Erne.  What my co-author and I have essentially written is a prototype first draft of a form of country by country reporting for governments and tax authorities.  It is intended to provide the basis for an ongoing form of international peer review that changes conceptions of appropriate state behaviour.  We are at the start of a process of working with a variety of stakeholders and experts to develop and refine this framework over the coming years.

3 Comments

Nick Meyne - 09.06.17

To your second point on ‘turf’and governance, if you can do a good enough job on defining the standards and templates it should be possible to create a secure, automated distributed solution without the need to have a single central authority and administrative effort. It would otherwise be hard to do this by piecemeal capture and shovelling of data between jurisdictions…. there would need to be a shared trusted platform, including reference data and a ‘trust framework’for identities of legal entities. The tough bit is in designimg and agreeing all this, of course. The technology is not the problem here…. maybe it could imspire some creative re-thinking of the ‘default’ centralized governance methods and parties?

Andrew Baker - 14.06.17

Nick this is a very good point and not one Richard or I have discussed in detail, but we have maybe touched on it in passing. The idea of a shared trust platform is a very good one. I will flag this to Richard.

Richard Murphy - 19.06.17

Consider it mentioned.

Thanks Nick

This may be discussed at the Tax Justice Network conference next month where this paper will be presented by one, either or both of us.

Post a comment