UPDATE: February 11, 2016
Below is my January 31, 2016 blog about the Google Tax row. I thought it would be useful to retweet it in advance of today’s PAC meeting, which I am looking forward to live streaming later today. The deal still seems to rest on shaky legal foundations, carrying the risk of undermining tax revenue generation in the long run. Nothing about HMRC’s statement of a couple of days ago has convinced me to change that opinion – it came across as being as sincere as a toddler proclaiming, through a mouth jammed with sweeties, that they haven’t eaten any sweeties and that they are really, really, telling the truth. Neither has this impression been helped by reports of protestations from Google that sounded as if they had refused to pay tax on sales since it is the law that taxes are based on income and not sales. This is absolutely correct of course. Nevertheless, the suspicion that they might have refused to pay taxes on sales leaves me wondering if they were asked to pay taxes on sales, and if they were, who was it that did the asking, and so on and so forth, and (I’m just saying) how this might be interpreted through the lens of the principle of legal certainty. In any case, whether or not HMRC and Google are tax sweethearts in reality, there is a perception that they are and this is not going to be improved unless there is complete transparency about how the details of how these tax deals are agreed and trust that the law was applied as it should be. I think that a bit of independent oversight is necessary – as suggested by Judith Freedman in her FT article of a few days ago. In the meantime, let PAC at it.
PS: If one has robust legal rules for paying taxes, one should be able to assess tax liabilities unilaterally without any consultation with the taxpayer. If a tax administration agency has to have a meeting with the taxpayer to get him/her to pay and/or explain why they have to, then the legal foundations for the tax liability cannot be thought of as being robust, but rather shaky in substantive terms, even if nobody has broken any rule or any law.
IT’S ALL ABOUT THE BASE: January 31, 2016
HMRC has made a tax deal with Google, a company worth billions of pounds, which allows them to pay £130 million in tax arrears. There is an unhealthy lack of transparency in the details of how HMRC came up with the £130 million figure and this has served to fuel the raging debate currently being played out in the media about whether or not Google, a multibillion pound corporation, has received special treatment from the tax authorities. Some commentators have reported that the deal amounts to Google paying an effective corporate tax rate of 3%, which is much less than the 20% statutory rate (see this report). My conclusions are that the deal seems to rest on shaky legal foundations and carries the risk of risk undermining tax revenue generation in the long run.
The starting place in thinking about all of this is to understand the principle that determines which tax jurisdiction profits are generated in. For a fully domestic firm, sales and production both take place within national boundaries, but this is not true for a multinational firm, which produces its goods and services in one country (or countries) and sells them in a different country (or countries). This raises the issue of what basis should be used to assess profit taxes of multinational corporation: the basis of the tax jurisdiction where sales take place, or the basis of the tax jurisdiction where the income is generated (i.e., where production takes place). The prevailing principle that all countries use is to base profit taxation on the location where income is generated and not on the location where sales takes place. This means that if a business sells 50% of its goods in the UK but produces all of these goods outside of the UK, then that business is not liable to pay any corporate taxes to the UK government. It does not mean that these goods escape taxation since there are taxes that need to be paid on imported goods, but those are sales taxes, not profit taxes.
Implementing profit taxation based on the location of where its income is generated is not always straightforward. Businesses that own foreign subsidiaries can use internal artificial accounting conventions (transfer pricing) to manipulate the amount of profit that they report in each tax jurisdiction, and even though there are rules in tax law that are specifically aimed at containing transfer pricing abuses, it is very difficult to monitor and identify them. Nevertheless, even with the risk of transfer pricing abuse, which is fundamentally an accounting concern, the general principle that profit taxes are based on the location where income is generated rather than where sales are generated remains. There are two main reasons for this.
First, there are efficiency reasons. Profit taxation based on the location of sales creates distortions in firm behaviour. This is economically inefficient in and of itself, but it can also lead tax jurisdictions to engage in welfare worsening tax competition. Consider the following example: Suppose there is a Kazakhstan exporter of widgets that produces widgets in its Kazakhstan factory and sells half of what it produces to the UK. If it does not own any subsidiary in the UK, then according to the general rules, it is not liable to pay any profit tax in the UK. Now suppose that the same firm, without changing anything about its production – widgets are still being produced in its factory in Kazakhstan – is thinking of opening a tiny one room office in the London that is staffed by one person. Maybe it wants to do this because it thinks it would be helpful with local promotion of the widgets and maybe, for legal reasons it would be convenient to have this office take the legal status of a fully owned UK subsidiary. If profit taxes were based on sales, then this small change, which has no marginal implications for how the Kazakhstan widget business is run, would trigger a 50% profit tax liability. Thus, it is pretty clear that the Kazakhstan firm would not entertain the notion of opening a subsidiary in the UK for very long. It would just not do it. In general, a legal tax regime based on sales and not income generation, would be unsustainable since it would generate large distortions in the behaviour of firms who would make decisions about whether or not to open local subsidiaries on the basis of tax convenience and not for business reasons. It would also trigger disputes between tax jurisdictions, and encourage welfare worsening tax competition between them. In the example, if the Kazakhstan widget producer did have a subsidiary in the UK, but then experienced a higher tax rate in the UK in comparison with China, it would simply shut-down its UK subsidiary and re-open it in China. This might induce the UK to offer a tax concession aimed at luring the subsidiary back, and then China would respond and a `race to the bottom’ in tax rates would be triggered.
Second, even if we wanted to depart from profit taxes based on the location of income generation, and instead assess profit tax liabilities on the basis of the location of firm sales, it would be administratively difficult to do so with sales of services. Sales of physical goods, both domestically produced and imported, are assessed for VAT tax purposes, but it is extremely difficult to do the same for imports of services. This means that in the case of services, it is not likely that there would be any reliable way of measuring domestic sales. We are left with the prevailing principle: profit taxation based on the location where the income is generated and not based on the location of where sales are made.
This is not to say that using sales as the base for taxation (not profits determined on the basis of the location of sales, which is a very different thing), might not be a good idea, and indeed sales are already used as a base for taxation — there is VAT on sales of goods and services, which is applied both to all sales of goods, whether they are domestically produced or imported. But there are problems with implementing sales taxes for services, both on the domestic front through evasion (e.g., who asks for a VAT receipt from the window washer that comes by the house occasionally?) and on the international front through problems in assessing taxes due – for imported services, there is not a physical good that crosses the border and whose value can be assessed by customs, as is the case for physical goods. If I order a Google Analytics product from the US and download it on my computer, it is quite difficult to have a system where that sale can become liable for UK VAT taxation. This is not only because it is difficult to keep track of the sale itself, but also because it is difficult to determine where the good is actually going to be used in the end. Maybe I am ordering and downloading the software while I am on holiday in France, or I could download it in the UK, but be doing it behind a VPN server, which is legal and which can mask my location (to get some idea of the complexities involved, see this guidance from HMRC on taxing digital services). This would make it very difficult for Google to collect VAT on the sale, even if they wanted to do so. Indeed at the most recent Public Accounts Committee meeting, which I attended, these issues were the topic of a lively debate that mainly missed these key points.
So admittedly, taxing sales of services is challenging, but this does not mean that one should therefore try to address the “Google problem” by distorting well established and solidly founded economic principles underlying the way that profits are taxed, especially as a knee jerk reactive scramble to find a quick remedy. In this article in the Guardian, Lord Lawson suggests taxing UK sales (the IFS suggests the same in their January 29th Observations article), and this is a reasonable thing to say; however, no reforms of the basic economic principles of taxation is implied in this, nor are they needed, because the principle of taxing the sales of goods and services is already in place. Rather, it is a problem of implementation of sales taxes that is the issue in the case of international sales of services and this needs to be the focus if the debate moves further in the direction of resorting to sales taxation.
It may be very frustrating for the public and government to see an international company such as Google, which has a large business interest in the UK, paying little taxes. But that is not a good reason for trampling over sound economic principles and for not sticking to the rules that are set out by the existing tax code. If the UK government or HMRC (rightly) feel that the current tax regime is less than satisfactory, then they should concentrate their efforts on reforming it, and endeavour to negotiate tax treaties with other tax jurisdictions – rather than negotiating what are being perceived as ad hoc deals with individual corporations. Indeed the OECD has been trying to move in this direction with its Base Erosion and Profit Shifting (BEPS) project and the EU’s recently announced Anti Tax Avoidance Package, which “contains concrete measures to prevent aggressive tax planning, boost tax transparency and create a level playing field for all businesses in the EU.’’ These initiatives provide an excellent basis for engaging in international tax negotiations.
But there is an even more fundamental reason for why these seemingly ad hoc deals are a bad idea, and that is that they seem to be based on shaky legal foundations. I cannot see how it is a good idea to improvise criteria for determining tax liabilities, which are not enshrined in the tax code (or if not improvise, be perceived as improvising), or even worse, to negotiate those criteria with individual businesses (or if not negotiate, be perceived as negotiating). This violates the basic principle of legal certainty, is rightly perceived as being unfair and unjust both by businesses (who feel unjustly targeted) and by the public (who perceive those same businesses as receiving special treatment), and creates a climate of business uncertainty that can discourage business investment in the UK. Abiding by the principle of legal certainty in tax matters is not in conflict with HMRC’s mandate to maximise its tax revenue collection applying a “risk management” approach: if one takes a long-run view of these objectives (as one should), uncertain and renegotiable tax rules carry the risk of risk undermining tax revenue generation in the long run.
I am an expert on the economics of international taxation. I have published papers on transfer pricing, international tax evasion, and international tax competition. I am also Editor-in-Chief of International Tax and Public Finance, one of the leading academic field journals in public economics, which has a special emphasis on the open economy or, more generally, interjurisdictional issues, the interaction of policies across jurisdictions and the effects of those policies on economic (and political economy) outcomes.