US Bureau of Economic Analysis data on effective tax rates in Ireland – An analysis.
Summary
A recent IIIS paper concluded that subsidiaries of US multinationals operating in Ireland have an effective tax rate of 2.2%. This is based on calculations derived from US Bureau of Economic Analysis (“BEA”) data. The BEA data is fatally flawed as a source for calculating an effective tax rate for US owned companies operating in Ireland. Critically it conflates the concepts of place of incorporation and proper locus of income for corporation tax purposes. Irish Revenue statistical reports confirm that the denominator in the effective tax rate calculation of 2.2% in the IIIS Paper is overstated by several hundred percent.
The 2012 Irish Revenue statistical report indicates an effective tax rate of at least 12.24% for companies in Ireland – this is likely to be somewhat understated for reasons outlined in this paper.
The BEA produces financial data on the overseas subsidiaries of US multinationals. This includes data on corporate taxes paid on the profits of those companies and data on the total income of those companies. That data has been used to calculate what are presented as “effective tax rates” for “US subsidiaries operating in Ireland” and to claim that such subsidiaries have “the lowest effective tax rate in the EU at 2.2%” a rate which is described as “not that dissimilar to effective tax rates in countries generally regarded as tax havens such as Bermuda”. This has directly led to international press coverage claiming that this demonstrates that Ireland is a tax haven e.g. “Why it’s completely accurate to call Ireland a tax haven”, Bloomberg 11 February 2014 by Jonathon Weil.
The purpose of this paper is to analyse the BEA data and whether it supports the contention that Ireland offers an effective 2.2% tax rate to US subsidiaries operating in Ireland thereby meriting the designation of “tax haven”.
1. Locus of corporate income for tax purposes
In order to calculate an effective tax rate for companies operating internationally one needs to determine what is the proper locus of the income for tax purposes i.e. is it Irish income, Bermudan income, US income etc.? It is only income which is Irish locus, under generally accepted norms of international tax law, and which therefore is properly within the Irish territorial charge to tax that ought to be included in the denominator in calculating an effective tax rate.
In order to properly analyse where is the proper locus of corporate income for tax purposes an understanding of the international tax rules on corporate tax residence, taxable branches etc. is required.
2. Where is a company tax resident?
One of the earliest and most basic questions of taxation is where is a company tax resident and following from this which country ought to tax its income.
A company may be “incorporated” (i.e. be created by filing documents with the companies office) in, say, Ireland but be “managed and controlled” (i.e. have its management personnel taking decisions) in, say, Bermuda. Is the company tax resident in Ireland or in Bermuda?
In the UK the issue was settled as long ago as 1905 in the De Beers case [3] . The judgement in that case stated:
“In applying the conception of residence to a Company, we ought, I think, to proceed as nearly as we can upon the analogy of an individual. A company cannot eat or sleep, but it can keep house and do business. We ought, therefore, to see whether (sic) [recte where] it really keeps house and does business. An individual may be of foreign nationality, and yet reside in the United Kingdom. So may a Company. Otherwise, it might have its chief seat of management and its centre of trading in England, under the protection of English law, and yet escape the appropriate taxation by the simple expedient of being registered abroad and distributing its dividends abroad. The decision of Chief Baron Kelly and Baron Huddleston, in the Calcutta Jute Mills v Nicholson and the Cesena Sulphur Company v Nicholson, now thirty years ago, involved the principal that a Company resides, for purposes of Income Tax, where its real business is carried on. Those decisions have been acted upon ever since. I regard that as the true rule; and the real business is carried on where the central management and control actually abides.” (emphasis added)
It can be seen that the court determined that the place of central management and control rather than incorporation was the preferable criteria for determining a company’s residence and that they were heavily influenced in coming to this conclusion by the stronger protection that a management and control test gave against tax avoidance.
Ireland inherited the UK tax system on gaining independence in 1922 and, consistent with international taxation principles, maintained management and control as its principal determinant of corporate tax residence. There have been two changes since 1922 to tighten up the corporate tax residence rules. Firstly, in 1999 certain Irish incorporated companies controlled from non treaty countries or with no connection with Ireland were deemed to be Irish resident – this was an ultimately successful attempt to discourage the use of such companies by groups with no connection whatsoever with Ireland where Ireland would have no visibility on the nature of the activities undertaken by the company concerned. Secondly, last year Ireland moved to eliminate the possibility of companies being resident nowhere by deeming Irish incorporated companies to be resident in Ireland if they are not resident elsewhere.
To this day the place of management / control is the preferred international norm for determining corporate tax residence. The OECD’s model tax treaty, which forms the basis of the vast majority of the world’s tax treaties, has for decades provided that the place of effective management ought to be the deciding factor in the case of two countries claiming taxing rights over the same company.
The US tax code is an outlier in having a pure focus on place of incorporation in determining the tax residence of a company and in ignoring the place of management. This is consistent with US provisions on taxation of individuals under which the US taxes its citizens even though they reside abroad – the US is again an outlier in this respect.
The BEA data reflects the US peculiar focus on place of incorporation and as we will see this means great care needs to be taken in interpreting the BEA data.
3. Branches
Determining the tax residence of a company is not necessarily the end of the analysis of where a company ought to be taxed. A company might be resident in, say, Country A, but have a “branch” [4] in, say, Ireland. In that instance international norms of taxation law would provide that Country A would have the right to tax the world wide profits of the company and Ireland would have the right to tax the Irish branch profits. To avoid double taxation of the Irish branch profits Country A might either grant a credit for Irish tax paid against Country A tax due or, alternatively, completely exempt the Irish branch profits from Country A tax – this might be done unilaterally under Country A’s domestic tax law or under the terms of a bilateral tax treaty between Country A [5] and Ireland [6] .
4. Territorial scope of Irish corporate taxation
Irish taxation law taxes the worldwide profits of Irish resident companies and the profits of Irish branches of non Irish resident companies. It also taxes some Irish source profits of non Irish resident companies – in particular rents or capital gains deriving from Irish land and buildings. This is consistent with the territorial scope rules of most other countries in the world [7] and with the principles set out by the OECD for the taxation of international profits. It is also consistent with the principles applied in Ireland, and in most countries in the world, for the taxation of individuals.
The place of incorporation of a company is largely irrelevant in determining its taxation status under Irish law as the country of tax residence is primarily based on the place of management and control rather than the place of incorporation. The following examples [8]illustrate this:
- An Irish incorporated company which is non Irish resident (because it is not managed and controlled in Ireland), which has no Irish branch and which has no Irish source rental income etc., will generally not be subject to Irish tax – in the same way that an Irish citizen resident abroad who has no Irish source income will generally not be subject to Irish tax.
- A foreign incorporated company which is Irish resident (because it is managed and controlled here) will be subject to Irish tax on its worldwide income.
- A non Irish incorporated, non Irish resident company with an Irish branch will be subject to Irish tax on its Irish branch profits.In none of the above cases could you get a meaningful measure of the effective tax rate in Ireland by calculating the total tax paid by such companies divided by their profits and attributing that to the country of incorporation – which is essentially the method used in the IIIS Paper.
5. BEA data collection methodology
The BEA data collection methodology means that its findings are fatally flawed in drawing conclusions about Irish effective rates of corporation tax. The BEA generally collects data on the basis of place of incorporation. In almost all circumstances an Irish incorporated company has to file BEA data as an “Irish” company even though it may not be Irish tax resident, may not have an Irish branch and may have no profits subject to Irish tax – for example even €1 of sales from Irish customers would be enough to cause the company to have to file as “Irish”.
This means that Irish incorporated companies that are not tax resident in Ireland will file as “Irish”. To include such companies in calculating Irish effective corporation tax rates is as misleading as if one was to include Irish citizens working abroad in calculating Irish effective income tax rates. It should be borne in mind that the use of Irish incorporated non Irish resident companies by US multinationals is common for US groups [9]and therefore the distortive effect on the statistics gathered for Ireland is particularly pronounced.
Furthermore non Irish incorporated companies that are tax resident in Ireland, or which have a taxable branch in Ireland, and which consequently are subject to tax in Ireland may be reported by the BEA as non Irish. It should be borne in mind that the usage of non Irish incorporated Irish resident companies is increasingly common as some aspects of foreign company law [10] (as distinct from tax law) are favoured by some international groups and therefore the distortive effect on the statistics gathered for Ireland may be particularly pronounced.
6. IIIS Paper
The IIIS Paper which calculates an effective Irish tax rate of 2.2% for US owned subsidiaries in Ireland uses as its denominator the “Irish” profits number from the BEA data. As outlined at 4. above this number is fatally flawed as it heavily based on place of incorporation rather than proper locus of income for tax purposes.
Other, less significant, flaws in the IIIS Paper include:
- In calculating effective tax rates the author has used as his denominator post tax profits rather than pre tax profits – this is contrary to his own stated methodology[11] and indeed any sensible methodology. The effect of this, for example, is that the French effective tax rate is stated as 35.9% whereas it ought, even on the author’s own methodology, to be 26.5% [12] .
- The German effective tax rate given of 20% does not compute on any basis – using pre tax profits as the denominator (the more sensible method) gives 21.7% while using post tax profits as the denominator (as the author has done in all other cases) gives 27.7%. [13]
The denominator included in the calculation of the supposed effective 2.2% tax rate includes “income from equity investments”. The Paper acknowledges that this may include income on which tax “may already have been paid”. In fact, in the vast majority of cases income from equity investments is highly likely to have suffered either Irish or foreign tax already as Ireland is not generally used as a holding company location for low taxed subsidiaries [14] . Whilst the paper provides a calculation of an effective tax rate adjusted for this of 3.8% (which is itself grossly understated for the reasons outlined in this paper) it only mentions the 2.2% rate in its summary and conclusion thereby giving the false impression that the lower number is more accurate. It is not surprising therefore that it is the 2.2% rate that has received the greatest amount of press coverage.
7. Irish Revenue Statistical Report
The Irish Revenue produces statistical reports which do not have the flaws contained in the BEA data. This allows us to calculate a far more accurate effective tax rate for Irish companies. Their 2012 statistical report gives data for 2011 – the same period for which the IIIS paper claims an effective 2.2% rate for Irish subsidiaries of US multinationals.
The Irish Revenue statistical report includes all income properly within the territorial charge to Irish corporation tax. This includes:
- (i) worldwide income of all Irish resident companies (irrespective of whether they are incorporated in Ireland or abroad).
- (ii) income of Irish branches of non Irish resident companies (irrespective of whether they are incorporated in Ireland or abroad).
- (iii) certain Irish source income of non Irish resident companies which do not have branches here (irrespective of whether they are incorporated in Ireland or abroad).
Unlike the BEA data it does not include income of Irish incorporated non resident companies falling outside categories (ii) and (iii) above.
These factors help explain why the BEA data reports US multinational’s total “Irish” corporate income as $147bn [15] in 2011 whereas the Irish Revenue reports income ofall (i.e. including non US owned) “Irish” corporate income as €40bn [16]. When we consider the size of the numerical difference and the fact that a significant proportion of the €40bn number relates to non US owned companies we can see that the IIIS paper has overstated the denominator required to calculate Irish effective tax rates by several hundred per cent.
Based on the Irish Revenue data we calculate an effective Irish tax rate for 2011 of 12.24% which for the reasons stated below we believe to be somewhat understated.
The numerator in our calculation is €4,902.9m. Key points are:
- This is the total corporation tax liability for 2011 for all Irish companies [17] as adjusted for items mentioned below.
- The number is before double taxation relief of €567.1m. In our view, it is not appropriate to reduce the effective tax rate for double taxation relief. Double taxation relief represents relief for taxes borne elsewhere i.e. this tax has actually been paid albeit to perhaps another country. The Irish double taxation relief system does not always give full relief for double tax suffered and the data on excess foreign tax credits is not available – if known this would increase the effective tax rate.
The number is after credit for Research and Development tax credits of €258.5m – this reduces the effective tax rate.
The denominator is Revenue’s total corporate taxable income number of €40,062.9m. This is the net number after deduction of capital allowances, trade charges, losses forward etc. These are normal deductible expenses in arriving at profits and are therefore properly deductible in calculating the denominator for an effective tax rate calculation [18] .
Some timing differences for tax deductions can arise as items such as capital allowances may be deducted in a different period for tax purposes than for accounts purposes but these differences reverse over time and should generally be smoothed out over the entire population of Irish companies.
The calculation should be adjusted for permanent differences. There are a number of permanent differences which the Revenue report does not provide data on – these would increase the effective tax rate if known e.g.:
- (i) Depreciation on a number of significant expenditures is not deductible for tax purposes e.g. depreciation on office buildings or retail premises.
- (ii) Certain entertainment, motor expense and non trade related expenses are non deductible.
- (iii) Unutilised losses are not included in the data and therefore the aggregate profit for the corporate sector is overstated. The amount of unutilised losses could be significant.
The €40bn number seems to be after expenses outlined at (i) and (ii) have been “added back” and before unutilised losses have been deducted – therefore our denominator is somewhat overstated. This together with the lack of data on excess foreign tax credits means that our effective tax rate of 12.24% is likely understated.
A significant permanent difference for which data is provided in the Revenue report is the Research and Development tax credit. This reduces companies’ effective tax rate and is adjusted for in the numerator as outlined above. Research and Development credits are a common feature of international tax systems and Ireland is not unusual in granting them.
The effective tax rate of 12.24% can be broadly rationalised as the trading rate of 12.5% plus an additional €265m on profits taxable at the higher 25% rate for non trading profits[19] broadly cancelled out by the €258.5m cost of the Research and Development tax credit in 2011.
8. Conclusion
The BEA data is fatally flawed in calculating an effective tax rate for US owned companies operating in Ireland. Critically it conflates the concepts of place of incorporation and proper locus of income for tax purposes. Irish Revenue statistical reports confirm that the denominator in the effective tax rate calculation of 2.2% in the IIIS Paper is overstated by several hundred percent.
The 2012 Irish Revenue statistical report indicates an effective tax rate of at least 12.24% – this is likely to be somewhat understated for reasons outlined at 7. above.
Please see www.bjdennehy.ie/taxation for further details.
B.J. Dennehy & Company, Accountants Limerick.